﻿                                   If symbols do not display correctly change
                                   your browser character encoding to unicode






CRUISE REPORT:  ARK-XXII_2
(Updated APR 2014)



Highlights


                         Cruise Summary Information

         CCHDO Section Designation  ARK-XXII_2
Expedition designation (ExpoCodes)  06AQ20070728
                           Aliases  ARK-XXII/2
                  Chief Scientists  Ursula Schauer / AWI
                             Dates  2007 JUL 28 - 2007 OCT 7
                              Ship  R/V Polarstern
                     Ports of call  Tromsø, Norway - Bremerhaven, Germany

                                                88° 40' N
             Geographic Boundaries  170° 45' W              177° 33' E
                                                72° 44' N

                          Stations  125
      Floats and drifters deployed  0
    Moorings deployed or recovered  8 meteorological buoys deployed
                                    1 Ice Mass Balance buoy deployed

                               Ursula Schauer

                          Alfred Wegener Institute
    Bussestrasse 24 • (Building F-123) • D-27570 • Bremerhaven • Germany
               Fax: +49(471)4831-1797 • Tel: +49(471)4831-1817 
                        Email: Ursula.Schauer@awi.de
























                                                                         579
Berichte                                                                 ----
zur Polar-                                                               2008
und Meeresforschung



Reports on Polar and Marine Research


The Expedition of the Research Vessel "Polarstern" 
to the Arctic in 2007 (ARK-XXII/2)




Edited by 
Ursula Schauer 
with contributions of the participants





















HELMHOLTZ                        ALFRED-WEGENER-INSTITUT FÜR
  GEMEINSCHAFT                   POLAR- UND MEERESFORSCHUNG
                                 In der Helmholtz-Gemeinschaft
                                 D-27570 BREMERHAVEN
                                 Bundesrepublik Deutschland




                                                               ISSN 1866-3192









Hinweis                                Notice
Die Berichte zur Polar- und            The Reports on Polar and Marine
Meeresforschung werden vom             Research are issued by the Alfred 
Alfred-Wegener-Institut für            Wegener Institute for Polar and
Polar-und Meeresforschung in           Marine Research in Bremerhaven*,
Bremerhaven* in unregelmäßiger         Federal Republic of Germany.
Abfolge herausgegeben.                 They appear in irregular intervals.

Sie enthalten Beschreibungen und       They contain descriptions and results
Ergebnisse der vom Institut (AWI)      of investigations in polar regions
oder mit seiner Unterstützung          and in the seas either conducted by
durchgeführten Forschungsarbeiten      the Institute (AWI) or with its 
in den Polargebieten und in den        support.
Meeren.

Es werden veröffentlicht:              The following items are published:

- Expeditionsberichte                  - expedition reports (incl. station
  (inkl. Stationslisten                  lists and route maps)
  und Routenkarten)

- Expeditionsergebnisse                - expedition results (incl. Ph.D. 
  (inkl. Dissertationen)                 theses)

- wissenschaftliche Ergebnisse         - scientific results of the Antarctic
  der Antarktis-Stationen und             stations and of other AWI research 
  anderer Forschungs-Stationen            stations
  des AWI

- Berichte wissenschaftlicher          - reports on scientific meetings
  Tagungen

Die Beiträge geben nicht               The papers contained in the Reports
notwendigerweise die Auffassung        do not necessarily reflect the
des Instituts wieder.                  opinion of the Institute.




                The "Berichte zur Polar- und Meeresforschung"
              continue the former "Berichte zur Polarforschung"




* Anschrift / Address
Alfred-Wegener-Institut                Editor in charge:
Für Polar- und Meeresforschung         Dr. Horst Bornemann
D-27570 Bremerhaven Germany 
www.awi.de                             Assistant editor:
                                       Birgit Chiaventone



Die Berichte zur Polar- und Meeresforschung (ISSN 1866-3192) werden ab 2008 aus-
schließlich als Open-Access-Publikation herausgegeben (URL: http://epic.awi.de)

Since 2008 the 'Reports on Polar and Marine Research (ISSN 1866-3192) are 
only available as web based open-access-publications (URL: http://epic.awi.de






The Expedition of the Research Vessel "Polarstern" 
to the Arctic in 2007 (ARK-XXII/2)


Edited by 
Ursula Schauer 
with contributions of the participants



































Please cite or link this item using the identifier
hdl: 10013/epic.30947 or http//hdl.handle.net/10013/epic.30947
ISSN 1866-3192




























                                  ARK-XXII/2




                        29 July 2007 - 7 October 2007
                             Tromsø - Bremerhaven




                       Fahrtleiter / Chief Scientist:
                                Ursula Schauer


                          Koordinator / Coordinator:
                              Eberhard Fahrbach






























CONTENTS


1.  Expedition ARK-XXII/2: Fahrtverlauf und Zusammenfassung                 
          
    Summary and itinerary      

2.  Weather conditions  

3.  Sea ice properties    
    3.1  Sea ice thickness measurements  
    3.2  Sea ice radar backscatter measurements for
         improved melt-pond and thin-ice cover analysis  
    3.3  Routine sea ice observations    
    3.4  Buoy deployments    
    3.5  Sea ice biology    

4.  Oceanography    
    4.1  Physical oceanography  
    4.2  XCTD observation    
    4.3  Deployment of ice-tethered buoys    
  
5.  GEOTRACES    
    5.1  A- trace elements  
    5.2  B- natural and anthropogenic radionuclides  
    5.3  C- related parameters    
    5.4  Coupling of methane and DMSP cycles in the
         marginal ice zone and on polar shelves  

6.  Marine biology  
    6.1  Zooplankton investigations  
    6.2  Biodiversity of polar deep-sea eukaryotic microbiota
         - molecular versus morphological approach  

7.  Marine geology  
    7.1  Parasound sediment echosounding  
    7.2  Bathymetry  
    7.3  Geological sampling  
    7.4  Physical properties  

APPENDIX  
A.1  Participating institutions  
A.2  Cruise participants  
A.3  Ship's crew  
A.4  Station list (see PDF)
A.5  Annex coring positions (see PDF)
A.6  Sediment core descriptions (see PDF)














1.  EXPEDITION ARK-XXII/2: FAHRTVERLAUF UND ZUSAMMENFASSUNG
    Ursula Schauer 
    Alfred-Wegener-Institut für Polar- und Meeresforschung


Der zweite Abschnitt der 22. Arktisexpedition des Forschungsschiffes 
Polarstern, ARK-XXII/2, war ein zentraler Beitrag zum Internationalen Polar-
jahr 2007/08 (IPY 2007/08). Er trug insbesondere zu zwei im 
IPY-Wissenschaftsplan aufgeführten Zielen bei:


   "1. Status: to determine the present environmental status of the polar 
       regions" und
   "2. Change: to quantify, and understand, past and present natural 
       environmental and social change in the polar regions; and to improve 
       projections of future change" (The Scope of Science for the 
       International Polar Year, http://www.ipy.org)


Zurzeit finden in der Arktis drastische Veränderungen statt: Das Meereis 
verringert sich, die oberen Wasserschichten werden wärmer und Strömungen 
verschieben sich. Daraus sind Auswirkungen auf den Austausch und den 
Transport von Stoffen und auf Ozean- und Eisorganismen zu erwarten. Für das 
Verständnis dieser Veränderungen ist eine umfassende Gesamtaufnahme als 
Ausgangspunkt für Langzeitbeobachtungen notwendig. Gemeinsam mit anderen 
arktischen Expeditionen im IPY 2007(1) diente ARK-XXII/2 diesem Ziel und 
übernahm dabei insbesondere die Erfassung der eurasischen und der zentralen 
Arktis. Gleichzeitig galt die Reise der Untersuchung von biogeochemischen 
Stoffkreisläufen und der Analyse von Ökosystemen in Eis und Ozean, sowie der 
quartären Vereisungsgeschichte des östlichen und zentralen Nordpolarmeers.

ARK-XXII/2 war eingebunden in die IPY-Projekte SPACE (Synoptic Pan-Arctic 
Climate and Environment Study, IPY-EoI #18), GEOTRACES: Spurenstoffe in der 
Arktis (IPY-EoI #45) und iAOOS (Integrated Arctic Ocean Observing System, 
IPY-EoI #80) (siehe http://www.ipy.org/development/eoi/index.htm), und 
lieferte einen Beitrag zu dem deutsch-russischen Projekt VERITAS (Variability 
and Export of Riverine Matter into the Arctic Ocean and late (Paleo-) 
Environmental Significance). Gleichzeitig ist ein Großteil der Arbeiten 
Bestandteil des durch die EU geförderten Integrated Programmes DAMOCLES 
(Developing Arctic Modelling and Observing Capabilities for Long-term 
Environment Studies).

Um dekadische Veränderungen zu erfassen, wurden hydrographische Schnitte, 
Zoo-planktonbeprobungen und Eisdickenbeobachtungen früherer Expeditionen, wie 
Oden 1991, Polarstern 1993, Polarstern 1995 und Polarstern 1996 wiederholt. 
Auf diese Weise können räumliche und zeitliche Variabilität unterschieden und 
damit die Entwicklung ozeanographischer, eisphysikalischer und biologischer 
Parameter über eine Dekade erfasst werden. Zusätzlich leistete ARK-XXII/2 
einen Beitrag zu einem internationalen Langzeitbeobachtungsprogramm von Ozean 
und Meereis durch Eisbojen, die in diesem Jahr auf verschiedenen Expeditionen 
erstmalig in großem Umfang arktisweit ausgebracht wurden.



___________________________
(1) NABOS (RV Victor Bujnitzky); LOMROG (IB Oden); AGAVE (IB Oden); Drift des 
    französischen Schiffes TARA und der russischen Eisstation NP 35; 
    AARI-Expedition (Academic Fedorov); Beaufortwirbel (IB Louis St.Laurent); 
    Transdrift XII (RV Ivan Petrov)


Ein wesentlicher Bestandteil der Reise war ein großes Chemieprogramm im 
Rahmen von GEOTRACES. Dabei kam erstmalig ein Ultra-clean-System zum Einsatz, 
mit dem in großem Umfang effektiv Wasserproben für Spurenmetalluntersuchungen 
genommen werden können. Parallel dazu wurde ein großes Spektrum von 
natürlichen Radioisotopen für Partikelflussuntersuchungen beprobt.

Die ursprüngliche Planung sah vor, dass ARK-XXII/2 vorwiegend den 
eurasischen Sektor der Arktis abdeckt. Die unerwartet niedrige Eisbedeckung 
des Sommers 2007 erlaubte jedoch, die Schnitte bis weit ins amerasische 
(kanadische) Becken hinein auszudehnen. Auf der anderen Seite zwang die weit 
nach Norden zurückgezogene Eisgrenze dazu die Bojen sehr viel weiter im 
Nordwesten auszulegen als eigentlich vorgesehen war.

Die Schnitte erstreckten sich von den Schelfgebieten der Barents-, der Kara- 
und der Laptewsee über das Nansen-, das Amundsen- und das Makarowbecken bis 
über den Alpha-Mendelejewrücken in das Kanadabecken. Auf allen Schnitten 
wurden in engem Stationsabstand CTD-Profile (Temperatur, Salzgehalt, 
Sauerstoffgehalt und Fluoreszenz) aufgenommen und eine Kombination von 
Standardproben genommen. In unregelmäßigen Abständen wurden zusätzlich 
Messungen und Beprobungen zur Dicke und zu Radar-Rückstreueigenschaften, 
sowie zu Organismen des Meereises, zur Verteilung von natürlichen 
Spurenstoffen und Radioisotopen im Ozean und im Eis und zur Verbreitung und 
der Anpassung von Zooplankton und zur Biodiversität polarer 
Tiefsee-Eukaryoten durchgeführt. Die physikalischen Eisuntersuchungen wurden 
vorwiegend als umfangreiche Fernerkundungsmessungen vom Hubschrauber aus 
durchgeführt und durch Arbeiten direkt auf dem Meereis ergänzt. In allen 
Becken, auf den Rücken und auf dem Karaseeschelf wurden Sedimentproben zur 
Bestimmung der spätquartären Veränderlichkeit des Flusswasserausstroms und 
der Vereisungsgeschichte der Arktis genommen.

Polarstern lief planmäßig am 28. Juli 2007 aus Tromsö aus. 54 Wissenschaftler 
aus 19 Instituten in 10 Ländern befanden sich an Bord. Am 29. Juli begannen 
wir auf der Zentralbank in der Barentssee den ersten Transekt, der entlang 
340 E nach Norden ins Nansenbecken führte. Bei etwa 81° 30'N trafen wir auf 
die Eisgrenze und bei 84° 30'N war das Packeis so dicht, dass ein Fortkommen 
nahezu unmöglich war. Wir brachen daraufhin am 6. August den Schnitt ab, um 
nach Osten zum zweiten Schnitt entlang dem 6lsten östlichen Längengrad zu 
gelangen. Auf dem Weg dorthin wurde die Eisbedeckung lockerer. Am 10. August 
begannen wir den zweiten Schnitt von 84° 40'N aus nach Süden in Richtung 
Franz-Joseph-land. Nach den vorliegenden topographischen Daten, die sich oft 
als fehlerhaft erwiesen, sowie nach Parasound- und Hydrosweepsurveys wurden 
Positionen für Sedimentproben im tiefen Nansenbecken, am Hang und auf dem 
Schelf bestimmt und geologische Kerne gezogen. An den Kernpositionen und 
dazwischen wurden wieder hydrographische Stationen und Netze für 
Zooplanktonproben gefahren. Um Zeit für die Stationen in der zentralen Arktis 
zu sparen, entschlossen wir uns zu relativ großen Stationsabständen und 
füllten die Zwischenräume mit XCTD-(Expendable CTD) Messungen, um die 
kleinräumigen Wassermassenstrukturen auflösen zu können.

Die Überfahrt auf den dritten Schnitt nordwestlich von Sewernaja Semlja 
führte durch sehr lockeres Eis in Sichtweite der Eiskante. Am 19. August 
begannen wir im östlichen Voronintrog unseren dritten Schnitt, der zunächst 
entlang ca. 86°E wieder nach Norden führte. Wir durchquerten das 
Nansenbecken, den Gakkelrücken und das Amundsenbecken bis zum 
Lomonossowrücken. Auf dem Rücken machten wir einen Abstecher nach Norden, um 
den Tiefenwasseraustausch an der Schwelle zwischen Amundsen- und 
Makarowbecken zu untersuchen, der durch das so genannte Intrabecken führt. 
Hier erreichten wir am 31. August bei 88°38'N die nördlichste Position 
unserer Fahrt. Die ungestörte Sedimentablagerung im Intrabecken wurde für 
geologische Kernprobennahmen genutzt. Anschließend setzten wir den 
ursprünglichen Schnitt bei etwa 88°N nach Osten ins Makarowbecken fort.

Informationen über die immer weiter zurückgehende Eisbedeckung in der 
eurasischen Arktis bewogen uns, den ursprünglichen Plan für die Auslegung von 
Eisbojen zu ändern und wir brachten am 2. September bei 87°51'N 170°W die 
erste ozeanographische und eine meteorologische Eisboje aus. Gleichzeitig 
beschlossen wir, das zügige Fortkommen in der lockeren Eisbedeckung zu nutzen 
und das Makarowbecken bis zum Alpharücken bei 85°42'N 135°W zu durchqueren. 
Eine weitere Verlängerung des Schnittes brachte uns bis zur Südflanke des 
Alpharückens bei 84°30'N 138°25'W und damit an den Rand des Kanadabeckens. 
Hier schlossen wir den Schnitt endgültig mit einer ausführlichen Beprobung 
von Meereis, physikalischen und chemischen Parametern der Wassersäule, 
Zooplanktonfängen und Sedimentkernnahmen ab. Sowohl am östlichsten als auch 
am südlichsten Punkt im amerasischen (kanadischen) Becken wurde der 
CTD-Schnitt durch XCTD-Abwürfe vom Helikopter aus um jeweils weitere 70 nm 
nach Osten bzw. nach Süden verlängert. Vom südlichen Alpharücken führte ein 
kurzer Schnitt über einen der Durchlässe zwischen Makarowbecken und 
Kanadabecken zum östlichen Ende des Mendelejewrückens.

Auf dem Weg zurück nach Westen stand im Vordergrund unserer Arbeit, geeignete 
Eisschollen für wenigstens zwei weitere Bojenarrangements zu finden. Um im 
halbwegs dichten Eis zu bleiben, konnte der Schnitt nur etwa 100 nm südlich 
vom vorigen Schnitt verlaufen, weshalb wir recht große Stationsabstände 
wählten. Die zweite ozeanographische Boje wurde am 10. September im 
Makarowbecken bei 86°38'N 177°33'E ausgebracht. Ein abschließender Test 
zeigte aber, dass die Boje nicht funktionierte. So wurde sie wieder 
aufgenommen und durch eine andere ersetzt und durch eine meteorologische Boje 
ergänzt. Nach einem weitgehend aus XCTDs bestehenden Schnitt über den 
Lomonossowrücken wurde am 13. September im Amundsenbecken eine weitere 
Bojengruppe ausgebracht, diesmal bestehend aus einer CTD-Boje, einer 
Strömungsmesserboje, einer Boje zur Turbulenzmessung unter dem Eis, einer 
Eismassenbilanzboje und einer Webcam. Im Abstand von etwa 50 nm um die 
ozeanographischen Bojen herum wurden per Hubschrauber 6 meteorologische Bojen 
ausgebracht.

Der Schnitt führte dann mit weiten Stationsabständen zurück zum Gakkelrücken. 
Beim ersten Kreuzen des Gakkelrückens im August bei 90°E hatten wir Anomalien 
der Temperatur und verschiedener chemischer Parameter gefunden, die eine 
hydrothermale Quelle vermuten ließen. Aufgrund der zeitlichen Verzögerung 
dieser Entdeckung durch die Dauer der Laboranalysen konnten wir die erste 
Fundstelle jedoch nicht ausgiebiger beproben. Deshalb fuhren wir ab dem 16. 
September einen Schnitt von 84°41'N nach Süden den Gakkelrücken entlang, um 
bei einem weiteren Anzeichen einer Anomalie eine umfangreiche chemische 
Beprobung vorzunehmen. Jegliche weitere Anzeichen solcher Anomalien waren 
jedoch nur sehr schwach. Am 19. September erreichten wir bei 82°12'N die 
Eisgrenze und kurz darauf das südliche Ende des Gakkelrückens. Unser Schnitt 
führte nun über den Kontinentalhang in die Laptewsee, wo das Stationsprogramm 
am 24. September bei 75°12'N, 121° endete und die Rückreise durch die 
Nordostpassage angetreten wurde. Am 30. September haben sich bei etwa 72°N 44 
Täuflinge der Polartaufe unterzogen. Am 7. Oktober 2007 lief Polarstern um 13 
Uhr in Bremerhaven ein.


Abb. 1: Fahrtroute ARK-XXII/2
Fig. 1: Cruise track ARK-XXII/2








SUMMARY AND ITINERARY

The expedition ARK-XXII/2 was a central contribution to the International 
Polar Year 2007/08 (IPY 2007/08). In particular it served two objectives 
formulated in the IPY science plan:


   "1. Status: To determine the present environmental status of the polar 
       regions" and
   "2. Change: To quantify, and understand, past and present natural 
       environmental and social change in the polar regions; and to improve 
       projections of future change" (The Scope of Science for the 
       International Polar Year; http://www.ipy.org)


Currently enormous changes take place in the Arctic Ocean: the sea ice is 
shrinking, the upper water layers become warmer and ocean currents are 
shifting. This will have consequences for ocean-atmosphere fluxes, for the 
oceanic and ice-related transport of substances, and for marine and 
ice-related organisms. Ultimately it will feed back to sub-polar climate. To 
understand these changes a comprehensive pan-Arctic survey is necessary as a 
benchmark for long-term observations and as a constraint for climate models.

ARK-XXII/2 aimed at meeting its objectives in co-operation with other IPY 
expeditions(1) thereby focussing on the Eurasian and central Arctic Ocean. 
Besides the physical system ARK-XXII/2 addressed biogeochemical tracer 
studies, the ecosystems in ice and ocean as well as the history of 
quarternary glaciations of the Siberian Arctic.

ARK-XXII/2 contributed to the IPY projects SPACE (Synoptic Pan-Arctic Climate 
and Environment Study, IPY-EoI #18), GEOTRACES: Geotraces in the Arctic 
(IPY-EoI #45), iAOOS (Integrated Arctic Ocean Observing System, IPY-EoI #80) 
(see http://www.ipy.org/development/eoi/index.htm), as well as to the 
German-Russian project VERITAS (Variability and Export of Riverine Matter 
into the Arctic Ocean and late (Paleo-) Environmental Significance) which is 
listed on governmental level in the bilateral research programme. Part of the 
work was funded through the EU Integrated Programme DAMOCLES (Developing 
Arctic Modelling and Observing Capabilities for Long-term Environment 
Studies).

In the context of the IPY programmes, ARK-XXII/2 took its share in covering 
part of the Eurasian sector of the Arctic. To identify decadal change one has 
to distinguish spatial and temporal variations of hydrography, sea ice, as 
well as biological and biogeochemical parameters. Therefore the expedition 
was designed to repeat largescale sections that were made in the nineties of 
the last century, such as Oden 1991, Polarstern 1993, 1995 and 1996. To 
ensure year-round observations of ice and upper ocean a number of ice-tethered 
buoys were deployed.

A large part of the programme was dedicated to GEOTRACES. In this context for 
the first time an Ultra Clean System was employed that enabled a systematic 
survey of trace metals in the Arctic. GEOTRACES also included sampling of a 
large spectrum of natural radio isotopes for particle flux studies.


__________________________
(1) NABOS (RV Victor Bujnitzky); LOMROG (IB Oden); AGAVE (IB Oden); Drift des 
    französischen Schiffes TARA und der russischen Eisstation NP 35; 
    AARI-Expedition (Academic Fedorov); Beaufortwirbel (lB Louis St.Laurent); 
    Transdrift XII (RV Ivan Petrov)


ARK-XXII/2 was originally planned to survey the Eurasian part of the Arctic. 
However, the extremely low ice cover in 2007 enabled us to extend the 
sections far into the Canadian Basin. On the other hand the unusual low ice 
cover constricted the deployment of the ice-tethered buoys and forced us to 
deploy the buoys much further downstream in the Transpolar Drift than it was 
intended.

The sections reached from the shelves of the Barents, Kara and Laptev Seas 
across the Nansen, Amundsen and Makarov Basins beyond the Alpha Ridge into 
the Canada Basin. On all sections CTD/water sampler casts (temperature, 
salinity, oxygen, fluorescence electronically and samples of nutrients and 
δ18O contents) were conducted in narrow station distances. In larger 
intervals stations were devoted to investigate thickness and back-scatter 
characteristics of sea ice, natural trace elements and radioisotopes, 
chemical composition of dissolved organic matter and their role as markers of 
water masses, and the distribution of organisms in and below the sea ice and 
in the water column and the eukaryotic diversity in the deep-sea sediments. 
North of the Kara Sea as well as along a cross basin section sediment cores 
were taken for the determination of the Late Quaternary variation of river 
runoff and of the Eurasian Arctic glaciation history. Helicopter- and 
ship-borne XCTD-casts were used to extend temperature and salinity sections in 
the Canadian Basin or to increase the spatial resolution across topographical 
features.

The cruise began on 28 July 2007 in Tromsø. 54 scientists from 19 institutes 
in 10 countries were onboard. On 29 July we started station work on the 
Central Bank in the Barents Sea and run a section along 34°E into the Nansen 
Basin. At 81°30'N we crossed the ice edge and at 84°30'N the pack ice was so 
dense that Polarstern made hardly any progress. We therefore broke off the 
section on 6 August and turned eastward heading for our second section at 
61°E. En-route the ice cover became significantly less. On 10 August we 
started at 84°40'N the section in southern direction towards Franz Josef 
Land. Guided by available bathymetry data - which turned out to be erroneous 
in large parts - and following Parasound and Hydrosweep surveys positions 
were determined for sediment coring in the deep Nansen Basin, at the 
continental slope and on the shelf. At and between the coring positions 
hydrographic and net stations were conducted.

The transit to the third section which started northwest of Severnaya Zemlya 
went through very loose ice in the vicinity of the ice edge. On 19 August we 
started the section in the eastern Voronin Trough and sailed northwards along 
86°E. We crossed the Nansen Basin, the Gakkel Ridge and the Amundsen Basin 
and reached the Lomonosov Ridge on 30 August. During the previous two weeks 
the light ice conditions had afforded fast progress so that we found time for 
a detour to the north to investigate the exchange of deep water across the 
sill of the Lomonosov Ridge between the Makarov and the Amundsen Basins. Here 
we reached with 88°38'N our northernmost position. At the sill a small basin 
(Intra Basin) is located where undisturbed sedimentation was used for 
geological coring. The section was then continued into the Makarov Basin at 
about 88°N.

Information about the ongoing seasonal retreat of the ice-cover in the 
Eurasian Basin induced us to modify our planned cruise track. Because of the 
light ice we were able to extend our cross-basin section but we also had to 
modify our plan for deployment of ice-tethered buoys. Instead of deployment 
on the upstream end of the Siberian branch of the transpolar drift we 
deployed the first oceanographic and meteorological buoy at 87°51'N 170°W. 
Then we proceeded to the Alpha Ridge at 85°42'N 135°W and extended the 
section further up to 85°23'N 136°17'W. At the easternmost as well as at the 
southernmost corners of the cruise track in the Canadian Basin, our sections 
were extended by XCTD casts. A last prolongation on September 7 brought us to 
the rim of the Canada Basin at 84°30'N 138°25'W where we took an extensive 
station to probe all parameters in ice, water and sediment. Then we turned 
west to cross one of the deep passages between the Makarov and the Canada 
Basin and reached the eastern edge of the Mendeleyev Ridge on 9 September.

During our return to the Siberian Arctic we focussed on searching suitable 
floes for at least two out of three more planned buoy arrays. The still 
retreating ice cover forced us to sail not more than 100 nm south of the 
previous section in order to meet thick and large enough ice floes. The 
second oceanographic buoy was deployed on 10 September in the Makarov Basin 
at 86°38'N 177°W. The final performance test revealed a malfunction of the 
buoy so that we recovered and replaced it (because of the reduction to three 
arrays we had a spare buoy). We crossed the Lomonosov Ridge conducting XCTD 
casts and deployed the last array, consisting of buoys that carry a CTD, a 
current meter, a turbulence meter and ice and snow mass sensors respectively 
and a Webcam. In about 50 nm distance from all oceanographic buoy arrays, 6 
meteorological buoys were brought out by helicopter.

The section was continued to the Gakkel Ridge. At the first crossing of the 
Gakkel Ridge in August at about 90°E, anomalies of temperature and of various 
chemical properties were detected that might have originated from a 
hydrothermal vent. Because of the delay of this discovery caused by the time 
needed for lab analyses this location had not been investigated in more 
detail. Therefore we surveyed the Gakkel Ridge from 84°41'N on southwards 
with CTD casts in order to detect any further anomalies in temperature or 
light transmission which then would have been sampled in detail. However, 
only very weak anomalies were observed. On 19 September we passed the ice 
edge at 82°12'N and soon after reached the southern end of the Gakkel Ridge. 
The section was continued up the continental slope to the Laptev Sea where 
the station work finished on September 24 at 75°12'N 121°E. We turned 
northwest and passed the Northern Sea Route to the western Barents Sea. On 
September 30 at about 72°N 44 candidates underwent the polar baptism. RV 
Polarstern returned to Bremerhaven on 7 October 2007.





2.  WEATHER CONDITIONS
    Manfred Gebauer and Hartmut Sonnabend 
    Deutscher Wetterdienst

When Polarstern left Tromsø on 28 July there was a low pressure system just 
over the middle of Norway. At the beginning the weather was rather smooth, 
later the ship was affected by strong easterly winds und wave heights up to 3 
m. Finally the weather grew calm under influence of a high over the Barents 
Sea.

While operating along 34°E, Polarstern was accompanied by fresh southerly 
winds from 4 to 5 Bft. The high was now stationary north of Franz Josef Land. 
Despite the high pressure the weather was changeable, prevailing conditions 
were fog or low stratiform clouds with occasional drizzle and sometimes 
danger of icing of the helicopters.

When 85°N was reached, the high near Franz Josef Land got weaker. 
Temperatures were mostly near -2°C, because low pressure systems that 
arrived from the Barents See often brought mild and moist air into the 
northern Arctic Ocean (Fig. 2.1). A new strong low arrived between Franz 
Josef Land and Severnaya Zemlya. The wind blew with 7 Bft from northwest, but 
due to the ice cover there was no heavy sea.

From then on the flight conditions were difficult due to bad visibility and 
ceiling during most of the scientific work (Figs. 2.2, 2.3). The distribution 
of air pressure with high pressure over western longitudes and low pressure 
systems over eastern longitudes of the Arctic Ocean lasted until the end of 
August. Winds were moderate with forces 3 to 5 Bft, visibility and clouds 
continued to be difficult, sometimes accompanied by freezing rain or danger 
of icing. Some small polar lows occurred.

During September the pressure distribution changed. The area around the North 
Pole was more and more influenced by high pressure, but still there were some 
small polar lows with intermediately stronger winds and dense snowfall. 
Temperatures were partly less than -5°C, the wind continued to blow mostly 
from southerly to southeasterly directions with 2 and 5 Bft (Figs. 2.4, 2.5).

At mid of September the ridge of high pressure above Polarstern was reduced 
by new weather fronts, accompanied again by fog. When arriving in the south 
of the Laptev Sea, a storm reached us from the Barents Sea. The air 
temperature rose and on the rear of this low was a strong storm blowing from 
northwest.

The journey through the Kara and the Barents Sea and along the Norwegian 
coast was partly laborious due to several stormy lows. Later the weather was 
fine up to the port of Bremerhaven.


Fig. 2.1: Time series of air temperature
Fig. 2.2: Distribution of Visibility
Fig. 2.3: Distribution of Ceiling
Fig. 2.4: Distribution of Wind Direction 31.07.- 24.09.07
Fig. 2.5: Distribution of Wind Force 31.07. - 24.09.07





3.  SEA ICE PROPERTIES
    Stefan Hendricks(1), Stefan Kern (2),  (1)Alfred-Wegener-Institut
    Volker Leinweber(1) Lasse              (2)lnstitute of Oceanography IfM HH
    Rabenstein(1), Gunnar Spreen(2),
    Andreas Winderlich(2)


INTRODUCTION

The Arctic sea ice cover has been suggested to be one key indicator of the 
amplification of global warming in the high northern latitudes. One important 
goal of the Polarstern expedition ARK-XXII/2 as part of the International 
Polar Year (IPY, 03/2007-03/2009) was to assess, together with other 
scientific expeditions within the IPY, the current status (area, thickness, 
type, structure) of the Arctic sea ice cover. Our contribution to this 
international effort comprised sea ice thickness measurements using various 
techniques: drilling, ground- and airborne electromagnetic sounding, laser 
profiling, deployment of one ice mass balance buoy and eight sea ice drift 
buoys (section 3.1). It comprises further airborne multi-frequency sea ice 
radar backscatter measurements, partly as satellite sensor under-flights, 
aiming at the development of methods to obtain melt pond and thin-ice area 
fraction and the thin-ice thickness. This suite of remote sensing data is 
accompanied with in-situ investigations of sea ice properties (temperature, 
salinity, density profiles, roughness) (section 3.2) Finally, hourly / 
bihourly day-round routine sea ice cover observations from the ships' bridge 
were conducted (section 3.3).

This report is organized as follows. First, the sea ice thickness 
measurements will be presented together with buoy deployment activities. Then 
the sea ice radar backscatter measurements will be described together with a 
quantitative analysis of sea ice properties that are relevant to interpret 
the radar data. The work and results described in these two chapters are 
based on measurements that have been carried out outside the Russian 
exclusive economic zone (REEZ) in the periods 28 July - 11 August and 24 
August - 20 September. The final chapter summarizes routine sea ice 
observations that were performed during the entire cruise like observations 
from the ship's bridge, digital photography, and results from routine 
drilling activities to support the work of the sea ice biology group and the 
oceanography group.



3.1  Sea ice thickness measurements
     Lasse Rabenstein, Stefan Hendricks, Volker Leinweber
     Alfred-Wegener-Institut

Objectives

The ultimate goal of sea ice thickness measurements during ARK-XXII/2 was to 
determine the sea ice thickness distribution within the Trans Polar Drift 
(TPD). These measurements are a continuation of a series of measurements in 
the TPD which have been conducted in irregular intervals by AWI since 1991. 
Former observations show a thinning of the TPD ice of 20% within 10 years. 
The last Arctic summer sea ice campaign took place in 2004. 2005 and 2006 
were years with a minimum in sea ice extent. Also the year 2007 shows a new 
record minimum in sea ice extent. Therefore it is of high interest how the 
sea ice thickness distribution has developed since 2004 and how its actual 
status is in 2007.

The instrument of choice is a helicopter based airborne electromagnetic (AEM) 
induction sounder, the so called "EM-Bird". Furthermore two ground based EM 
devices, an EM31 and a SLINGRAM instrument, were used during 15 ice stations. 
Finally an electric driller was used for high accuracy point measurements.

The AWI sea ice physics group used AEM and EM31 instruments during previous 
campaigns. These methods are well established and a lot of validation 
measurements were done. It was found that EM methods underestimate thickness 
over deformed ice. To deepen the understanding of the induction process in 
deformed ice the SLINGRAM method was used for the first time on sea ice. The 
advantage of SLINGRAM is a variable frequency and coil separation, in 
comparison to AEM and EM31, where these parameters are fixed.


Airborne Electromagnetics (AEM) 

Introduction

Airborne measurements are a very useful way to obtain statistically robust 
probability distribution functions of sea ice thickness because of two 
advantages. First, every ice thickness can be sampled and second, the 
measurements can cover a regional scale of a few hundred kilometres. For this 
purpose, the Alfred Wegener Institute maintains a helicopter based system 
that can be operated with every helicopter, which is certified for carrying 
an external sling load. The so called EM-Bird consists of a cylindrical 
instrument and a towing cable with a length of 20 meter. The instrument 
itself is operated in an altitude of 10 to 15 meters above the sea ice. A 
dragskirt is mounted on the rear for stability during the flight with a 
typical speed of 80 knots. The system has a weight of 100 kg and a total 
length of 3.4 meter and is therefore small enough for takeoff and landing 
operations directly from the helicopter deck.


Fig. 3.1.1: EM-Bird on the helicopter deck of Polarstern. A special cart was 
            used for transport, takeoff and landing operations.


The instrument consists of a pair of rectangular coils, one for generation 
and one for the reception of low frequency electromagnetic fields. The 
transmitter coil emits a harmonic signal, the so called primary field, which 
is the source of induction processes in all conductive mediums in close 
vicinity to the coil, mainly the ocean water. The induced eddy currents are 
the source of a secondary electromagnetic field which is detected together 
with the primary field at the receiver coil of the EM-Bird. From the complex 
ratio of secondary to primary field the distance between the instrument and 
the sea water interface can be computed. The EM-Bird used during ARK-XXII/2 
has a frequency of 4.06 kHz and a coil spacing of 2.77 meters, with a 
horizontal coplanar coil configuration. Samples are taken at a rate of 10 Hz, 
yielding a point spacing of 3 to 4 meters at average flight speeds. Other 
parts of the EM-Bird are a near infrared laser altimeter system and a GPS 
receiver. All data are processed fully digitally within the EM-Bird and sent 
by a wireless data link to the helicopter where it is recorded by a standard 
laptop PC. All data can be observed by the operator in real time and the 
laser height of the system is additionally displayed to the pilot by a 
standard avionic altimeter.

Sea ice thickness can be computed by taking the difference between the 
distance to the sea water derived by the EM system and the readings of the 
laser altimeter. Because the return signal of laser always reflects the 
uppermost reflecting surface, the result of the measurements are the ice plus 
snow or total thickness if a snow layer is present on the summer sea ice in 
the Arctic.


Work at sea

Data Acquisition

Measurements were performed with a wide regional coverage of the sea ice of 
the Transpolar Drift. In total more than 4,000 km of sea ice was profiled 
during 23 measurement flights. In summer Arctic airborne operations are 
always hampered by hazards like in-flight icing of the helicopter and poor 
visibility due to fog patches. Therefore some flights had to be aborted 
during measurement operations since the weather showed continuously poor 
conditions during the cruise.


Fig. 3.1.2: Map of EM-Bird flight tracks (triangles)


In general, the flight tracks followed triangles with equally sized sides 
with a length of 40 nautical miles. On the corners and in the middle of the 
sides the helicopter ascended to an altitude of roughly 500 feet for system 
calibration of the EM-Bird and radio contact to the bridge of Polarstern. The 
ascents divided the whole flight in profiles with a length of 15 to 20 
minutes. In addition to the EM ice thickness measurements observations of the 
sea ice surface with a digital camera were conducted by a second person in 
the front of the helicopter. The images were taken roughly every 5 minutes 
and geo-located with a waypoint of a handheld GPS. For a more detailed 
analysis of the surface properties like melt pond coverage of the sea ice a 
nadir pointing video camera was also mounted on the helicopter (see section 
3.3: Sea ice observations).


Data Processing

The low temperatures of the Arctic play an important role for the performance 
of electronic components of the EM-Bird. In addition problems with static 
charges can arise during takeoff procedures, hence the bird has to be 
disconnected from the power source and cannot be heated at this time. 
Therefore some electronic components showed significant temperature driven 
drift behaviour especially in the beginning of the flight. A drift correction 
is applied by taking zero level measurements in an altitude of more than 300 
feet at the beginning and the end of one profile where no signal of the sea 
water can be received.

Furthermore sites with open water can be used to calibrate the system during 
the flight. Over open water the EM derived distance and the readings of the 
laser altimeter are identical. Therefore any larger leads were flagged by the 
operator manually in the data stream for later identification of suitable 
calibration points in the data processing.

The conductivity of the ocean water is necessary for the calculation of sea 
ice thickness. The value is taken from the keel salinometer of Polarstern. It 
is assumed that the conductivity is regionally stable in all areas of the 
measurements. Errors in the final data product may arise due to local fresh 
water concentrations caused by stronger melting in the loose ice pack. If 
possible the conductivity was checked over large leads by measuring the 
response of open water at different heights of the EM-Bird. In general a good 
agreement was found and the possible error is assumed to be in the order of 
10 cm, which has been found in the variability of the ice thickness result 
over open water sites. For data processing convenience an average 
conductivity during the flight is calculated with the Polarstern data and 
sampled into 50 mS/m steps. The conductivity of the surface sea water varied 
significantly between 2200 mS/m and 2700 mS/m during the cruise depending on 
the occurrence of melting and the geographical location.

The conductivity value is used to generate an analytical relation between EM 
readings and distance to the sea water. The relation can be approximated with 
a series of two exponential functions. The inverse of this function is then 
used to calculate the distance of the EM-Bird to the sea water. From this 
distance the laser range is subtracted to get total thickness. As a final 
step the ice thickness data is geo-located with the onboard GPS antenna.


Fig. 3.1.3: EM data taken in different altitude. Displayed distance of the EM 
            Bird to the ice surface is the in phase component (real part of 
            the complex ratio of secondary to primary field) versus the laser 
            range. The black line gives the theoretical values for open 
            water. At a given laser height sea ice causes a reduction of the 
            in phase component, while measurements over open water coincide 
            with the theoretical curve.


Preliminary Results

All flights show a most frequent ice thickness of equal or less than 1 meter 
(see Table 3.1). No clear thick multiyear ice class could be identified in 
the individual thickness distribution functions. There are only very weak 
indications for a modal thickness of 2.3 meters in the second flight on 8 
September. This finding coincides with visual observations of thick deformed 
ice which also carried a lot of sediments. For all other flights it can be 
assumed that the same ice type was surveyed during the complete cruise.

The measurements can be roughly divided into two zones: The Eurasian and the 
central Arctic. Both are divided by a time of two weeks, where Polarstern was 
within the REEZ and no EM ice thickness measurement could be carried out. 
Most of the flights were conducted in the central Arctic and only 5 surveys 
were done before entering the REEZ.

The results of the preliminary data processing shows that the ice was 
slightly thicker in the Eurasian zone (20 cm in average ice thickness) than 
in the later phase of the cruise in the central Arctic. It has to be 
mentioned that melting was still going on in the mid of September. For later 
interpretation therefore spatial and temporal effects on the ice thickness 
distribution have to be decoupled. All ice thickness statistics for both 
zones and all measurements are given in Table 3.2. The last flights of the 
cruise were dedicated to survey the ice thickness at the very northern ice 
edge. These flights included a lot of open water which is reflected by the 
much higher open water fraction.

The open water fraction is defined of the part of the thickness profile which 
is thinner than 10 cm. This value reflects the accuracy of the system over 
level ice.

The probability density function (pdf) of both zones are displayed in figure 
3.1.4. Additionally to the open water, the fraction of thin ice is also 
raised compared to the beginning of the cruise where all leads where 
completely ice free.


Tab. 3.1.1: Results of the individual EM-Bird profiles. The conductivity 
            value represents the value used for the processing.

Date        Flight  Conduc-  Length    Modal      Mean     Standard    Median
                    tivity           Thickness  Thickness  Deviation  Thickness
                    [mS/rn]   [km]      [m]        [m]        [m]        [m]
----------  ------  -------  ------  ---------  ---------  ---------  ---------
2007/08/03    #1     2650    166,6      1,00       1,32       0,78      1,16
2007/08/06    #1     2650    226,0      1,00       1,40       0,73      1,22
2007/08/06    #2     2650    133,5      0,80       1,30       0,71      1,09
2007/08/07    #1     2700    255,2      0,90       1,31       0,64      1,14
2007/08/10    #1     2650    150,1      0,80       1,40       0,88      1,19
2007/08/28    #1     2550    224,9      1,00       1,34       0,87      1,16
2007/08/28    #2     2500    219,1      1,00       1,30       0,81      1,13
2007/09/01    #1     2400    215,9      0,70       1,13       0,81      0,94
2007/09/03    #1     2350    182,7      0,80       1,22       0,73      1,02
2007/09/03    #2     2350    37,1       0,80       1,36       0,76      1,16
2007/09/04    #1     2300    221,4      0,90       1,43       0,73      1,27
2007/09/04    #2     2300    110,7      0,70       1,16       0,63      1,00
2007/09/06    #1     2300    111,4      0,90       1,34       0,68      1,18
2007/09/08    #1     2250    211,3      0.70       1,18       0,70      0,99
2007/09/08    #2     2250    218,6      1.00       1,48       0,85      1,26
2007/09/09    #1     2200    216,5      0,40       0,88       1,00      0,60
2007/09/09    #2     2200    48,3       0,80       1,24       0,87      1,02
2007/09/10    #1     2400    147,2      0,80       1,30       0,80      1,10
2007/09/15    #1     2550    149,7      0,60       1,01       0,76      0,86
2007/09/16    #1     2550    221,7      0,90       1,14       0,71      1,01
2007/09/17    #1     2550    225,4      0,50       0,81       0,78      0,64
2007/09/18    #1     2500    138,9      0,00       0,46       0,56      0,26




Both zones combined give the summer sea ice thickness distribution of the 
Transpolar Drift in 2007 (Fig. 3.1.5). The most prominent ice thickness is 90 
cm which is significantly lower than measurement in previous years. For 
example measurements in the Laptev Sea in 1994 showed a modal ice thickness 
of 2 meter in the Transpolar Drift.

Therefore it can be concluded that in general the sea ice thickness 
measurements during ARK-XXII/2 support other findings of a decreasing ice 
thickness and a shrinking of the multiyear ice zone, especially in the 
Russian Arctic.


Fig. 3.1.4: Probability Density Functions (pdf) of ice thickness measurements 
            in the Russian (grey) and in the Eurasian Arctic (line)

Fig. 3.1.5: Probability Density Function (pdf) assembled with all measurements 
            in the Transpolar Drift in late summer 2007. Marked are the local 
            maxima of the pdf, which give the fraction of open water and the 
            modal thickness of the profiled sea ice.


Tab 3.1.2: Statistics of the sea ice thickness distribution measured in 
           different zones

    Zone               Modal       Mean    Standard     Median     Open
                     Thickness  Thickness  Deviation  Thickness    Water
                        [m]        [m]        [m]        [m]     Fraction
                                                                    [%]
    ---------------  ---------  ---------  ---------  ---------  --------
    Eurasian Arctic     0,90       1,34       0,74       1,17      0,7
    Russian Arctic      0,80       1,15       0,81       1,01      5,7
    Both Zones          0,90       1,20       0,80       1,04      4,7



EM31 ground-based EM measurements 

Introduction

Measurements of sea ice thickness in the Transpolar Drift utilizing 
electromagnetic induction devices have been carried out at Arctic Polarstern 
cruises from 1993 to 2003. This time series has been continued using a 
Geonics EM31 device during all ice stations. The spatial coverage of the 12 
stations ranges from the European basin and the Russian Arctic to the Makarov 
Basin. Besides obtaining the sea ice thickness distribution surveys were 
conducted for retrieving supplement data for airborne scatterometer data 
(section 3.2) and mapping of possible buoy deployment sites (section 3.4).


Fig. 3.1.6: EM31 placed in a kayak and towed over the sea ice


The instrument operates at a frequency of 9.8 kHz and a coil spacing of 3.66 
meters. For better signal to noise ratio a horizontal coplanar coil 
configuration was chosen. The EM readings are recorded by an autonomous data 
logger at frequency of 0.5 Hz together with geographical positions by an 
external handheld GPS device. All devices are mounted to a kayak for easy 
access to every type of ice including melt ponds and even open water. To 
shelter the EM31 from the external conditions the instrument is placed inside 
the kayak, while data logger and GPS are mounted outside allowing a quick 
inspection of the data acquisition process at all time.

The selection of floes for ice stations was mostly restricted by the 
possibility to access the floe from the ship and therefore mainly by the 
parameters size, degree of ridging and thickness. In general thick multiyear 
ice floes were selected for ice stations which will result in a bias of the 
obtained total ice thickness distribution. The area of the survey on the ice 
was normally limited by floe edges and the occasional poor visibility 
yielding in an average profile length of more then 2 km and a total length of 
roughly 30 km of ground EM data.


Fig. 3.1.7: Map of ice stations (triangles) with EM31 ground based EM 
            measurements of sea ice thickness


Calibration

Calibration procedures of the EM31 instrument were carried out during several 
ice stations along the cruise. These were necessary since the instrument 
showed an offset from the analytical relation of EM readings and distance to 
the sea water - ice interface. In addition the Inphase reading was always in 
saturation. This problem could not be solved onboard and was not considered 
to be critical since the ice thickness is calculated with the apparent 
conductivity only which showed besides the offset a reasonable behaviour.

For calibration, values of the apparent conductivity are taken with a known 
distance of the EM31 to the sea water interface. This is equal to ice 
thickness for an instrument directly placed on the ice. A good location of 
level ice was investigated by ice drilling to retrieve the ice thickness on 
an area of roughly 5 m x 5 m. Then an average of thickness values of the 
different holes is used as reference distance. As a second step, the kayak 
was lifted to different heights above the sea ice. The distance between ice 
surface and bottom of the kayak was measured with a ruler tape and the value 
for the apparent conductivity was recorded. Four different heights were 
realized: 1) on the ice, 2) hip-height, 3) shoulder height and 4) height 
above the head. At some ice stations the kayak floated freely in the sea 
water to get calibration points for very small distances to the conductive 
medium. To all heights a value of 18 cm was added which was the distance 
between the centre of the coils and the bottom of the kayak.


Fig. 3.1.8: Result of EM31 calibration procedure. Marker are calibration 
            point represented by exponential relation (dotted curve). Lower 
            three curves are numerical solution for 2,200, 2,400 and
            2,600 mS/m as conductivity of sea water.


From analytical calculation it is known that the retrieved ice thickness 
depends only weakly on the conductivity of the sea ice. For all the ice 
stations it is assumed that the conductivity of the sea ice plays no role 
despite the fact that a lot of brine channels were present during the cruise. 
Therefore all calibration results of different ice station were used to 
construct one relation between apparent conductivity and ice thickness which 
was used for all ice stations (see Fig. 3.1.8). The relation has the 
following form:
                        z=-1/a2_ln((sa-a0)/a1)-ho

(where: z = ice thickness, sa = apparent conductivity, h0 = instrument offset 
of EM31 in kayak, a = calibration coefficients). The calibration coefficients 
were obtained by a fit of an exponential function to the calibration points 
taken on the ice stations and are listed in Table 3.3.


Tab. 3.1.3: Calibration coefficients obtained during several ice stations

                               a0     28.8821
                               a1   1196.07
                               a2      0.75357


The average error made by the fit compared to the actual data points can be 
estimated with roughly 10 cm, which lies well within the instrument error.


Processing

With the relation between apparent conductivity and ice thickness obtained by 
calibration procedure a conversion of the retrieved EM readings in ice 
thickness is a simple process. In general the files from the data logger are 
used as an input for a computer programme in the IDL language. The data is 
already synchronized with the GPS position by the data logger. Since pure GPS 
positions are of limited use on a drifting ice floe a correction is applied 
utilizing the GPS position and gyro heading of Polarstern. With both 
information a Polarstern Reference Frame (PRF) is established. For the 
reference frame a cartesian coordinate system in meters is used with the 
Polarstern Trimblel GPS antenna as origin a northwards pointing y-axis. The 
GPS antenna of Polarstern is used to correct for ice drift, while the change 
of the gyro heading can be taken to correct for floe rotation. Case studies 
have shown that a good agreement in PRF-positions can be achieved for 
repetitions of validation lines with a spacing of a few hours.

From the ice thickness data a probability density function (PDF) can be 
calculated. But this pdf can be biased since the kayak is never pulled at a 
constant speed over different types of ice and sometimes measurements were 
continued while the kayak remained at the same spot for a certain time. To 
get a statistical correct pdf the profile is re-sampled to constant point 
spacing. Since the ice drift correction is not perfect a few data points are 
removed which show only a short variation in position and no change in ice 
thickness. All points with invalid GPS positions are also not regarded for 
the final ice thickness distribution.


Preliminary Results

All data has been taken into account for preparing the final ice thickness 
results except data which has been taken during surveys of buoy deployment 
sites. During these surveys the profiles were focused to a few areas yielding 
no representative mapping of the floe.


Tab. 3.1.4: Statistics of the sea ice thickness distribution measured by the 
            EM31

            Station     Date        Mean    Standard    Median    Profile
                                 Thickness  Deviation  Thickness  length
                                    [m]        [m]        [m]      [km]
            -------  ----------  ---------  ---------  ---------  ------
               01    2007/08/02     2.27       0.80       2.46     1.72
               02    2007/08/05     1.85       0.80       1.60     1.49
               03    2007/08/07     1.60       0.36       1.51     1.39
               07    2007/08/24     2.07       0.93       1.73     1.79
               08    2007/08/28     1.49       0.80       1.14     4.08
               09    2007/08/31     1.69       0.77       1.43     2.46
               10    2007/09/02     2.15       0.87       2.01     2.64
               11    2007/09/05     2.42       1.39       1.95     2.54
               12    2007/09/07     2.03       0.80       1.78     2.52
               13    2007/09/10     1.83       1.30       1.60     4.06
               14    2007/09/14     1.64       1.00       1.48     2.11
               15    2007/09/16     1.60       0.89       1.32     3.17


Values for ice thickness are higher compared to airborne EM measurements 
which reflect the selective choice for the ice stations. The average ice 
thickness is also raised by the fact that no very thin ice was profiled, 
because it was not accessible for measurements in general. In the pdf (Fig. 
3.1.9) a modal thickness of 1.3 meter of the most dominant ice type can be 
observed, while there is weak indication for a multiyear ice mode of 2.3 
meters. But since there were always ice deformation features at almost every 
ice station this secondary mode of the pdf can also be explained by the 
thickness of younger ridged or possible rafted sea ice.

Taken all ice station into account the mean ice thickness amounts to 1.86 
meter with a standard deviation of 1.01 meter. The median of the distribution 
gives a slightly lower value of 1.6 meter.


Fig. 3.1.9: Probability density function (pdf) of sea ice thickness obtained 
            with ground- based EM measurements with data of all ice stations


Slingram ground-based EM measurements 

Work at Sea

The SLINGRAM instrument was tested during one ice station, two times on level 
ice and one time to profile an ice ridge. To operate the instrument three 
persons are necessary. One person carries the transmitter coil, one person 
the receiver coil and the third person writes the data to a protocol. The 
measurements were performed using a horizontal coplanar loop mode. Therefore 
it was necessary to hold both, transmitter and receiver, in a horizontal 
position, which could be achieved with a level. The whole procedure is time 
consuming. Nevertheless all measurements were done during one ice station. 
They include two ridge profiles of 50 meter length with a point spacing of 
two meters and measurements over level ice. The instrument works with the 
same principle as the EM31 or the EM-Bird. The transmitter produces a primary 
magnetic field and the oceans inductive response is measured at the receiver 
coil. Eight frequencies were used: 440, 880, 1,760, 3,520, 7,000, 14,000, 
25,000 and 56,000 Hz. Furthermore three different coil spacings have been 
used: 5 m, 10 m and 20 m.

The measurements over level ice were done to compare the results of the 
Slingram instrument with 1D model results. Therefore the level ice thickness 
at the measurement site was determined by drilling.

The overview about the Slingram Ridge profiling is given in figure 3.1.10. 
The ridge profiling was done using a coil spacing of 10 and 20 m only. Each 
profiling started with the receiver on the first point and it ended when the 
transmitter reached the last point. Therefore profiling with a separation of 
10 meters included 21 data points and the profiling with a 20 meter 
separation included 16 measurements. On every point all eight frequencies 
were measured. Both profiles were orientated in an angle of -45° to the 
orientation of the ice ridge, and with an angle of approximately 90° to each 
other. For every data point the height over sea surface was determined using 
a levelling laser. Furthermore ice thickness was measured on both profiles 
using an EM31.


Fig. 3.1.10: Overview about the two Slingram profiles over an ice ridge. On 
             each profile 26 points were measured



Preliminary Results

Figure 3.1.11 shows the results of Slingram measurements over level ice for a 
coil separation of 10 m and 20 m. The midpoint between receiver and 
transmitter was fixed. The ice thickness was known by an initial drilling 
survey. Therefore a ID forward model could be calculated to compare the model 
results with the measurements. The graph in figure 3.1.11 shows the ratio of 
primary to secondary field in percentage for all frequencies. The results for 
the 10 and 20 meter mode are in good agreement with the model. Only for the 
higher frequencies the results differ quantitatively from the model. In the 5 
m mode the results were in no agreement with the model and are not shown in 
the graph. The results over level ice show that the Slingram method can be 
used for sea ice applications.

Results of the ridge profiling are shown in figure 3.1.12. Combined 
levelling, snow thickness and EM31 measurements provided data of ice and snow 
thickness, including freeboard and draft. The solid lines show the Slingram 
results for a coil separation of 10 m in percentage of the primary field. The 
dark lines are the inphase and the bright ones the quadrature. The data are 
not yet corrected for the elevation difference between receiver and 
transmitter. The raw data show no significant correlation to the structure of 
the ridge, as it is determined with the EM31. A reason for this is most 
probably the 3D nature of the thickness problem in deformed ice and the 
missing height correction.


Fig. 3.1.11: Slingram results over level ice. Shown is the ratio of primary 
             to secondary field versus frequency. The left graph shows the 
             results for a coil separation of 10 m and the right graph for a
             separation of 20 m.

Fig. 3.1.12: The left graph shows profile 1 and the right graph profile 2. 
             The EM31 ice thickness is shown in black, including freeboard 
             and draft, and the snow thickness in grey. The solid lines are 
             the Slingram results.


Video recordings of sea ice surface parameters In addition to the EM-Bird 
measurements the ice situation was recorded during parts of the flights with 
a nadir-looking digital video camera. The camera was mounted inside a metal 
box to the landing gear of the helicopter. The box was isolated and equipped 
with a heating element to shelter the camera of the harsh condition during 
the flight. The system consisted of a Sony Digital Video Camera Recorder with 
the following technical specifications:


Tab. 3.1.5: Technical specifications of the Camcorder

                                               F1.6-2.4
            focal length                       6-72 mm
            diameter of objective              58 mm
            Opening angle horizontal           ca. 42°
            Opening angle vertical             ca. 33°
            Effective number of pixels on CCD  400000


The flights were recorded using 4:3 format to Sony Mini-DV tapes by "Single 
Play", so the maximum length of one recording was 60 min. Table 3.1.6 shows 
the recorded movies with their respective date, time, coordinates and length. 
Figure 3.1.13 shows all coincident data of ice thickness and video recording.

In total, 08:19:23 hours of video data were recorded. Assuming a flight 
height of 35 m, the area covered by one frame of the movie is around 27 x 21 
m or 570 m2.

In addition to the visual documentation of the ice situation, the videos can 
be used to estimate the coverage of the observed area by melt ponds, sea ice 
and open water. To achieve this, a Matlab algorithm has been developed during 
the cruise. This algorithm classifies the pixels of single frames of the 
movie by their brightness and identifies them after calibration as sea ice, 
melt pond or open water. In a second step, the areas identified as melt 
ponds, which are connected to open water, are classified as frozen open 
water.

The software gives reasonable results, but there are still some difficulties 
to overcome concerning the automatic brightness adaptations of the camera 
during the flight and masking the shape of the bird and the dirt on the 
protecting glass window.


Tab. 3.1.6: List of all video recording taken during EM-Bird measurement 
            flight

                    Start                                   End
---------------------------------------  -------------------------------------
Date  Time  Latitude      Longitude      Time      Latitude      Longitude      Length
----------  ------------  -------------  --------  ------------  -------------  --------
03.08.2007  
  16:00:14  82°17.7161 N  029°16.4692 E  16:12:39  82°32.6989 N  030°16.2661 E  00:12:25
  16:15:49  82°36.5872 N  030°32.5874 E  16:27:18  82°50.3459 N  031°33.2801 E  00:11:29
  16:31:16  82°49.6020 N  031°35.0619 E  16:47:03  82°32.9973 N  032°49.9767 E  00:15:47
  16:51:26  82°28.2229 N  033°10.2398 E  17:02:45  82°16.4028 N  033°58.2857 E  00:11:19
28.08.2007  
  10:39:42  87°02.9614 N  104°22.6998 E  10:53:20  86°48.6182 N  100°44.1192 E  00:13:38
  10:56:01  86°47.8792 N  100°33.7163 E  11:09:40  86°32.4456 N  097°22.4393 E  00:13:39
  11:12:35  86°33.7119 N  097°06.9774 E  11:26:06  86°50.3290 N  094°13.3823 E  00:13:31
  11:28:51  86°51.4749 N  094°01.0910 E  11:40:12  87°05.8347 N  090°52.7294 E  00:11:21
28.08.2007  
  15:17:08  87°04.9881 N  111°40.9976 E  15:31:47  87°04.0039 N  117°19.1139 E  00:14:39
  15:35:29  87°05.1920 N  117°09.4166 E  15:46:08  87°18.6246 N  114°30.7017 E  00:10:39
  15:49:45  87°19.3542 N  114°20.3927 E  16:02:21  87°35.1218 N  110°28.6849 E  00:12:36
  16:06:26  87°34.4330 N  110°23.4044 E  16:19:38  87°19.2644 N  107°11.2551 E  00:13:12
01.09.2007  
  14:11:39  88°04.1836 N  160°35.8495 E  14:13:17  88°04.2544 N  160°18.0482 E  00:01:48
03.09.2007  
  14:52:15  87°43.9869 N  161°49.0833 W  15:06:42  87°45.3999 N  153°31.4537 W  00:14:27
04.09.2007  
  09:17:05  87°10.6261 N  146°27.3298 W  09:17:36  87°11.2502 N  146°28.2624 W  00:00:31
08.09.2007   
  11:29:10  84°43.4537 N  146°44.6277 W  11:45:41  84°26.1937 N  145°02.7110 W  00:16:31 
  11:48:10  84°25.6635 N  144°59.8908 W  12:04:07  84°08.4448 N  143°30.3685 W  00:15:57 
  12:07:12  84°08.4686 N  143°35.0653 W  12:22:09  84°08.9914 N  146°44.7200 W  00:15:57
09.09.2007   
  09:01:31  85°04.1669 N  164°24.0936 W  09:16:06  85°15.4743 N  161°25.4462 W  00:14:35 
  09:19:07  85°16.1741 N  161°13.8356 W  09:32:44  85°26.6185 N  158°00.6745 W  00:13:27 
  09:35:41  85°25.9655 N  158°01.5749 W  09:53:17  85°05.5649 N  158°01.1665 W  00:17:36 
  09:56:15  85°05.3293 N  158°02.0116 W  09:59:31  85°01.6687 N  158°01.2747 W  00:03:16
09.09.2007   
  12:03:45  85°04.9239 N  164°41.1488 W  12:17:12  85°14.7524 N  167°45.5123 W  00:13:27 
  12:19:53  85°16.0452 N  167°55.9689 W  12:30:31  85°23.4635 N  170°30.8678 W  00:10:38 
  12:33:45  85°21.4176 N  170°53.9415 W  12:36:10  85°18.3962 N  170°55.5130 W  00:02:25 
  12:39:30  85°14.2562 N  170°58.2847 W  12:41:28  85°11.6770 N  171°00.1395 W  00:01:58 
  12:55:57  85°05.7821 N  168°00.2283 W  13:05:48  84°52.8497 N  168°00.1506 W  00:09:51
10.09.2007   
  15:31:15  86°43.4185 N  176°40.9585 E  15:42:03  86°55.8795 N  174°27.9449 E  00:10:48 
  15:44:43  86°56.3252 N  174°23.0767 E  15:58:59  87°12.1632 N  170°59.5621 E  00:14:16 
  16:04:33  87°13.8627 N  173°18.3394 E  16:13:21  87°10.0143 N  176°10.2976 E  00:08:48 
  16:15:27  87°06.9015 N  176°19.9155 E  16:33:48  86°41.8685 N  177°23.4377 E  00:18:21
15.09.2007        
  13:33:18  85°52.7721 N  116°06.2881 E  13:49:22  85°34.2055 N  114°20.3608 E  00:16:04 
  13:52:00  85°33.4614 N  114°16.8876 E  14:07:00  85°15.5707 N  112°49.3017 E  00:15:00 
  14:09:28  85°16.0807 N  112°45.1936 E  14:24:09  85°31.2466 N  110°09.4184 E  00:14:41 
  14:26:47  85°31.9861 N  110°01.6030 E  14:28:11  85°33.3822 N  109°45.6701 E  00:01:24
16.09.2007   
  12:16:14  84°39.7946 N  102°31.2567 E  12:30:16  84°48.1923 N  099°30.4372 E  00:14:02 
  12:32:31  84°48.3591 N  099°27.3379 E  12:47:03  84°56.2509 N  096°05.6496 E  00:14:32 
  12:50:16  84°56.2889 N  096°07.6495 E  13:04:58  85°08.4233 N  099°21.3631 E  00:14:42 
  13:07:47  85°09.0479 N  099°31.9153 E  13:18:49  85°17.3151 N  102°05.9464 E  00:11:02
17.09.2007   
  10:35:50  84°12.6845 N  109°37.6441 E  10:46:50  84°13.0596 N  112°10.9981 E  00:11:00 
  10:49:43  84°13.0531 N  112°10.3871 E  11:04:01  84°12.4891 N  115°27.6639 E  00:14:18 
  11:07:48  84°13.1707 N  115°23.7404 E  11:21:34  84°30.9297 N  113°52.1853 E  00:13:46

                                                                    Total time: 08:19:23


Fig. 3.1.13: Map of coincident EM-Bird ice thickness measurements and video
             recording of sea ice surface properties (symbols)



3.2  Sea ice radar backscatter measurements for improved melt-pond and 
     thin-ice cover analysis
     Stefan Kern, Gunnar Spreen, Andreas Winderlich
     Institute of Oceanography (IfM HH)

Background and objectives

Melt Ponds form regularly on summer Arctic sea ice. They are on average 10 m2 
in size, between a half and one meter deep, and can cover up to 50% or more 
of the sea ice area. Coverage and depth varies with the sea ice type and its 
degree of deformation, as well as the snow thickness at the beginning of the 
melt period. During freeze-up melt ponds are the first open water areas to 
re-freeze because of the comparably low surface salinity (even when melted 
through because of the strong vertical stratification of the ponds' water 
column). In a physical sense, melt ponds are simply areas of open water. This 
means, that due to the low albedo of open water (0.06) compared to that of 
sea ice (melting: 0.62; snow covered: 0.82), substantially more solar 
radiation is absorbed, further enhancing the melt process. Depending on the 
depth of the melt pond and upon the thickness and the type of the ice 
underneath, melt ponds show albedo values between 0.15 and 0.3.

Sea ice concentration retrieval algorithms based on satellite passive 
microwave observations tend to underestimate the ice concentration under the 
presence of melt ponds. In this case, no differentiation can be made between 
the open water of a melt pond and the open water of the leads or breaks 
between the ice floes. Therefore, quantitative estimation of the melt pond 
cover fraction on sea ice on a regular basis would support a more accurate 
retrieval and verification of the sea ice concentration during summer. It 
would further support numerical modelling of the sea ice decay during summer 
by providing a more accurate estimate of the summer-time sea ice albedo.

Thin ice (with a thickness below about 30 cm) develops in large areas during 
freezeup and in leads, breaks and polynyas during winter. Thin ice areas are 
sites of enhanced sea ice formation during winter. The associated brine 
rejection into the ocean can trigger oceanic deep convection and water mass 
modification. Thin ice areas are sites of enhanced winter-time 
ocean-atmosphere heat exchange as compared to thick ice. As thin ice can be 
deformed easily, continuous thin ice formation and deformation under the 
action of tidal and inertial ocean surface motion can add a substantial 
amount of ice to the total annual sea ice mass balance. Finally, frost 
flowers developing on thin ice have been suggested lately to play a 
significant role in the halogen chemistry in the lower troposphere.

The observed accelerated retreat of the summer Arctic sea ice cover leaves 
larger open water areas to re-freeze during fall freeze up. Thus, regular 
observations of the thin ice area and a quantitative estimation of its 
thickness will become increasingly important in the near future and will 
allow us to better quantify the salt input into the ocean during sea ice 
formation.

First attempts to get information about the melt pond cover fraction are 
either based on high-resolution visible frequency satellite observations 
(Landsat-TM), or data obtained from active microwave instruments at 5.3 GHz 
(C-Band) (RADARSAT-1 Synthetic Aperture Radar (SAR)) or 13.4 GHz (Ku-Band) 
(SeaWinds Quikscat Scatterometer), the latter two instruments having the 
advantage that they are independent of daylight and cloud cover. One result 
of these attempts is, that the combination of lower frequency (C-Band or 
below) with higher frequency (10 GHz (= X-Band) or above) could improve the 
quality of such observations.

Several methods exist that allow us to identify thin ice and to estimate its 
thickness using satellite data in the infrared (IR) (clear sky only) and/or 
microwave (MW) frequency range (independent of clouds). Using data acquired 
in the IR frequency range requires detailed additional information about the 
net surface heat and radiation fluxes. Such data are difficult to obtain in 
the remote polar regions and thus have often to be taken from a numerical 
model with a grid-cell size, that is much larger than the thin ice areas, 
typically resulting in an under-estimation of the flux values. Airborne 
active MW data acquired at co-polarization (see below) have been used to 
obtain information about the thickness of thin ice. Results are quite 
diverse, with maximum thin-ice thickness values to be obtained between 10 and 
100 cm, depending on the used frequency (1.0 (L-Band) to 10.0 GHz (X-Band)). 
As mentioned for the melt ponds, a combination of different frequencies (and 
polarizations) has been suggested to yield an improvement in thin-ice 
identification and retrieval of its thickness.

Aims

• to find the most appropriate frequency / polarization / incidence angle 
  combination to unambiguously identify melt ponds with active MW radiometry, 
  and to obtain an estimate of the melt-pond cover fraction

• to find the most appropriate frequency / polarization / incidence angle 
  combination to unambiguously identify thin ice with active MW radiometry, 
  and to improve existing methods to estimate the thickness of thin ice

In order to achieve these aims we combined helicopter-borne radar backscatter 
measurements of sea ice with in-situ measurements of relevant sea ice 
parameters. Measurements of the in-situ ice properties are needed to 
interpret the Multi3Scat radar backscatter signal. The backscattering of the 
radar electromagnetic waves can be divided in the main processes surface and 
volume scattering. While the surface scattering is mainly a function of the 
surface roughness and the near surface conductivity, the volume scattering is 
influenced by the porosity and conductivity of the scattering medium. 
Porosity and conductivity can be described by the ice density, salinity, and 
the amount of liquid water (relevant in case of snow), and determine also the 
penetration depth of the electromagnetic radiation into the scattering 
medium. This depth determines at which layer the main scattering processes 
take place. It is not just a function of density and conductivity of the 
scattering medium but also of the used frequency. For the frequencies used by 
the Multi3Scat sensor (see section Work at sea - The Multi3Scat sensor) the 
maximum penetration depth into sea ice should be below 50 cm. To obtain this 
information the following quantities are measured on the ice:

1) surface properties
   • snow thickness DS
   • snow surface temperature TS
   • snow-ice interface temperature TI
   • air temperature TA
   • photography of snow and ice surface in front of a scale to get e.g. the 
     snow grain size

2) volume properties (profiles of only the uppermost meter of the sea ice)
   • temperature
   • salinity
   • density

Additional ice cores were drilled and stored in one piece at -20°C. They are 
available for the preparation of thick- and thin-sections for analysis of ice 
grain, air bubble and brine pocket size and distribution at home. It is 
important to mention, that the salinities and densities measured here, are 
the pure ice values with most of the brine removed. Especially the warm 
porous summer ice sampled on this cruise looses most of its brine through the 
large, opened brine channels when the ice core is removed from the borehole. 
Only small parts of the brine in closed brine pockets remain in the ice. In 
the upper few centimetres, which are most important for our later analysis, 
above the water level this problem might not be severe, but in the lower part 
of the ice core this effect is definitely not negligible.


Work at Sea

The Multi3Scat Sensor

We used the multi-frequenmulti-polarization mono-static homodyne low-IF 
Doppler-Scatterometer (Multi3Scat) of the University of Hamburg to obtain 
measurements of the radar backscattering characteristics of sea ice. Figure 
3.2.1 a) shows the components of the Multi3Scat schematically. Table 3.2.1 
gives an overview upon its technical specifications. In this table the 
polarization combination HH means, that the Multi3Scat transmits and receives 
electromagnetic radiation at horizontal polarization; VV means the same for 
vertical polarization. These are the co-polarized channels. HV and VH are the 
so-called cross-polarized channels, meaning that the radiation is transmitted 
at either horizontal or vertical polarization and received at the not 
transmitted polarization, respectively. Thus, the Multi3Scat allows to 
measure quasi-simultaneously at five different frequencies and four different 
polarization combinations.

The radar signal is generated in five modules separately for each frequency 
band by two phased-locked oscillators with external reference frequency input 
by a Direct Digital Synthesizer (DDS), offset against each other by a 
frequency in the kHz range (about 700 Hz at L- and about 10 000 Hz at 
Ku-Band). Microwave switches in the transmit and receive path, which are 
triggered via the steering unit with another reference signal from the DDS, 
allow to switch between transmission and reception of the radar signal. For 
each frequency band the signal is transmitted via a circulator into a 
multiplexer, where the signals of all bands are combined and subsequently fed 
into the parabolic dish antenna via a dual-polarization broad-band feed after 
decomposition into H- and V-polarization. The received signal is first 
decomposed into the five frequencies, fed into the corresponding module via 
the circulator and then mixed with the signal of the receive path. The 
resulting kHz signal is subsequently amplified and filtered before it is 
digitized and streamed on a hard disc (Fig. 3.2.1).


Fig. 3.2.1: a) Schematic overview of the components of the Multi3 Scat with, 
               from left to right, the DDS, the steering unit, the MW unit, 
               and the digitizer; 
            b) Schematic overview of one frequency band.

Fig. 3.2.2: Picture of the radar antenna looking out of the helicopter 
            together with the IR camera (to the left above the antenna) and 
            the Video camera (to the right above the antenna)


Tab. 3.2.1: Technical specifications of the Multi3Scat

Scatterometer Typ            Homodyne low-IF Doppler-Scatterometer
Antenna                      Parabolic Dish, Ø 96 cm
Polarization                 VV, HH, HV, VH
Nominal Flight altitude [m]  30-300
Speed above ground [m/s]     30-50
Incidence angle * [°]        20 to 70 (can be altered during flight)
Pulse repetition [Hz]        81920 (typical)

Frequency-Bands                   L            S          C          X         Ku
                             -----------  ----------  ---------  ---------  ---------
Frequency [GHz]                  1.0          2.4        5.3       10.0       15.0
Wavelenqth [cm]                 30.0                     5.7        3.0        2.0
Emitted Power [mW]               150          100         40        10         10
Antenna Beamwidth
(2-Way; 3dB) [°]                13.6          5.6        2.5        1.4        0.9
Footprint: * = 23° [m x m]   20.7 x 22.6   8.5 x 9.2  3.8 x 4.1  2.2 x 2.3  1.4 x 1.5
           * = 53* [m x m]   31.7 x 54.0  13.0 x      5.8 x 7.9  3.3 x 5.4  2.1 x 3.5



The Multi3Scat is mounted in an aluminium frame, which is designed to be used 
with a MBB BO-105 helicopter. A lever at the end of this frame carries the 
antenna and can be steered during flight to allow incidence angles between 
20° and 65°. This lever also carries a high-resolution video camera (704 x 584 
pixels) and a medium resolution IR camera (384 x 288 pixels, 0.08 K 
temperature resolution), both looking at the radars' footprint and allowing 
to monitor the ice situation (type) and to obtain the IR (surface) 
temperature distribution within the radar's footprint along track at the same 
incidence angle as the radar (Fig. 3.2.2). A gyro measures continuously the 
helicopters' pitch- and roll-angle, which is streamed to the hard disc 
together with data of the helicopters' altimeter and the radar data. Two GPS 
receivers complete the system, one feeding the position into the data stream 
of the IR camera, while the position of the other one is acquired together 
with the radar data.


Observations

Our work at sea can be divided into two parts: the helicopter-borne 
(Multi3Scat-) measurements and the sea ice property measurements at so-called 
ice stations, which will be separately described now in detail.


In-situ Measurements

In total sea ice was sampled at 22 positions. Details can be found in Tab. 
3.2.2 and figure 3.2.3. The ice sampling work can be divided in three types. 
The main type are 13 ice stations beside Polarstern, hereafter called ice 
stations, where always several ice cores were drilled and additional data 
from other groups are available. These ice stations lasted 3 hours at 
minimum. The two ice stations in the Russian exclusive economic zone (EEZ) 
(stations number 5 and 6 in Table 3.2.2 and white circles near 
Franz-Joseph-land in Fig. 3.2.3) differ from the other stations, as here due 
to Russian regulations no sea ice remote sensing work was allowed. Therefore 
only plain salinity and temperature ice core profiles were taken to 
complement the sea ice biology data sets (Chapter Sea ice Biology). The 
second type of ice stations is done by helicopter flights on a floe. The 
sampling is the same as for the normal Polarstern ice stations but time is 
limited; therefore fewer samples could be taken. But it is advantageous that 
helicopter Multi3Scat over-flights are secured for these stations and that 
they can be performed parallel to all other ship station work without using 
up ship time. Three such helicopter stations were performed. Towards the end 
of the cruise young ice was sampled at six positions from on board Polarstern 
using the ship's mummy chair.


Tab. 3.2.2: Overview of ice work done during the cruise. Ice samples were 
            either taken at ice stations beside Polarstern, on floes visited 
            by helicopter, or by ice fishing from the mummy chair on board 
            Polarstern.

                Overview of Multi3Scat ice stations during ARK-XXII/2

                Polarstern,                                           Number of
                helicopter                Number  Number   Number of  core for
                or ice fish-  Multi3Scat  of ice  of lead  melt pond  structure
Nr.    Date     ing station   overflight  cores   samples  samples    analysis
--- ----------  ------------  ----------  ------  -------  ---------  ---------
 1  2007-08-02  Polarstern        no        8        0        0           1
 2  2007-08-07  Polarstern        no        6        0        0           0
 3  2007-08-10  helicopter        yes       2        0        0           0
 4  2007-08-11  helicopter        yes       2        0        0           0
 5  2007-08-12  Polarstern        no        3        0        0           0
 6  2007-08-15  Polarstern        no        4        0        0           0
 7  2007-08-24  Polarstern        yes       5        0        1           2
 8  2007-08-28  Polarstern        yes       3        2        0           1
 9  2007-08-31  Polarstern        no        3        0        1           0
10  2007-09-02  Polarstern        yes       3        1        4           1
11  2007-09-04  helicopter        yes       1        0        1           0
12  2007-09-05  Polarstern        yes       3        0        3           1
13  2007-09-07  Polarstern        yes       3        1        2           0
14  2007-09-10  Polarstern        yes       3        2        2           1
15  2007-09-13  Polarstern        yes       4        2        1           0
16  2007-09-16  Polarstern        yes       3        3        1           0
17  2007-09-17  six ice
 ↓       ↓       fishing          no        0        9        0           1
22  2007-09-19   stations
-------------------------------------------------------------------------------
Total           13 Polarstern,   11/16     56       20       16           8
                 3 helicopter, 
                 6 ice fishing 
                   stations


The positions of all 22 ice stations are shown in figure 3.2.3, circles 
denote ice stations beside Polarstern, triangles helicopter ice stations, and 
diamonds ice fishing from on board Polarstern. For 11 out of the 16 normal 
and helicopter ice stations a simultaneous Multi3Scat over-flight could be 
accomplished. In figure 3.2.3 crosses mark all Multi3Scat measurements (see 
Table 3.2.3 for an overview of Multi3Scat measurements).


Fig. 3.2.3: Overview of data taken at sea. Circles denote positions of ice 
            stations beside Polarstern, and triangles ice stations accessed 
            via helicopter. Crosses give the rough position of Multf3Scat
            helicopter flights. Diamonds mark the locations where ice fishing 
            of thin ice from on board Polarstern was performed. The black 
            line together with black and white arrows marks the track of 
            Polarstern. A few key dates are given for better orientation. 
            Underlain are ice concentrations from 6 September 2007 derived 
            from AMSR-E 89 GHz data.


A typical ice station consists of the following steps

  1. mark a track on the ice. 
  2. helicopter Multi3Scat over-flight above this track. 
  3. taking samples at at least three positions along the track.

The track length was typically between 300 and 600 m. The direction of the 
track was chosen due to the following requirements: a) If simultaneously or 
within a time range of a few hours the acquisition of satellite SAR 
(synthetic aperture radar) data was scheduled, the track was aligned along 
the SAR look direction. b) If new ice areas were accessible within walking 
distance of the ship, these were made part of the track. c) Positions of 
sampling stations of the other groups on the ice were avoided in order to 
keep the surface as undisturbed as possible for the Multi3Scat measurements. 
d) The track has to be easily accessible from the ship and for the helicopter 
pilots. As example the set up of the last ice station on 16 September 2007 
with the accordant Multi3Scat tracks is shown in figure 3.2.4.

The track was marked with blue plastic bags filled with snow, which were 
distributed in distances of around 100 m along the track. In most cases this 
was sufficient but in future large flags or larger bags with different colour 
should be used for better visibility. To not further disturb the surface and 
get out of the Multi3Scat measurement the people having laid out the track 
now left the ice.


Fig. 3.2.4: Set up of the ice station track on 16 September 2007. Circles 
            mark the positions of the ice cores taken. Grey lines mark the 
            helicopter flight tracks of the over-flights with different 
            angles of the MOP Scat All positions were referenced to the 
            Polarstern GPS position to correct for linear drift during
            the ice station time.


After the Multi3Scat over-flights (see section Work at sea: Multi3Scat 
Measurements) the actual ice work started. A sledge was used to transport the 
equipment to the measurement points. Figure 3.2.5 gives an overview of a 
typical ice station together with the measurement equipment used on the ice.


Fig. 3.2.5: Picture of a typical ice station with the on the ice used 
            measurement equipment


The main part at every measurement point was the drilling of at least one ice 
core of 1 m length. After drilling the ice core was put in a half pipe for 
analysis and sawing (Fig. 3.2.5). At first the ice core was photographed 
(entire core + every 20 cm segment). Then the temperature profile was 
measured in 5 cm steps with a rod thermometer (Fig. 3.2.5). This thermometer 
was also used to measure the air temperature, the snow surface temperature 
and the snow-ice interface temperature. For the salinity and density analysis 
the ice core was sawed in slices afterwards. As for the Multi3Scat 
backscatter comparison only the upper part of the ice is important, the upper 
55 cm of the ice cores were sampled in 5 cm slices. From the rest of the core 
only two to three additional 5 cm slices were taken to have some continuation 
of the profiles. The ice slices were put in plastic bags. Slices with an 
undisturbed shape and thus could be used for density analysis were marked. At 
the end of the ice station the ice core slices were transported back on board 
and in frozen state the weight of the marked slices was measured. By knowing 
the diameter (9 cm) and the length (5 cm) of the core slice the ice density 
can be calculated from the weight. At the following day after melting of the 
ice the salinity and the initial values conductivity and temperature were 
measured with a handheld salinometer "Cond 3151" (Fig. 3.2.5). The same 
salinometer was used on the ice to measure the salinity and temperature in 
the borehole, in melt ponds, and leads. Some ice cores were continued beyond 
the first 1 m all through the ice to get the ice thickness and some deeper 
samples of temperature, salinity and density. But this was not done 
systematically. With a ruler stick the snow thickness and ice freeboard was 
obtained. The ice freeboard was always measured, even when the core was not 
drilled all through the ice. The borehole always filled up with water after a 
while. If this water level then represents the sea surface level is unsure. 
Likewise the snow and ice surface as well as the snow grain size was 
photographed in front of a snow scale (Fig. 3.2.5). Additional ice core and 
water samples from melt ponds and more important leads were taken after the 
refreezing had started (i.e. basically after 24 August). If possible the lead 
and melt pond ice was also analyzed in slices like the ice cores. If slush 
was present at the thin ice surface it was sampled and investigated 
separately. To get further insight in the crystal structure of the sea ice at 
interesting measurement sights additional ice cores were drilled and stored 
at -20°C temperature on board Polarstern. They will be analyzed by thin cut 
sections after the cruise. See Table 3.3.2 for an overview what samples had 
been taken on what dates. If Multi3Scat over-flights were performed the 
position of every ice core was determined with a GPS connected to a notebook 
or with a handheld GPS. To bring the ice core positions in accordance with 
the Multi3Scat measurements sometimes taken hours before, they had to be 
referenced to the Polarstern position at that time. By this procedure linear 
drift is corrected for and all measurements can be related to each other, as 
can be seen by the good match in figure 3.2.4.


Multi3Scat Measurements

The Multi3Scat measurements can be divided into two groups:
  • measurements at ice stations (group a)
  • measurements carried out along long profiles in order to obtain a sample 
    data set representative for a larger region or to obtain a sample data 
    set as coincident as possible with a satellite sensor over-flight (group b).

Table 3.2.3 gives an overview about all measurement flights. The intention 
behind the measurements of group a) is to repeat Multi3Scat measurements at 
different incidence angles over sea ice with properties, which are 
investigated by carrying out in-situ measurements as has been described in 
section Work at sea: In-situ Measurements.

Once the profile was laid out, we carried out Multi3Scat measurements at five 
different incidence angles (20, 30, 40, 50, and 600) by flying along the 
profile from always the same direction with constant speed and at constant 
altitude. In case of a potential satellite sensor overpass the look direction 
of the Multi3Scat was chosen to be similar to that of the satellite sensor. 
The length of each flight leg was chosen such that the part over the profile 
on the floe was situated close to the middle of the leg in order to ensure 
almost identical tracks. The length of these flight legs varied depending on 
limitation by time, weather, and daylight between one and five miles.

Flights of group b) that were under-flights of satellite sensor overpasses 
were carried out such, that at least during one flight leg look direction and 
incidence angle of the Multi3Scat and the satellite sensor were similar. 
Different flight pattern were used. However, due to i) the limited chance of 
obtaining satellite images within the range of Polarstern and/or the 
helicopter, ii) the deviation from the originally planned cruise track to the 
north and east, and iii) the fact, that, e.g. Envisat ASAR images have to be 
ordered two weeks in advance, the total number of potential satellite image 
under-flights is small: 6 out of 24 (see Table 3.2.3).

Flights of group b) that were not under-flights of satellite sensor 
overpasses were carried out mostly at two different incidence angles (35 and 
50°) over approximately the same track, also at constant speed and altitude. 
The intention here was to cover a larger region and by this to obtain radar 
backscatter data resembling a larger variety of surface properties than was 
possible with the ice floe over-flights (group a). The data acquired along 
these long-distance flights will be used as test and application data sets 
for the methods to be developed in the future. Those Multi3Scat data acquired 
as satellite sensor under-flights will be used as training and validation 
data sets.

Unfortunately, we lost the L-Band due to the failure of more than one 
microwave component so that after 11 August Multi3Scat measurements were 
carried out with only four frequencies (S- to Ku-Band, see Table 3.2.1). 
Moreover, some problems in the streaming of the data to hard disc caused loss 
of data and/or a decreased signal to noise ratio in some cases (see Table 
3.2.3, x (co) and x (-)).


Table 3.2.3: Overview of all Multi3Scat flights. Given are (from top to 
             bottom) the date, the purpose indicated by an "X" (floe: 
             ice-station floe over-flight (compare Table 3.2.2), SAR: 
             satellite synthetic aperture radar under-flight, other: neither 
             floe nor SAR), the approximate total distance along which data 
             have been acquired, the number of incidence angles used, and the 
             five frequency bands. For meaning of "x", "x (co)" etc. see 
             legend in right part of lowermost table. The flight of 30 
             September was an open water flight for control purposes.


Date     03.08.  07.08.  10.08.  11.08.  24.08.  24.08.  27.08.  28.08.  02.09.  02.09.
-------  ------  ------  ------  ------  ------  ------  ------  ------  ------  ------
Floe                       X       X               X               X               X
SAR                                        X               X
other      X       X                                       X               X
Length    15      20      10       5      40      10     100(-)  10(-)   60(-)    10
[miles] 
angles     1       5       5       5       1       5      1(+)     5       2       5
L         co      co      co      co       -       -       -       -       -       -
S          x       x       x       x       x       x      x(co)  x(co)   x(co)     x
C          -       -       -       -       x       x      x(co)  x(co)   x(co)     x
X          x       x       x       x       x       x       x      x       x        x
Ku         x       x       x       x       x       x      x(-)   x(-)    x(-)     x(-)




Date     03.09.  04.09.  05.09.  07.09.  08.09.  09.09.  10.09.  12.09.  13.09.  15.09.
-------  ------  ------  ------  ------  ------  ------  ------  ------  ------  ------
Floe                       X       X                       X               X  
SAR                                                                X               X
other      X       X                       X       X
Length    120     60      10      20      10      120     10      70      10      80
[miles]
angles     2       2       5       5       1       2       5       1       5       1
L          -       -       -       -       -       -       -       -       -       -
S        x(co)     x       -       x     x(co)     x       x       x       x       x
C          -       -       -       -       x       x       x       x       x       x
X          x       x       x       x       x       x       x       x       x       x
Ku       x(-)      x       -       x     x(co)     x       x       x       x       x


Date     16.09.  17.09.  18.09.  30.09.  Legend
-------  ------  ------  ------  ------  ---------------------------------------
Floe       X                             L, S, C, X, and Ku: The frequency bands
SAR                X       X             Length (-): Data of some bands missing
other                              X     Angles (+): More than one angle used but 
Length    10      45      90      20     not along same the track
[miles]
angles     5       2      1(+)     2     At the frequency bands:
L          -       -       -       -     -: no data acquired / signal too weak / 
                                            too noisy
S          x       x       x       x     co: only co-polarized data fine
C          x       x       x       x     x: all data fine
X          x       x       x       x     x(-): first all data fine, then none fine
Ku         x       x       x       x     x(co): first all data fine, then only co-pol. 
                                                fine



Some remarks to the ice conditions

The aims of the Multi3Scat measurements were twofold, targeting both, melt 
pond and thin ice identification. Ice and weather conditions limited 
observations of open melt ponds to the first four flights (see Table 3.2.3). 
The lacking permission to carry out Multi3Scat measurements in the Russian 
EEZ and the growing distance between the ship's position and the nearest 
point outside the Russian EEZ restricted flights to the period between 11 
August and 24 August. At that time, however, most melt ponds were already 
covered almost completely with a thin ice layer. In the following period 
until we left the sea ice covered area on 19 September almost continuous 
northward advection of mild air masses inhibited the expected cooling and 
considerably limited new ice formation in breaks and leads. Therefore, 
Multi3Scat measurements targeting our thin ice work were restrained to thin 
ice with thickness values below about 10 cm.

Preliminary Results

In the following we will first summarize results from the in-situ 
measurements carried out during the ice stations listed in Table 3.2.2(except 
of ice stations 5 and 6, which are situated within the Russian EEZ and where 
only routine ice observations were carried out to support the work of the sea 
ice biologists and the oceanography group). After that we give an example of 
Multi3Scat measurements carried out during ice station 7 of August 24, 2007, 
together with the obtained in-situ measurements.


Summary of in-situ measurements

Ice Core Temperature

The temperature of the upper part of the sea ice is mainly depending on the 
surface air temperature also on short time scales. For the summer conditions 
of this cruise the temperature below 50 cm depth stays fairly constant at 
between -0.7°C and -1.5°C. Therefore the ice core temperature profiles can 
be separated into three classes:

a) Stations with relatively high air temperature. In our case these were the 
   nine stations with number 1, 2, 3, 4, 5, 6, 7, 9, and 15 given in Table 
   3.2.3. These are mainly stations from the beginning of the cruise but also 
   the second last station number 15 at 2007-09-13 had with -0.6°C a 
   relatively high air temperature. The air temperature of these stations 
   varied between -1.1°C and 1.2°C. The mean temperature profile of the 
   upper first meter of all ice cores from these stations is shown in figure 
   3.2.6a. Until 30 cm depth the temperature stays fairly constant at -0.2° 
   C. Below the temperature is steadily decreasing until it reaches a minimum 
   of -1.0°C at 92.5 cm depth. With 9 out of 16 stations this profile is 
   typical for the majority of ice cores taken during this cruise. This is 
   also reflected in the number of available measurements for each ice depth 
   shown in the right graph of figure 3.2.6a. In total 56 ice cores were 
   taken. Until a depth of 90 cm the number of measurements stays around 35. 
   As not all ice cores could be recovered completely and thus had a shorter 
   length than 1 m the number of measurements drops thereafter.

b) Stations with relatively low air temperature. These were six stations with 
   numbers 8, 10, 11, 13, 14, and 16 (see Table 3.2.2) towards the end of the 
   cruise. The air temperature here was between -7.2°C and -3.0°C. The mean 
   temperature profile of the cores of these stations is shown in figure 
   3.2.6b. Starting with a cold surface layer the temperatures stay below 
   -1.5°C until 30 cm depth. Below this depth the temperatures stay at about 
   -1.2°C. The variability of the temperature values is much larger here than 
   for the "high" air temperature case in figure 3.2.6a, as can be seen from 
   the error bars, which denote plus-minus one standard deviation. This may 
   also result from the smaller amount (about 15) of available measurements, 
   shown in the right part of figure 3.2.6b.

c) A profile with depth homogeneous temperature. On this cruise this case 
   could only be observed for the ice station on 5 September 2007. The air 
   temperature was -1.5°C and the temperature profile of the three ice cores 
   stays up to 1 m depth with some variations at around -1°C. Figure 3.2.6c 
   shows the mean temperature profile of this station.


Fig. 3.2.6: Mean temperature profiles of the first meter from ice cores of 
            ice stations with a) relatively high (-1.2 to 1.2°C) air 
            temperature, b) relatively low (-7.2 to 30°C) air temperatures, 
            and c) for the ice station at 200 7-09-05 with -1.5°C air 
            temperature. The right graph of each plot shows the number of 
            used measurements for every depth. Error bars denote the 
            plus-minus one standard deviation interval.


Ice Core Salinity

For the ice core salinity profiles no different classes could be identified 
clearly. Figure 3.2.7 shows the mean salinity profiles of all 56 ice cores 
taken. The most distinct feature is the raise of salinity between 25 cm and 
50 cm from about 0.2 psu to 2.0 psu. Most individual ice cores show a sharp 
rise of salinity in this depth range but the increase for an individual ice 
core in general is much steeper than expressed in the mean salinity profile 
in figure 3.2.7. This is also reflected by the large standard deviations in 
the 25 to 50 cm range (error bars in figure 3.2.7). Below 55 cm depth the 
standard deviations stay high and the profile looks not as smooth anymore. 
This is due to the reduced number of available measurements (right graph in 
figure 3.2.7) for this part of the ice cores. Only the first 55 cm of all ice 
cores were regularly sampled for salinity and density. In the lower part of 
each core only two to three slices were taken at different depths for every 
core and thus not every core incorporates in the mean value in the lower 
part. The number of measurements plot at the right side of figure 3.2.7 shows 
a sharp drop from about 55 measurements for the first 55 cm to 14 
measurements for the rest of the first meter. For the depths 57.5 cm and 62.5 
cm only 0 and 2 measurements were available, respectively. They are therefore 
excluded for the mean value calculation.


Fig. 3.2.7: Mean salinity profile of the upper meter of all ice cores from 
            all stations. The right part of the graph gives the number of 
            measurements for every depth.


Ice Core Density

The ice density only can be obtained for sawed ice core slices which 
cylindrical shape is not disturbed. As we only measure the weight of each 
core slice the volume has to be known to calculate the density. Unfortunately 
this is only given for a small part of the collected ice core slices. Very 
often the ice was broken or so porous that no clear volume was defined. 
Figure 3.2.8 shows the mean density profile of all obtained density samples. 
Even for the first, always completely sampled 55 cm of the ice cores the 
number of measurement graph at the right side of figure 3.2.8 only gives 
values of about 19 measurements out of the 56 available cores. That the 
lower, sparely sampled part of the cores is not even worse represented, is 
based on the free position choice for these samples, where undisturbed core 
parts were favoured. Nevertheless, the density profile shows the expected 
shape with lower densities in the upper part of the cores and homogeneous 
densities below 30 cm. The mean density of the first meter of all cores is 
833 kg /M3, the mean density for the homogeneous part 25 to 100 cm is 873 kg 
/M3.


Fig. 3.2.8: Mean density profile of the upper one meter of all ice cores from 
            all stations. The right part of the graph gives the used number 
            of measurements for every depth.


Other Samples

During ice stations and ice fishing stations 20 samples of thin ice were 
taken with ice thicknesses between 2 and 12 cm. The mean bulk ice salinity of 
these samples without an eventually present slush layer is 9.5±3.5 psu. If 
there was slush on top of the ice its salinity was measured separately. The 
mean slush salinity amounts to 20±7 psu. There was only one case (ice fishing 
on 18 September 2007) where the thin ice surface was dry and the salinity of 
the small snow patches on top was only 0.2 psu. Without this case the mean 
slush salinity rises to 21±5 psu. During the ice station on 16 September 2007 
frost flowers were found on the thin ice. They had a salinity of 28 psu. The 
air temperature was with -4.3°C still quite high for frost flower growth.

The 16 melt pond samples can be divided into samples from fresh melt ponds 
and those melted through the ice. All melt ponds with a water salinity below 
15 psu are declared fresh, even if some of these might have had some 
connection to sea water. The often strong vertical stratification of the 
water of the melt ponds keeps the salinity close to the surface low. The mean 
water salinity of the fresh melt ponds comes then up to 8±4 psu. The mean 
fresh melt pond ice salinity amounts to 1.2±0.9 and if slush salinity could 
be measured it has a mean salinity of 7±4 psu. The same values for the melted 
through melt ponds amount to 28±4 psu for the pond water, 6±3 psu for the 
ice, and 17±6 psu for the slush.

So in summary we have gathered a quite complete data set to better interpret 
the Multi3Scat data acquired of the ice station floes. Not shown here is the 
archive of digital photographs of each ice core, taken when it was lying in 
the half pipe which has a scale on one side. There is also an archive of 
digital photographs of the snow grain size (if possible close to the snow 
surface but also at the snow-ice interface), the snow surface roughness, the 
surface roughness of the new ice in leads or melt ponds (often determined by 
the slush layer), and finally of melt pond freeboard heights, i.e. the height 
distance between the new thin ice on the melt pond and the surrounding thick 
ice.


The Ice Station of 24 August 2007 within an acquired Envisat ASAR image

In the following data from 24 August 2007 are shown and discussed. Figure 
3.2.9 shows the location of the Multi3Scat flights carried out on that day. 
That day was unique because this has in fact been the only day (as we know so 
far at least), when an ice station took place within an acquired Envisat ASAR 
image. So first a profile was marked on the ice floe as described in section 
Work at sea: In-situ Measurements and as is shown in figure 3.2.9 b); the 
profile was marked parallel to the look direction of Envisat ASAR, which flew 
over the floe at 11:54 UTC on that day.


Fig. 3.2.9: a) Flight tracks of the Multi3Scat on August 24, 2007, overlaid 
               upon the sea ice concentration obtained from 89 GHz AMSR-E 
               data. The short legs close to the star, which marks the 
               position of the Polarstern, are floe over-flights. Solid 
               arrows denote the direction of the Multi3Scat flights. The 
               black box indicates the approximate location of the map given 
               in image b). 
            b) Map showing the location of the profile and the places where 
               three of the five ice cores were drilled (circles) relative to 
               the Polarstern together with the flight tracks of the 
               Multi3Scat (grey lines from upper right to lower left). 
               Photographs in c) and d) show the scenery at core 1 and 3, 
               respectively.

Fig. 3.2.10: Left: Relative radar cross section (RCS) as obtained with the 
                   Multi3Scat over the ice station floe (see Fig. 3.2.9) from 
                   an altitude of 200 feet at a speed of 80 knots with an 
                   incidence angle of 20° at (from top to bottom) 5-, C-, X-, 
                   and Ku-Band on Aug. 24, 2007. 
             Right: RCS ratios calculated from the RCS values shown left. The 
                   lead where the Polarstern was anchored at the ice floe is 
                   marked by two solid vertical lines and the normal arrow. 
                   The ice station floe follows to the right Locations where 
                   ice cores number 1 and 5 were taken are marked by the two 
                   dashed vertical lines and the numbers. The bold arrow 
                   indicates the profile marked on the ice floe.

Fig. 3.2.11: As figure 3.2.10 but for an incidence angle of 40°


After the profile was marked we carried out Multi3Scat flights. First we flew 
two 20 miles long profiles with an incidence angle of 40°, 200 feet altitude, 
80 knots speed, starting at around 11:15 UTC; during the first leg we looked 
against, during the second leg in the look direction of the Envisat ASAR 
(Fig. 3.2.9 a). After completion of these flights we carried out Multi3Scatt 
measurements over the ice floe along the profile as described in section Work 
at sea: Multi3Scat Measurements using five different incidence angles, 
approaching the floe from the top right, i.e. northeast (see Fig. 3.2.9 a), 
lines close to the star, and b)). Almost exactly during the Envisat ASAR 
over-flight the Multi3Scat measurement at 40° incidence angle was carried 
out. Unfortunately, the ice floe started to rotate due to the ice drift so 
that at the time of our Multi3Scat over-flights the orientation of the 
profile on the floe was not parallel to the look direction of Envisat ASAR 
anymore but off that direction by about 30° (Fig. 3.2.9 a).

The figures 3.2.10 and 3.2.11 show the results of the Multi3Scat measurements 
carried out along the profile at incidence angles of 20 and 40°, 
respectively. Given are the Radar Cross Section (RCS) derived from the 
Multi3Scat measurements (left panels) and the RCS ratios as calculated from 
these RCS values (right panels), for (from top to bottom) S-, C-, X-, and 
Ku-Band and all available polarization combinations for the first 64 seconds 
of the flight. These RCS values are obtained by calculating the spectral 
power density for each frequency band and each polarization combination. In 
the resulting spectra the backscattered signal can be located as a Doppler 
peak shifted towards lower frequencies relative to the IF (intermediate 
frequency) peak. The magnitude of this Doppler peak is a measure for the 
amount of the backscattered energy. By integrating over the topmost 6 dB of 
this Doppler peak the RCS value is calculated with an along-track spatial 
resolution of approximately 10 meters. Note that these RCS values (all RCS 
values shown in this report) are relative ones, i.e. they have not been 
calibrated so far. First flights over calibration targets have been carried 
out already before the cruise, however, allowing a proper calibration of the 
RCS values at home.

What can we learn from figures 3.2.10 and 3.2.11? First of all, all shown 
frequency bands show a reasonable separation of the co- (HH- and 
VV-polarization) from the cross- (HV- and VH polarization) polarized data, 
except perhaps Ku-Band. All frequency bands pick up nicely the change from 
thick melt-pond covered sea ice to thin sea ice (mainly dark nilas) in the 
lead where the Polarstern was located, and back to the thick melt-pond 
covered ice of the ice station floe by showing a pronounced decrease of the 
RCS values at the beginning of the lead and an increase of the RCS values at 
the end of the lead. The change in the RCS values amounts between 15 and 25 
dB. The solid black vertical lines given in figures 3.2.10 and 3.2.11 mark 
the mentioned transition between thin and thick sea ice. This change is more 
abrupt at an incidence angle of 20° than at 40° (compare Figs. 3.2.10 and 
3.2.11), because the footprint is smaller at 20° than at 40° incidence angle, 
shadowing effects by floe margins and ridges are smaller at a steeper 
incidence angle, and the radar backscatter of thin ice (as was present on the 
lead) is more influenced by the underlying water at steep than at shallow 
incidence angles. Within the lead the RCS values are varying quite a lot, 
particularly at the beginning (left) and more pronounced at 40° than at 20° 
incidence angle. This is caused by varying ice conditions as will be shown 
later.

The RCS values over the thick ice tend to be quite smooth at S-Band, while 
they are extremely variable with a number of pronounced dips at X- and 
Ku-Band; this applies to both incidence angles shown. Since the ice station 
floe exhibited rather homogeneous surface properties except that it was 
covered with melt ponds and since these melt ponds were covered with thin 
level ice, it seems likely that the observed dips in the RCS values are 
caused by the melt ponds.

It is known, that the VV/HH RCS-ratio at lower frequencies, say C-Band and 
below (see Table 3.2.1), takes values well above 0 dB over open water and 
thin ice. In particular, this RCS-ratio tends to decrease from a high open 
water value (L-Band: 10 dB) to values close to 0 dB with increasing ice 
thickness - up to about 30 - 50 cm. This effect increases with increasing 
incidence angles. Therefore we would expect that the VV/HH RCS-ratio is 
larger than zero decibel over any open water or thin ice area encountered 
during Multi3Scat flights carried out with shallow incidence angles. In fact, 
the VV/HH RCS-ratio obtained at an incidence angle of 40° takes values of up 
to 8 dB and 6 dB at S-band and C-Band, respectively, over the lead. At X- and 
KuBand there is also an indication for elevated VV/HH RCS-ratios, however 
less pronounced than at C- and S-Band. Corresponding RCS-ratio values remain 
close to 0 dB at an incidence angle of 20° (Fig. 3.2.10). At shallower 
incidence angles, i.e. 50° and 60°, we observed VV/HH RCS-values of up to 12 
dB and 16 dB, respectively, at S-Band over the lead at this ice station (not 
shown). From figures 3.2.10 and 3.2.11 there seems to be no indication, 
however, that elevated VV/HH RCS -values can be observed over melt ponds as 
well.

What else can we take from these two figures? At 20° incidence angle the 
VV/VH and HH/HV RCS ratios change to lower values over the lead than over the 
thick meltpond covered sea ice, with the largest (smallest) change at S-Band 
(Ku-Band). In contrast, these ratios tend to increase, also over the lead, at 
40° incidence angle -but only at 5- and C-Band. This might deserve future 
investigations.


Fig. 3.2.12: Zoom of seconds 28 to 48 of RCS (left) and RCS-ratio (right) of 
             figure 3.2.11 for (from top to bottom) S-, C-, and X-Band 
             together with the mean infrared (IR) temperature and its 
             standard deviation (thick and thin solid lines in lowermost 
             panels) as obtained simultaneously with the IR- camera. Letters 
             A to G refer to surface types shown in figure 3.2.13. Grey areas 
             mark the approximate position of the lead. Double arrows mark 
             examples where RCS dips coincide with elevated IR temperature 
             values; the dashed lines above these arrows intend to show this 
             coincidence at the other frequency bands as well. The location 
             of core stations one and five are marked by the normal arrows.

Fig. 3.2.13: a) Different ice types encountered during the lead over-flight 
             (compare Fig. 3.2.12); A to G show: thick ice with melt ponds, 
             light nilas, thick ice, dark nilas, bra shed thin ice, dark 
             nilas, thick ice with melt ponds, respectively. b) The ice floe 
             as it looked around core stations one and five viewed by the 
             helicopter video camera. White arrows indicate the flight 
             direction. Black circles mark the location of the plastic bags.


Figure 3.2.12 shows a 20 s (approximately 800 m) long blow-up of figure 
3.2.11 together with our coincident helicopter-borne IR temperature 
measurements. These are basically a measure of the surface temperature. Data 
from Ku-Band have been omitted here. An elliptical region resembling the 
footprint size of the Multi3Scat at C Band was used to calculate the mean 
IR-temperature and its standard deviation inside this region. Note that 
consecutive ellipses overlap so that abrupt temperature changes are smoothed 
a bit. Note also, that these IR-temperature values have not yet been 
calibrated so that any interpretation should be based on the relative I
temperature changes rather than the absolute values. The air-temperature 
during that day hovered around -1°C and did not change too much the days 
before. Therefore, differences in the IR temperature are expectedly small; 
the maximum difference amounts to 2 K.

Our in-situ measurements revealed an ice surface temperature of the thick ice 
of about -1.0°C (see also Fig. 3.2.14), while the melt pond surface 
temperature was around 0.0°C. This would make a difference of 1 K which is 
smaller than the observed one. However, since there might be small 
differences in the infrared emissivity between the coarse grained rotten 
surface of the thick ice and the relatively smooth surface of the new ice we 
cannot simply use one emissivity value to directly translate the 
IR-temperature measurements into surface temperatures. A difference in this 
emissivity value (typically around 0.98) of 0.01 can already cause a change 
in the calculated surface temperature of almost 3 K. So we cannot interpret 
the observed IR-temperatures without knowing exactly the typical infrared 
emissivity of the involved surface types.

Nevertheless, two different regimes of surface temperatures can be 
discriminated: IR temperature values around 269 K and IR temperatures at or 
above 270 K. The former regime can be associated with the thick sea ice as is 
bordering the lead (see Fig. 3.2.13 a). Temperatures of the latter regime 
occurred in two ways, either as a quite homogeneous distribution as observed 
over the lead ice, or as distinct peaks as observed over the ice station 
floe. It is likely that these peaks show the melt ponds, which can be 
expected to be a bit warmer than the surrounding thick ice. So melt ponds and 
the surrounding thick ice can be clearly discriminated from each other. 
Double arrows in figure 3.2.13 indicate where melt ponds were located along 
the Multi3Scat measurement profile, which coincide with dips in the observed 
RCS values at X-Band, all polarization combinations. There is also clear 
evidence that also at CBand, and to a smaller degree even at S-Band, melt 
ponds, which are covered with a thin ice layer, cause a decrease of the RCS 
values relative to those observed over the surrounding sea ice. The fact that 
these dips seem to be quite smoothed at 5Band can be explained by the 
footprint size at this frequency (see Table 3.2.1), which is larger than most 
of the melt ponds encountered along the profile.

At this station ice cores were drilled at five locations, at two of which 
additional ice cores for structure analysis were taken. Figure 3.2.9 C) and 
d) showed examples of the ice surface at core stations one and three. A quite 
high coverage of melt ponds (30 - 40%) can be identified. Figure 3.2.14 
illustrates the vertical temperature (a) and salinity (b) profiles obtained 
for the upper meter of the five ice cores.


Fig. 3.2.14: a) Vertical temperature profiles as measured in-situ in the ice 
             cores drilled at the ice station on August 24, 2007. b) Vertical 
             salinity profiles as obtained from these cores as has been 
             described in section Work at sea: in-situ Measurements.


The temperature profiles show an almost iso-thermal upper layer of about 40 
cm thickness with temperatures around -0.3°C, and then a gradual decrease to
temperatures around -0.9°C at a depth of 90 cm. Two of the five cores 
exhibited one to two values at the melting point. The salinity profiles 
indicate a fresh upper layer of about 30 - 40 cm thickness with a bulk 
salinity below 0.5 psu, and then an increase to salinity values between 2 and 
3 psu at a depth of about 100 cm - as can be called typically for multiyear 
sea ice. The density profiles (not shown) revealed a less dense upper layer 
(750 kg/m3) and an increase in density (on average) at approximately 40 cm 
depth, i.e. that depth where the temperature and salinity profiles start or 
have their largest changes, to values around 925 kg/m3. Digital photography 
of the ice cores (not shown) reveals a quite coarse grained upper layer, 
which is followed by relatively clear ice containing quite a number of air 
bubbles.

In summary, this ice floe seems to reveal properties that are typical for 
multi-year ice, which should be resembled in accordingly high values of the 
radar backscatter intensity. This will be checked after calibration of the 
obtained RCS values (Figs. 3.2.10 to 3.2.12) and remains a future task to do, 
as is the investigation of the structure of the two ice cores drilled in 
addition for thick and thin section analysis (see Table 3.2.2)



3.3  Routine sea ice observations

     Stefan Hendricks(1), Stefan Kern (2),  (1) Alfred-Wegener-Institut
     Rainer Kiko(3), Takashi Kikuchi(6),    (2) Institute of Oceanography lfM HH
     Maike Kramer(3), Volker Leinweber(1),  (3) lnstitut für Polaräkologie
     Sebastian Mechler(4), Sabine           (4) Optimare
     Mertineit(1), Lasse Rabenstein(1)      (5) VNIIO All-Russia Research Institute for
     Alice Schneider (3), Peter Semenov(5),     Geology and Mineral Resources of the
     Stefan Siebert(3), Gunnar Spreen(2),       World Ocean
     Andreas Winderlich(2)                  (6) Japan Agency for Marine Earth Science
                                                and Technology


Visual Observations from the Bridge

Introduction

To obtain a continuous record of the sea ice state, standardized visual 
observations were performed from the ship's bridge every hour during day 
time. During night observations were conducted more sparely depending on the 
observers' working shifts.

Visual observations are the only possibility to regularly obtain a broad 
range of sea ice parameters extended with photographs and comments about 
distinctive features. Starting with explorers and whalers visual observations 
from ships represent the longest time series about the Arctic sea ice state. 
Today still all ice going research vessels are asked to carry out such 
standardized observations to enlarge the existing database. Besides being a 
valuable dataset on its own, these observations can be used as in-situ 
comparison to otherwise obtained measurements like remote sensing data. 
Nevertheless, it has to be taken into account that all observations strongly 
depend on the personal view and estimation of each individual observer and 
such the absolute accuracy is low. Additionally Polarstern favours open water 
and thin ice areas for steaming, which may bias ice thickness observations.


Work at Sea

The bridge observations were executed following the ASPeCt (Antarctic Sea Ice 
Processes and Climate) observation protocol conceived for Antarctic sea ice, 
as to our knowledge no special standardized protocol exists for the Arctic. 
All observations were directly entered in a computer. The appropriate 
programme and a tutorial to train the sea ice observers were used from Worby, 
A. P., 1999: Observing Antarctic sea ice: A practical guide for conducting 
sea ice observations from vessels operating in the Antarctic pack ice. A 
CD-ROM was produced for the Antarctic Sea Ice Processes and Climate (ASPeCt) 
programme of the Scientific Committee for Antarctic Research (SCAR) Global 
Change and the Antarctic (GLOCHANT) programme, Hobart, Australia. The 
following parameters are recorded in the protocol: date, time, latitude, 
longitude, total ice concentration in tenth part, open water classification 
(e.g. narrow breaks 50 - 200 m), melt pond coverage, comments. Up to three 
different ice types are described in more detail starting with the primary, 
the most thickest, type. For every particular class the following parameters 
are recorded: ice concentration, ice type e.g. first year 0.7 - 1.2 m, ice 
thickness estimated from tilted floes with a ruler stick, which was attached 
to the ship's starboard side, floe size class e.g. medium floes 100 - 500 m, 
topography class e.g. ridges, new snow covered, ridge area and height, snow 
type e.g. old melting snow, snow thickness. At last also meteorological data 
is recorded: air temperature, sea temperature, wind speed, wind direction, 
cloud octas, visibility and weather code. Additionally, photos showing the 
views to port side, ahead and starboard side were taken to have a visual 
impression of the ice situation afterwards. An example of 2007-08-03, 06:00 
UTC is shown in figure 3.3.1.


Fig. 3.3.1: Pictures taken during routine sea ice observation at 2007-08-03, 
            06:00 UTC. Left: view to port side, middle: view ahead, right: 
            view to starboard side



Preliminary Results

The ship entered ice covered waters northeast of Svalbard on 1 August 2007. 
Even though the ice thickness was mainly below 2 m, the ship could only sail 
slowly and had to do a lot of ice ramming for the first ten days due to 
convergent ice drift and thus a very closed ice cover. During the rest of the 
cruise very favourable ice conditions for ship operation were encountered. 
Open leads allowed fast steaming and only few ice ramming was needed. 
Polarstern left the ice cover on 19 September 2007 in the Laptev Sea. Later 
another band of sea ice was encountered in front of Vilkitzky Strait. On 25 
September the last sea ice was observed.

In total 525 observations on 54 days were carried out, resulting in 9.7 
observations per day. Taking into account that for many hours several ship 
stations took place during these days, this represents good observation 
coverage in time. During long stations with unchanging sea ice conditions, 
less observations were needed.

The mean ice concentration during the cruise amounts to 81%. A plot of the 
observed ice concentration during the cruise is shown in figure 3.3.2. The 
main observed ice concentrations are 8 to 10/10. The high number of 10/10 ice 
concentration observations starting in September is due to the onset of 
refreezing of the open water areas. After the main part of Arctic sea ice was 
left on 19 September ice concentrations got very scattered until the last 
observation on 25 September 2007.


Fig. 3.3.2: Total sea ice concentration observed in tenth part from the 
            bridge of Polarstern between 2007-08-01 and 2007-09-25. The 
            dashed line denotes the mean sea ice concentration of 81%.


The average level ice thickness of the ice area excluding open water is 95 
cm, which matches well with the average ice measured by the helicopter-borne 
EM-bird (see Section 3.1). Figure 3.3.3 shows the observed ice thickness of 
the primary and thus thickest ice class during the cruise. For the primary 
ice class the average ice thickness amounts to 121±46 cm. During the first 
part of the cruise the thickest ice was observed with a significant part of 
observations around 2 m. Afterwards the majority of observations lay between 
75 cm and 150 cm.


Fig. 3.3.3: Sea ice thickness of the primary ice class observed from the 
            bridge of Polarstern between 2007-08-01 and 2007-09-25. The 
            dashed line denotes the mean primary sea ice thickness of 121 cm.


5% of the sea ice area consisted of ridges with an average ridged ice 
thickness of 116 cm. The mean snow depth on sea ice for this cruise is 7 cm, 
only regarding the snow covered ice the snow depth increases to 9 cm. 74% of 
the ice was snow covered and 29% of the ice was covered with melt ponds. 
Figure 3.3.4 shows the melt pond coverage in tenth part during the cruise. 
The average value of 3/10 was also the most common coverage observed. The 
majority of observations lay between 2 and 4/10 melt pond coverage.



3.3.1  Sea ice thickness around biological stations determined by drilling

       Lasse Rabenstein, Stefan Hendricks, Volker Leinweber 
       Alfred-Wegener-Institut


Background

At some positions, assistance was given to investigations of the sea ice 
ecosystem by determining ice thickness distribution.


Work at Sea

To determine the thickness distribution of larger areas around sea ice 
biology stations we flew along randomly chosen straight profiles within 
approximately 50 x 50 nautical miles. Along these lines we drilled single 
holes with a spacing of at least the correlation length of the thickness 
distribution. Since we did not know the correlation length of the area of 
interest we used a correlation length of a sea ice thickness distribution 
from the Beaufort Sea in 1980 which was determined by Rothrock et al 
(Rothrock, 1986). Therefore the separation of the drill holes was chosen to 
be two nautical miles to guarantee the statistical independence of the 
samples. In order to minimize a bias in the statistic by the choice of a 
proper helicopter landing site we always drilled 50 m on the right of the 
helicopter. In total 75 holes were drilled on four different profiles (see 
Fig. 3.3.4). Based on the Rothrock article this leads to a standard deviation 
of mean ice thickness of approximately 0.3 m. This is only valid as long as 
the thickness distribution is isotropic and stationary.

During two ice stations drilling profiles with 95 and 50 holes respectively 
were done with a point spacing of 5 m. On each point ice and snow thickness, 
freeboard and melt pond depth was measured.


Fig. 3.3.4: Drill hole sites along the four helicopter flights. The solid 
            line is the cruise track of Polarstern.


Preliminary Results

Figure 3.3.5 shows the results of the ice thickness drilling along the four 
helicopter profiles. The maximum of the Probability Density Function (PD F) 
can be found at 1.2 meters. As mentioned above the error is approximately +1- 
0.3 m. Within the error margins this agrees to the results obtained with the 
EM Bird in other parts of the Trans Polar Drift.

Figure 3.3.6 shows the result of the drilling along a profile of 250 m 
length. The ice thickness, including freeboard and draft, the melt pond depth 
and the snow thickness is displayed. The thinnest ice can be found underneath 
the melt ponds. The thinning can be explained mostly by surface melting since 
no draft anomalies under the melt ponds can be observed.


Fig. 3.3.5: Probability Density Function of the sea ice thickness along the 
            four helicopter based drilling profiles

Fig. 3.3.6: A profile consisting of 49 drilholes with measurements of snow 
            thickness, sea ice draft, freeboard and melt pond depth


Reference

Rothrock, 1986, Ice Thickness Distribution - Measurement and Theory, in 
    Geophysics of Sea Ice, Ed. N. Untersteiner, p551-576, Plenum Press.



3.4  Buoy deployments

Background

As a contribution to the International Buoy Program (IABP) eight 
meteorological buoys and one Ice Mass Balance buoy (IMB) were deployed on 
several sites in the TPD. At the same time more buoys were deployed by other 
science vessels all over the Arctic Ocean.

Work at Sea

In total nine drifting buoys were deployed. Four of them were especially 
designed for meteorological measurements on sea ice. They include a 
cylindrical shaped main part, with the computer and batteries inside and a 
pole with an ARGOS antenna, a GPS receiver and a temperature and barometric 
pressure sensor on its top. The high position of the sensors at the top of 
the mast guarantees reasonable measurements even after heavy snow fall. The 
buoy is fastened to the sea ice either inside a 50 cm deep drill hole or by 
three ice screws. Three of the four buoys are "Ice Beacons" constructed by 
METOCEAN. The fourth meteorological buoy is an ICEB-CANI6-103, also 
constructed by METOCEAN.

Four buoys are Surface Velocity Profiler (SVP). They consist of a GPS 
receiver, a temperature sensor, a barometric pressure sensor and an ARGOS or 
an IRIDIUM antenna respectively. All sensors are mounted directly to the main 
plastic body. SVP's are designed for usage in open water but can also be used 
for deployment on sea ice. If the buoys are covered by snow, the 
meteorological data may be wrong. To avoid these we deployed the SVPs at the 
top of consolidated ice ridges.

One buoy is an Ice Mass Balance Buoy (IMB). It measures air temperature, 
barometric pressure, GPS position, ice thickness, snow thickness and ice 
temperature. The IMB was part of the "super buoy station", which includes in 
addition to the IMB some oceanography buoys and a webcam (see physical 
oceanography chapter). The IMB consists of the following components (see Fig. 
3.4.1):

1) A main part similar to those of the four meteorological buoys.
2) A thermistor string of 3.5 meters in length installed in a hole with 5 cm 
   in diameter. It measures a vertical temperature profile with thermistors 
   every 10 cm.
3) A vertical pole with an above ice and under ice acoustical sensor to 
   measure ice and snow thickness. For installation a hole with 10 cm in 
   diameter was drilled.

Ideally an IMB is deployed on 2 meters thick level ice. Since no 2 meters 
thick level ice could be found, the IMB was deployed in a broad zone of 
approximately 2.5 meters thick heavily weathered but deformed multi year ice. 
The deployment sites of buoys are shown in Fig. 3.4.2.


Fig. 3.4.1: Buoy types deployed during ARK-XXII/2. Left: Ice Mass Balance 
            Buoy (IMB) Middle: Meteorological buoy (Ice Beacon) Right: 
            Surface velocity profiler (S VP)

Fig. 3.4.2: Map of all buoy deployment sites. Circles mark ice beacons, 
            triangles the surface velocity pro filers and the diamond the ice 
            mass balance buoy


Preliminary Results

All buoy data are sent in regular intervals to the database of the 
international Arctic buoy programme (IABP). This information is freely 
distributed to the scientific community. Download information can be found on 
the IABP homepage.

Acknowledgements

We thank the crew of the Polarstern, the team of FIELAX, and the team of 
HeliService International for their excellent collaboration. We thank Ursel 
Schauer for her kind way of guiding us through thick and thin ice. Finally, 
all of you who helped us physically, mentally, or with some extra equipment 
on the ice, in the helicopter, and on the vessel: Thanks a million!



3.5  Sea ice biology

     Rainer Kiko(1), Maike Kramer (1),  (1) Institut für Polarökologie, Kiel
     Stefan Siebert(2), Alice           (2) Institut für Zoologie, Kiel
     Schneider(1) Karel Bakker(3),      (3) Royal Netherlands Institute for 
     Lilith Kuckero(4)                      Sea Research, Texel
                                        (4) Alfred-Wegener-Institut für 
                                            Polar- und Meeresforschung, 
                                            Bremerhaven

Background and objectives

Sea ice covers large areas of the polar oceans: in the Arctic, approx. 7 
million km2 are covered with ice in summer time, 14 million km2 in winter 
time. This ice cover plays a crucial role not only for geophysical processes, 
but also for the biology in the polar regions. Sea ice is not a solid block, 
but a matrix permeated with brine channels, which vary in diameter from 
micrometers to centimetres. These brine channels make up one habitat for a 
special community, the sympagic (ice-associated) community. It comprises 
viruses, bacteria, fungi, microalgae, protozoans and small metazoans. With 
temperatures between approx. 0°C and -22°C, the brine channels represent one 
of the coldest environments on earth. Salinity in the brine channels is 
coupled to the environmental temperature and rises when temperatures drop 
below the freezing point, as only water and not the contained solutes 
crystallize. The brine salinity can vary between approx. S = 220 at -22°C ice 
temperature and S = 2 - 3 during meltwater flushing in summer near 0°C.

Our studies focused on sea ice metazoans (also called sympagic meiofauna), 
multicellular animals in the size range of about 20 µm to 2 mm, as well as on 
underice amphipods. The meiofauna, which has been found in Arctic sea ice so 
far, comprises mainly of copepods, rotifers, turbellarians and nematodes. 
While sea ice algae received considerable attention since the beginning of 
biological sea ice studies, sympagic meiofauna has been studied only 
recently. Community composition in the Arctic pack ice and the factors 
influencing it are still not fully understood, especially as seasonality is 
concerned. Virtually nothing is known about the feeding ecology of sympagic 
meiofauna. These animals are however expected to play an important role in 
the sympagic ecosystem and in cryo-pelagic coupling, being a potential 
mediator of biomass and energy from algae, bacteria and protozoans to higher 
trophical levels.

The predicted loss of perennial sea ice in the Arctic Ocean will lead to 
major changes, if not destruction of an ecosystem, which is nearly as large 
as Australia (7 million km2) (Intergovernmental Panel on Climate Change, 
2007). Species dependent on perennial Arctic sea ice will probably be 
extinguished at the end of the current century. Therefore, studies on all 
topics of sea ice biology are urgently needed, in especially the 
identification of endangered species, description of their life cycles and 
physiology and the collection of genetic information (even complete genomes), 
as well as the observation and modelling of changes in the whole ecosystem, 
which might occur due to a change from perennial to seasonal sea ice cover.



Work at sea

Quantitative sampling Ice properties

The sympagic meiofauna community was analysed qualitatively and 
quantitatively, and the sea ice habitat characterized in terms of several 
environmental parameters. For this, several ice cores, cut in sections for 
vertical resolution, and pump samples of under-ice water were taken at eleven 
standard stations (Fig. 3.5.1, Table 3.5.1). At most stations eleven ice 
cores were drilled with a motor-powered KOVACS ice corer (internal diameter: 
9 cm) in areas of non-deformed ice (Table 3.5.2). The lowermost part of 
completely sectioned cores was cut into five segments of 1 cm length, the 
next segment was 5 cm thick, then a segment of variable thickness followed 
and the remaining parts of the cores were cut into segments of 20 cm. One 
core was taken to analyse in-situ temperature directly on the floe. Two cores 
were retrieved and stored at -20°C, one core for texture analysis (AG Sea ice 
physics, AWI), another one to search for the presence of carbonate crystals 
within sea ice (AG Sea ice biology, AWI). Another core was retrieved, 
sectioned and melted directly at 4°C in the dark for determination of bulk 
salinity, chlorophyll a (chl a) and phaeopigment (phaeo) content, as well as 
nutrient concentrations (phosphate, silicate, nitrate and nitrite). 
Furthermore aliquots from this core were preserved in formalin/hexamine for 
counting of algae. For determination of chl a and phaeo a determined volume 
was filtered onto whatman CF/F filters, extracted with 90% acetone, 
homogenized and measured fluorometrically. The fifth core taken during each 
standard station was cut into sections, melted directly in the dark at 4°C 
and a determined volume was filtered onto precombusted CF/F filters. These 
were stored at -80°C and will be used for the determination of stable carbon 
and nitrogen isotopes of the particulate matter. Ice-core sections for 
meiofauna analyses were generally melted with an excess of 0.2 µm filtered 
seawater (FSw, 200 ml per 1 cm of ice core) and enriched over a 20 µm gauze. 
One meiofauna core from each standard station was sectioned completely, 
worked up as described above and then fixed with 2% borax buffered 
formaline. These samples will be analysed for species composition, abundance 
and biomass in the home laboratory. For each of the eleven standard stations, 
meiofauna from three ice-core bottom sections (lowermost 5 cm of the cores) 
was sorted and counted alive onboard the vessel. One bottom section 
(lowermost 10 cm) was fixed with ethanole and stored at 4°C for determination 
of unidentifiable objects (e.g. eggs) with DNA sequencing techniques. Two 
bottom sections (lowermost 2 cm) were immediately flushed with 50 ml FSw, the 
resulting liquid filtered over a 20 µm gauze, the samples fixed in 
glutaraldehyde or picric acid formaldehyde (PAF) and stored at 4°C. These 
samples will be used for gut content analyses in the home laboratory using 
transmission and scanning electron microscopy. Another 10 cm bottom section 
was also melted in an excess of FSW, and approx. 200 ml of the resulting 
sample was thereafter filtered onto a 0.02 µm filter and the filter stored at 
-80°C. These samples will be used for analysis of hidden diversity within sea 
ice also using DNA sequencing techniques.



Sub-ice layer

At each standard station, temperature and salinity profiles in the sub-ice 
water layer (0 - 6 m below the ice underside) were measured in-situ with a 
conductivity meter lowered through a core hole. Discrete water samples for 
analysis of algal pigments were collected from 0 m and 5 m depth below the 
ice with a polyethylene tube with a valve at one end. The unequipped end of 
the tube was lowered into the water through a core hole with the valve 
closed. At the sampling depth, the valve was opened and closed again and the 
tube with the enclosed water sample was hoisted to the surface. An integrated 
sample from 0 - 10 m water depth under the ice was taken, leaving the valve 
open while lowering the tube, and closing it at the sampling depth. For 
determination of chl a and phaeo concentration, samples were treated as 
described above. Also subsamples for determination of nutrient concentration 
and algal composition were taken. Organisms from the under-ice water (0 m and 
5 m depth below the ice) were quantitatively sampled with an under-ice 
pumping system equipped with a standardized water meter and inserts of 
plankton gauze (mesh size 55 µm) to concentrate the organisms. Samples were 
fixed with 2% borax buffered formaline. Enumeration of species and stages 
from the sub-ice layer will be done in the home laboratory. At some of the 
stations, also some specimens from nonquantitative pump samples (0 m) were 
fixed in glutaraldehyde and PAF for gut content analyses. Abundances of 
under-ice amphipods will be estimated from under-ice video images taken during 
ten standard stations.


Other ice habitats

Samples for analysis described above (except for live counting and gut 
content analysis) were also retrieved from several melt pools and from newly 
formed ice in melt pools and leads (Fig. 3.5.1, Tab. 3.5.1).


Qualitative sampling

Meio- and macrofauna was also collected for various experiments and analyses 
concerning feeding ecology. Individuals were isolated alive from bottom 
sections of ice cores, from ice pieces collected by ice fishing (through a 
hole in the bottom of a mummid chair), from newly formed ice in melt pools 
and from under-ice water and zooplankton net samples. Part of the experiments 
were already conducted onboard. For further experiments in our home 
laboratory, cultures of sympagic meiofauna and sympagic algae were 
established during the cruise.

For morphological studies with light or electron microscopy in the home 
laboratory, single organisms isolated form qualitative samples were fixed in 
2 - 4% borax buffered formalin, glutaraldehyde or PAF.

In order to gain information about the role of sympagic meiofauna for the 
sympagic food web, we conduct feeding experiments with dominant sympagic 
meiofauna species as predators or grazers and different sympagic meiofauna, 
protozoans or ice algae as food. In basic predator-prey experiments, only one 
prey taxon is offered to the predators and the influence of predator and prey 
density on the predation rate (functional response, concurrence) are studied. 
Additionally, in food-choice experiments, different food is offered to the 
meiofauna at the same time, and preference for particular food sources is 
studied. Grazing experiments on ice algae are conducted not only in 
suspensions of algae but also on surfaces, in order to simulate in-situ 
conditions.

For the biochemical analyses of fatty acids as well as stable carbon and 
nitrogen isotopes, ice and sub-ice fauna was sorted alive onboard, and stored 
deep-frozen (-80°C) until analysis in the laboratory. Several fatty acids are 
not produced by the animals themselves, but are taken up with their food. 
Fatty acids can thus present information about the diet consumed in-situ. The 
stable isotope ratios of carbon (δ13C) and nitrogen (δ15N) provide a 
time-integrated measure of the trophic position in a certain food web, and 
can also provide information on the major carbon source of an organism.

Also the above-mentioned gut-content analyses will give information about 
in-situ diets of sympagic meiofauna and sub-ice fauna.


Preliminary results

In total, samples were collected on 31 stations (Fig 3.5.1). On eleven 
stations, so called standard stations, the above described complete sampling 
programme of the ice proper and the sub-ice layer was performed. On eleven 
stations, newly formed ice on melt ponds or in leads was quantitatively 
sampled for meiofauna abundance, chl a and phaeo concentration, POC/PON, 
nutrients, algal composition and bulk salinity. During all 31 stations 
qualitative samples were taken, in order to gather material for experiments, 
and for lipid and isotope analyses


Fig. 3.5.1: Sampling stations plotted together with the cruise track and ice 
            concentrations derived by satellite imagery for the 6 Sept. 
            crosses: qualitative sampling; circles: standard stations; 
            diamonds: thin ice sampling. (Map provided by Gunnar Spreen, IFM 
            Hamburg)


Tab. 3.5.1: Date, time, position and kind of programme performed

  Date       Time 
(yymmdd)     (UTC)       Lat       Lon    Programme
--------  -----------  -------  --------  ---------------------------------
 070802   13:30-21:00  81°57'N   34°02'E  standard
 070803   18:00        82°24'N   34°02'E  ice fishing qualitative
 070804   10:55-12:20  82°48'N   33°45'E  qualitative (Heli)
 070805   10:00-15:00  83°30'N   33°57'E  qualitative
 070806   16:00        84°07'N   34°41'E  ice fishing qualitative
 070807    9:30-17:00  84°29'N   36°08'E  standard
 070811   16:15        84°32'N   60°37'E  ice fishing qualitative
 070812   12:00-18:00  83°38'N   60°23'E  standard
 070815   13:00-18:00  82°30'N   60°48'E  standard
 070817   12:30-12:50  82°08'N   69°12'E  ice fishing qualitative
 070820   13:30-19:00  82°09'N   86°20'E  standard
 070823   14:10        83°49'N   88°06'E  qualitative
 070823   15:10        83°51'N   88°16'E  Qualitative
 070823   16:00        83°52'N   88°22'E  qualitative
 070824    9:30-17:30  84°35'N   89°50'E  standard
 070827   21:50-4:00   87°03'N  104°48'E  standard
 070830   18:35        88°21'N  142°51'E  ice fishing qualitative
 070831   18:30        88°08'N  150°05'E  standard, meltpond qualitative
 070902   14:00-18:00  87°49'N  170°36'W  Meltpond quantitative
 070905   18:30-23:15  85°42'N  135°02'W  qualitative
 070907    9:30-12:00  84°30'N  138°22'W  Meltpond quantitative
 070907   19:30-19:50  84°30'N  138°26'W  ice fishing qualitative
 070910   13:30-18:00  86°38'N  177°33'E  standard, meltpond quantitative
 070912-  22:15-10:00  86°24'N  135°49'E  standard, meltpond quantitative
  070914                                    
 070916   12:30-20:00  84°39'N  102°44'E  standard, meltpond quantitative
 070917   10:00        84°11'N  108°56'E  ice fishing qualitative, thin
                                            ice quantitative
 070917   16:45        84°06'N  109°57'E  ice fishing qualitative, thin
                                            ice quantitative
 070918    1:00        83°46'N  113°08'E  ice fishing thin ice quantitative
 070918    6:45-7:25   83°25'N  115°25'E  ice fishing thin ice quantitative
 070918   14:20-15:40  83°04'N  116°56'E  ice fishing thin ice quantitative
 070919    0:20-1:10   82°37'N  118°24'E  ice fishing qualitative, brash
                                            ice quantitative


Tab 3.5.2 Parameters and samples gathered during standard ice stations

   parameter                             ice cores    sub-ice water
   -----------------------------------  -----------  ----------------
   temperature                           complete         0-6 m
   (bulk) salinity                       complete         0-6 m
   brine salinity                        complete          --
   relative brine volume                 complete          --
   POC, PON, δ13C, δ15 N                 complete    0 m, 5 m, 0-10 m
   NO2, NO3, PO43, SiO4                  complete    0 m, 5 m, 0-10 m
   chlorophyll a, phaeopigment a         complete    0 m, 5 m, 0-10 m
   algae                                 complete    0 m, 5 m, 0-10 m
   fauna abundance, diversity           5 cm bottom        --
     (alive)                             sections
   fauna abundance, diversity             > 20 µm        0 m, 5 m
     (formaldehyde fixed)  
   fauna diversity                       complete        > 55 µm
     (ethanole)  
   fauna gut content                      > 20 µm          --
     (PAF, glutharaldehyde fixed)  
   fauna diversity, taxonomy,          10 cm bottom         0 m
   morphology(PAF, glutharaldehyde,      sections
     formaldehyde fixed, deep frozen)   
   fauna fatty acids / alcohols           > 20 µm        > 55 µm
   fauna δ13C, δ15 N                    2 cm bottom        --
                                         sections
   hidden diversity                       > 20 µm          0 m


Standard stations

In order to describe the typical environmental parameters found within the 
ice and in the sub-ice layer data from the stations on the 7 August and 13 
September are presented and discussed in the following.

Ice thickness on 7 August was 150 cm with a positive freebord of 8 cm, on the 
13 September an ice thickness of 101 cm with a positive freebord of 9 cm was 
found. On 7 August temperature at the surface of the ice was higher at about 
-0.1°C and decreased monotonously to -1.1°C at the bottom (Fig. 3.5.2). On 13 
September temperature at the surface was lower at -1.5°C, increased then to 
-0.8°C at 22.5 cm depth and then decreased again to -1.5°C at the bottom. 
Hence freezing in the upper parts of the ice occurred, due to low atmospheric 
temperatures, but heat loss from the ocean did not take place through the ice 
proper as otherwise a monotonous profile could be expected. Temperatures in 
the sub-ice layer during the station on 7 August showed a typical summer 
profile with higher temperatures of -1.5°C at 0 cm and 10 cm, -1.6°C at 20 
cm and 30 cm directly under the ice and thereafter a temperature of -1.7°C 
until 600 cm below the ice. This indicates bottom melting of the ice, which 
is supported by the salinity data. These show a freshened layer of 20 cm 
thickness (S = 30.1 - 31.5) directly under the ice, thereafter salinity rose 
to 33.5 and stayed almost constant down to 600 cm depth. In contrast on 13 
September salinity and temperature were homogenous in the first 600 cm under 
the ice. Temperature was at -1.7°C and salinity at 31.2, which indicates 
that bottom melting stopped and heat was transferred to the atmosphere from 
the ocean through leads and melted through melt pools.


Fig. 3.5.2: Temperature and bulk salinity as measured within the ice on 7 
            August and 13 September


Chl a profiles from both stations show a typical L-shaped profile (Fig 
3.5.3), with lower chl a concentrations around 1 µg/L in the upper part of 
the core, rising to high chl a concentrations in the bottom layer, with 46.0 
µg/L on 7 August and 6.9 µg/L on 13 September. Low chl a concentrations in 
the upper part were probably due to nutrient limitation (data not shown). Chl 
a concentration in the water column on 7 August at 0 cm depth below the ice 
was 3.7 µg/L, at 500 cm depth it was 0.4 µg/L. Higher concentrations directly 
under the ice are probably due to a release of algae from the ice through 
bottom melting. On 13 September chl a concentrations directly under the ice 
were lower with 0.44 µg/L at 0 cm depth. At 500 cm depth the same 
concentration as on 7 August of 0.4 µg/L was found. The lower chl a 
concentration directly under the ice coincides with the end of bottom melting 
and release of ice algae to the water column.


Fig. 3.5.3: Chlorophyll a concentrations in the ice (melted sample; upper 
            panel) and sub-ice layer (lower panel) on 7 August and 13 
            September


The described environmental parameters will help to interpret the meiofauna 
abundances throughout the ice and in the water column. These will be the 
result of sorting, identification and counting of the species to be found in 
the formalin fixed samples in the home laboratory.



Sea ice meiofauna

Bulk abundances of metazoans in ice-core bottom sections (alive counts on 3 
replicates for each of the 11 standard stations) ranged between 9 Ind L-1 and 
375 Ind L-1 (in melted ice); abundances of protozoans were an order of 
magnitude higher (Fig. 3.5.4). The sea ice metazoans were dominated by 
rotifers, red acoel platyhelminthes, nematodes, the harpacticoid copepods 
Halectinosoma sp. and Tisbe sp., and different cyclopoid copepods in live 
counts of bottom sections of the 11 standard stations. Also white acoel 
platyhelminthes with similar shape as the red ones were found, but only in 
low numbers. Among protozoans, ciliates were dominant, but also the 
foraminifer Neogloboquadrina pachyderma occurred in high numbers. These 
findings are in accordance with results from former cruises. High abundances 
of eggs were found in the ice especially during the second half of the 
cruise, when also egg-carrying females of the copepod Halectinosoma sp. were 
observed.


Fig. 3.5.4: Bulk abundances of sea ice meio fauna in ice core bottom sections 
            from 3 replicates for each of the 11 standard stations. a: 
            protozoa, metazoan eggs and metazoa, b: only metazoa


All 11 stations differed significantly in terms of sympagic meiofauna 
communities (global ANOSIM based on Bray-Curtis similarities, calculated from 
fourth-root transformed abundances of metazoans in bottom sections counted 
alive; significance level 5%), i.e. variability between stations was 
generally considerably higher than variability between replicate cores. Pair 
wise comparison of stations (pair wise ANOSIM), as well as multidimensional 
scaling and hierarchical agglomerative clustering, revealed no distinct areal 
pattern and showed that at some stations patchiness was high.

In addition, mainly in qualitative samples, we found some taxa, which have 
not been described for the Arctic sea ice as yet. Elongated white acoel 
plathyelminthes as well as elongated white plathyelminthes with eye spots 
(probably rhabditophores) were found at some stations, usually in low 
numbers. Adults and copepodides of an unidentified calanoid copepod (probably 
Limnocalanus sp.) were found in the ice at two stations. Most interestingly, 
we also found cnidarians, in the ice, which is the first record of sympagic 
cnidarians for this area, and the second record of sympagic cnidarians for 
the Arctic.

The newly formed thin ice on melt ponds, which could be found during the 
second half of the cruise, was often inhabited exclusively by ciliates, which 
were always the dominant heterotrophs, and rotifers. Sometimes also red acoel 
plathyelminthes and few individuals of other meiofauna taxa were found.



Sub-ice fauna

The sub-ice fauna was dominated by Oithona spp., different copepod nauplii, 
Calanus finmarchicus, C. glacialis, and an unidentified calanoid copepod 
(probably Limnocalanus sp.). Also the foraminifer N. pachyderma was frequent 
in the sub-ice layer. In addition, some other calanoid copepods (Acartia sp., 
Metridia longa), few rotifers and single individuals of the ptereopod 
Limacina sp. were found. Amphipods found associated with the ice were 
Gammarus wilkitzkll, Apherusa glacialis and Onisimus spp. Preliminary 
analysis of the under-ice videos revealed also the presence of ctenophores 
and the polar cod Boreogadus saida in the sub-ice layer. Abundances and 
diversity of the amphipods seem to be higher in the vicinity of Svalbard, 
Franz Josef Land and Severnaya Zemlya than in the Central Arctic.



Food web studies

For gut content analyses with the electron microscope, altogether 40 samples 
(20 each for transmission and scanning electron microscopy) were collected at 
13 stations. 32 of these samples are from different ice habitats (level ice 
and newly formed ice on melt ponds), eight samples are from the sub-ice 
habitat.

For analyses of fatty acids and alcohols, or of stable carbon and nitrogen 
isotopes, 139 samples were collected, comprising samples for ten different 
ice taxa (67 samples) and six different sub-ice taxa (72 samples). For most 
taxa, replicate samples could be taken. Most of the samples comprised animals 
from only one station, so that possible regional differences can be detected. 
Samples from different types of habitats (sub-ice, first-year ice, multiyear 
ice, thin ice on melt ponds) taken for several taxa will give insight into 
the influence of the habitat on feeding.

The data from the gut content and biochemical analyses will give a 
comprehensive picture of the in-situ feeding habits of all dominant sympagic 
metazoans.


Predation experiments were conducted onboard the vessel using

  • cnidarians as predators and rotifers as prey, with varying predator and 
    prey densities
  • cnidarians as predators and copepod nauplii as prey
  • the copepod Halectinosoma sp. as predator and sympagic ciliates as prey, 
    with varying predator densities
  • red and white acoel plathyelminthes as predators and ciliates as prey.

The experiments revealed that the cnidarians prey on rotifers as well as on 
copepod nauplii, being able to ingest prey individuals of their own body 
size. Halectinosoma sp. preys on sympagic ciliates, with predation rates 
depending on predator densities. Predation of turbellarians on ciliates was 
not observed. Further predation experiments, as well as grazing experiments 
and food-choice experiments, will be conducted at the home laboratory. For 
this purpose, cultures of all dominant sympagic meiofauna species and of ice 
algae have been successfully established.

The data from the feeding experiments conducted during the cruise, and from 
further experiments will give predation rates and information about 
concurrence / functional response of the sympagic organisms. Furthermore, 
viewed in conjunction with abundances of sympagic metazoans, ciliates, and 
algae, it will be possible to estimate the predation and grazing impact of 
sympagic metazoans, and hence their overall role within the ecosystem.




4.  OCEANOGRAPHY

4.1  Physical oceanography

      Takashi Kikuchi(3), Sebastian      (1)Alfred-Wegener-lnstitut
      Mechler(4), Sergey Pisarev(5),     (2)Finnish Institute of Marine Research
      Benjamin Rabe(1), Bert Rudels(2),  (3)Japan Agency for Marine Earth 
      Ursula Schauer(1), Andreas            Science and Technology
      Wisotzki(1)                        (4)Optimare
                                         (5)P.P. Shirshov Institute of Oceanology


Objectives

The circulation and water mass properties of the Arctic Ocean have been 
changing considerably during the past decades. The aim of the oceanographic 
part of this cruise was devoted to the IPY theme of quantifying the ocean 
changes by documenting the present state of the water mass distribution in 
the Eurasian basins and shelf seas. Due to the light ice conditions we could 
even include the northern Canada Basin.

Water advected from the North Atlantic has become warmer since the early 
nineties and the question is still open whether the additional oceanic heat 
has contributed to the decrease of the sea ice. The Atlantic inflow has also 
become more saline, but at the same time the hydrological cycle has increased 
and with that the fresh water supply to the Arctic. Less sea ice might in 
turn have affected the interaction between ocean and atmosphere and altered 
the modification of the water on the Eurasian shelves and even in the central 
Arctic Ocean. Furthermore, changing atmospheric patterns (Arctic Oscillation) 
influence the input, storage and circulation of the fresh water which is 
supplied to the Arctic Ocean through continental runoff, precipitation, and 
Pacific water inflow and which shields the warm Atlantic water from the 
surface.

Specific questions are: Has the warmer inflow to the Barents Sea lost all its 
additional heat before entering the central Arctic? To what extent is the 
Fram Strait branch water exposed to the surface while flowing along the 
continental slopes and ridges and where is it isolated from the surface by 
the fresh water layer of the Siberian river run-off? Is Atlantic water 
upwelling at the shelf edge so that its heat can be released? Where do Fram 
Strait and Barents Sea branches mix and which fractions recirculate at which 
time scales towards Fram Strait within the various basins? What is the 
current position and direction of the Transpolar Drift of fresh waters from 
the Pacific and from the Siberian rivers? Can we identify individual pulses 
of warm saline inflow and speculate about the timing of their outflow to the 
subarctic Atlantic?

To address these questions hydrographic sections were repeated that were 
taken in the Eurasian Basin during the cruises Oden 1991 and 2005 as well as 
Polarstern 1993, 1995 and 1996.


Fig. 4.1.1: Location of CTD stations. Numbers denote sections. AB: Amundsen 
            Basin, NB: Nansen Basin, MB: Makarov Basin, CB: Canadian Basin


Work at sea

Ship-borne measurements

Along the five sections (Fig. 4.1.1), altogether 191 CTD profiles were taken 
at 127 stations and water samples were collected. 142 casts were carried out 
with a standard CTD/rosette water sampler (described below) and 49 casts were 
taken with the ultra-clean system of the GEOTRACES programme (described in 
chapter 5). Both systems had Seabird CTD components (SBE 911+) with double 
temperature and conductivity sensors. The standard CTD system, SN 485, from 
Sea-Bird Electronics Inc SBE911+ was equipped with duplicate temperature and 
conductivity sensors (temperature sensors SBE3, SN 2423 and 2460, 
conductivity sensors SBE4, SN 2078 and 2054 and pressure sensor Digiquartz 
410K-105 SN 68997). The CTD was connected to a 5BE32 Carousel Water Sampler, 
SN 202, with 24 12-liter bottles. Additionally, a Benthos Altimeter PSA-916, 
SN 1228, a Wetlabs C-Star Trans missometer, SN 946, and a SBE 43 dissolved 
oxygen sensor, SN 743, was mounted on the carousel. The SBE 43 contains a 
membrane polarographic oxygen detector. The algorithm to compute oxygen 
concentration requires also measurements of temperature, salinity and 
pressure which are provided by the CTD system. To calibrate the oxygen 
profiles 279 water samples at 30 CTD casts were collected and measured 
onboard with Winkler titration. Continuous profiles of the DOM concentration 
were obtained with two fluorometers, a Wetlabs ECO-CDOM, SN 742, and a Dr. 
Haardt Back Scat Model 1184.6, SN 12030. Salinity of 245 water samples was 
measured using a Guildline salinometer with Standard Water Batch P148 for 
calibration of the salinity sensor. At 15 stations, 60 water samples at 5 l 
were collected for Technetium measurements.

In areas with heavy ice, the sections were extended by helicopter-borne XCTD 
casts. XCTDs were also launched between CTD casts to increase the horizontal 
resolution in frontal zones. Underway measurements with a vessel-mounted 
narrow-band 150 kHz ADCP from RD Instruments and two Sea-Bird 5BE45 
thermosalinograph were conducted to supply temperature, salinity and current 
data. The thermosalinographs are installed in 6 m depth in the bow thruster 
tunnel and in II m depth in the keel. The salinity of both instruments was 
controlled by taking water samples. Unfortunately, already at the first 
section, on 4 August, the ADCP failed and no current measurements were taken 
thereafter.

In order to provide year-round measurements of temperature, salinity, 
velocity and under-ice turbulence, ice-tethered platforms (ITPs) with various 
instrumentation were deployed. These platforms contribute to an 
"International Arctic Ocean Observation System" (iAOOS) that aims at a 
persistent observation network.

The oceanographic work of the cruise was part of the EU-funded Integrated 
Project "DAMOCLES" (Developing Arctic Modelling and Observing Capabilities 
for Long-term Environment Studies) and the BMBF-funded Project 
"North-Atlantic".

First results of the CTD observations

The oceanography work started on Sunday, 29 July 2007, with a station at 
72°35'N and 26°0'E to test the two CTD systems, the AWI standard CTD/rosette 
and the NIOZ ultra-clean (UC) Titan CTD system. All bottles were closed to 
test for leaking and rosette malfunctions. All sensors and bottles were found 
to work properly. During the cruise there were, however, constant problems 
with position 4 on the AWI rosette and in the end it was considered 
unreliable, leaving 23 working bottles on the AWI rosette.

The sampling of the first real section, taken along the 34°E meridian 
commenced at 75°N on Monday 30 July at 10 am with station 228 using the UC 
CTD. The AWI rosette was, however, considered the main instrument and was 
used at most stations. On several stations both CTD systems were used, 
sometimes with multiple casts to supply the required water volumes. Only at a 
few stations just the UC CTD was used to save time. On the sections prepared 
and shown here the stations with only UC CTD have been merged with the AWI 
CTD casts. Especially the AWI CTD was found to well reproduce the numerous 
intrusions observed in the water column, while the UC CTD data will need 
extensive post cruise editing to remove spikes and instabilities in the 
profiles. The conductivity sensors on both CTDs have been preliminary 
calibrated against salinity samples measured onboard during the cruise and 
the values shown on the sections are reliable.

Section 1 (Fig. 4.1.2) started in the Barents Sea over the Central Bank and 
ran northward across the Grand Bank and the continental slope into the Nansen 
Basin, almost extending to the Gakkel Ridge. Because of heavy ice and slow 
progress the section was terminated with station 260 at 84°30'N and 36°09'E 
before the Gakkel Ridge was reached. The section captured the Atlantic water 
entering through the Barents Sea Opening between Norway and Bear Island, 
which this year also covered the Central Bank. The salinity was high, >35.1, 
and the temperature of the upper layer were above 6°C. The polar front was 
located at the southern edge of the Grand Bank and above the bank the upper 
salinity was lower, partly due to ice melt, partly due to the presence of 
less saline Arctic water. A temperature minimum, T< -1.6°C, with salinity 
34.5 - 34.6 was located between 50 and 100 m, indicating the depth and 
characteristics of the winter homogenised upper layer. The deeper layers 
above the bank and in the northern depressions were occupied by Atlantic 
water, less saline and less warm than farther to the south.

The Fram Strait Atlantic water core was located above the continental slope, 
close to the shelf break and close to the sea surface, with its thermocline 
less than 100 m deep. The surface heating and mixing as well as the mixing of 
heat from below increased the temperature at the shelf break and thus has 
obviously separated the winter mixed layer in the Barents Sea from the less 
saline, but colder winter mixed layer of the Nansen Basin. The Atlantic water 
at the slope was warm, occasionally above 3°C and saline, above 35. Below the 
Atlantic core, the weak salinity minimum of the Arctic Intermediate Water 
(AIW) at about 1,000 m and the Nordic Sea Deep Water (NDW), both less saline 
than the Eurasian Basin Deep Water (EBDW), could be recognised.

Farther offshore the temperature and salinity of the Atlantic water decreased 
and intrusions were observed. Below the Atlantic core a salinity minimum was 
present, slightly less dense and located higher in the water column than the 
AIW at the slope. The salinity minimum was also distinguished by many 
small-scale intrusions. This minimum, being colder than the upper layer, 
stems from the Barents Sea while the upper water is one loop of the Fram 
Strait inflow branch that recirculates within the Nansen Basin along the 
Gakkel Ridge.

The second section (Fig. 4.1.3) ran from 84°39'N, 60°57'E southward to the 
Barents Sea shelf northwest of Franz Josef Land. Only on the northernmost 
station were the characteristics of the Barents Sea branch water and the 
return circulation of Fram Strait water to be seen. Further south, the 
warmer, more saline, newly entered water from the west dominated. The 
southern part of the Atlantic core was even closer to the sea surface, at ~50 
m depth, than it was at the 34°E section and here the temperature minimum 
from the winter homogenisation of the upper layer was absent.

This suggests that Atlantic water is entrained into the upper mixed layer and 
may influence and reduce the ice formation. Close to the slope less saline 
and colder water was present, as a detached eddy in the upper 300 - 600 m on 
one station and at the slope as intrusions around 300 - 500 m, close to the 
bottom. This indicates an outflow of water from the Barents Sea, presumably 
flowing northward along the Victoria Channel.

After section 2 was finished on 17 August with station 271 Polarstern sailed 
eastward along the shelf and slope, occupying stations north of Franz Josef 
Land and in the western part of St. Anna Trough. North of Franz Josef Land 
similar cold, low salinity intrusions into and below the Atlantic water were 
observed as at 60°E, showing the persistence of the outflow from the northern 
Barents Sea. In the St. Anna Trough a colder layer, denser than the outflow 
from the Victoria Channel, was present. This would then be a part of the main 
inflow of the Barents Sea branch water to the Arctic Ocean.

In the northern Voronin Trough, some oceanography stations were then occupied 
in connection with geology work. The main section (Fig. 4.1.4) extended from 
the Kara Sea shelf west of Severnaya Zemlya at 81'34'N, 86°11'E across the 
Nansen Basin and the Amundsen Basin and then across the Lomonosov Ridge into 
the Makarov Basin. It repeated a long section taken by Polarstern in 1996 as 
the section along 34°E was a repeat of the Polarstern section taken in 1987 
and the section along 60°E towards Franz Josef Land was roughly similar to a 
Polarstern section taken north of Franz Josef Land in 1993. The new section 
would therefore, in addition to the extended spatial coverage also bring a 
temporal dimension to the observations.

The scattered stations on the Kara Sea shelf as well as the southernmost 
stations on the section along 86°E showed cold, low salinity water below a 
warmer, more saline Atlantic core. The Atlantic water was, however, colder 
and less saline than that observed at the slope farther to the west, 
indicating that the northern Kara Sea shelf was dominated by water from the 
Barents Sea inflow branch. The water was colder and denser than the Barents 
Sea outflow from the Victoria Channel and also colder and denser than that 
observed in the western St. Anna Trough. This confirms earlier findings that 
the Barents Sea branch enters the Arctic Ocean along the eastern flank of the 
St. Anna Trough and then turns eastward, following the Kara Sea continental 
slope and also makes a turn into the Voronin Trough west of Severnaya Zemlya.

On the shelf and above the upper continental slope the temperature minimum 
layer, located at about 50 m depth, was more saline than that found in the 
interior Nansen Basin, 34.5 as compared to 34.2, indicating different sources 
for the upper layers in the northern Kara Sea and the interior of the Nansen 
Basin. The higher salinity of the temperature minimum and the lower 
temperature in the Atlantic layer below suggest that the upper layer observed 
here initially has formed through the interaction between sea ice and the 
cooler Barents Sea branch, while the upper layer in the Nansen Basin is 
derived from the interaction between sea ice and the Fram Strait inflow 
branch north of Svalbard.

These two temperature minima were separated from each other by a narrow strip 
without a temperature minimum, located where the Atlantic water core rose 
close to the sea surface (50 m). This situation indicates the possibility of 
Atlantic water here being entrained into the mixed layer, providing heat to 
the atmosphere and to ice melt.

The denser part of the Barents Sea branch was present as a wedge above the 
continental slope, extending down to 1,200 - 1,500 m. Strong intrusions and 
interleaving between the waters of the Barents Sea branch and the Fram Strait 
branch were observed in the Atlantic layer as well as in the intermediate 
water below. The intrusions were seen on the slope at the front between the 
two branches. Beyond the front a narrow band of Fram Strait branch water 
contained almost no intrusions. On the other side of this Fram Strait branch 
the intrusions again increased in prominence. However, the structure of the 
intrusions on either side of the Fram Strait branch was similar. The largest 
intrusions were found in the upper part between the temperature maximum and 
the salinity maximum, where the water column is stably stratified in both 
components. Below the salinity maximum one or two larger intrusions with 
several smaller features were observed, and below the salinity minimum, where 
again both components were stably stratified, several smaller intrusions were 
encountered. Such intrusions were observed both at the basin side and the 
shelf side of the Fram Strait branch and coincided with the coldest part of 
the dense Barents Sea branch inflow. Since the intrusions appear to be 
created by interaction between the two inflow branches the water column must 
recirculate somewhere farther to the east and bring the Barents Sea branch 
water and the intrusions to the offshore side of the Fram Strait branch, 
forming a narrow recirculation loop.

Beyond the warm (>3°C) and saline (>35) Fram Strait branch the core 
temperature and core salinity decrease to slightly above 2°C and 34.93 
respectively and stayed around these values almost up to the Gakkel Ridge. At 
the Gakkel Ridge a front appeared to be present and the temperature decreased 
to a little above 1°C and the salinity was reduced to below 34.9. Also the 
intermediate low salinity layer from the Barents Sea branch decreased in 
salinity to 34.86 - 34.87. These properties then held from the Gakkel Ridge 
over the entire Amundsen Basin to the Lomonosov Ridge, where the temperature 
and salinity again increased. The differences in the return flows in the 
Nansen and the Amundsen Basins could be explained by a larger prominence of 
Barents Sea branch water in the Amundsen Basin. Such explanation would not 
hold for the higher temperatures and salinities encountered on the Amundsen 
Basin side of the Lomonosov Ridge. One possibility is that different vintages 
of the inflow branches are observed. The warmer, more saline water at the 
Lomonosov Ridge would then represent a more recent and warmer Fram Strait 
inflow than the more sluggish Amundsen Basin circulation carrying colder 
water that has entered a few years earlier.

Because of the light ice conditions we found the time to make a detour to 
investigate a small intra-basin in the ridge, where a through-flow of deep 
water from the Makarov Basin to the Amundsen Basin had been observed by Oden 
in 2005. The intention was to determine if this flow was a permanent feature 
or if the flow direction of the densest water at the sill could change over 
time, to study the water mass structure in the intra-basin and to determine 
the characteristics of the densest water at the sill to the Makarov Basin. It 
has previously been suggested that to explain the characteristics of the 
Makarov Basin bottom water an inflow of Amundsen Basin deep water across the 
Lomonosov Ridge was necessary, and the sill at the intrabasin could provide a 
path for such an inflow. In 2005 such a flow was not detected and also now 
the densest water encountered at the bottom of the sill derived from the 
Makarov Basin. Therefore any deep flow from the Amundsen Basin to the Makarov 
Basin must either be very intermittent or perhaps the deep water circulation 
and exchanges have changed and the flow from Amundsen Basin to the Makarov 
Basin has been disrupted for a longer period, possibly as a response to 
warmer conditions in the Arctic. One station in the intra-basin was special. 
A thin, almost 2°C warm Atlantic layer was overlaying a 1,000 m thick low 
salinity intermediate layer, resembling an eddy of almost pure Barents Sea 
branch water. Similar characteristics were not observed elsewhere on the 
cruise.

The stations at the Makarov Basin side of the Lomonosov Ridge showed low 
salinity in the intermediate layers, indicating that the denser water of the 
Barents Sea inflow, was entering the Makarov Basin from the Amundsen Basin 
through the intra-basin and flowed along the Lomonosov Ridge back towards 
Siberia. The Atlantic core was cooler, <1°C, on the Makarov side than on the 
Amundsen Basin side (1.2 - 1.3°C). The water column below the Atlantic layer 
was characterised by decreasing temperature and increasing salinity with 
depth, making the water column stable in both properties. This intermediate 
layer was more saline and warmer in the Makarov Basin than in the Amundsen 
Basin and the curves in the q-S diagrams were almost straight with little 
tendency toward lower temperatures and salinities. Closer to the Alpha Ridge 
the intermediate water again became cooler and less saline at mid-depth 
suggesting the presence of a second inflow of less saline Barents Sea branch 
intermediate water.

As Polarstern crossed the Alpha Ridge into the Canada Basin large changes in 
water mass properties were observed. The temperature maximum of the Atlantic 
water dropped to 0.5°C and inside the Canada Basin the last station (342) on 
the section displayed very smooth θ-S characteristics and a temperature 
maximum of just 0.4°C. This was one of the few stations taken on the cruise 
where no intrusions were present in the Atlantic water core. This was the 
common situation in the southern Canada Basin 10 to 15 years ago, and the 
presence of such a water column in the northern Canada Basin close to the 
Alpha Ridge indicates that the Atlantic water is circulating around the 
Canada Basin, some leaking out along the continental slope to the Makarov 
Basin and partly continuing to the Eurasian Basin and Fram Strait, some 
re-enter the Canada Basin at the Alpha Ridge.

The intermediate water appeared to follow a similar circulation. The water 
below the coldest Atlantic water was warmer and more saline than that in the 
central Makarov Basin, suggesting that it contained not only contributions 
from the Barents Sea inflow but also from shelf-slope convection and the 
corresponding entrainment occurring within the Canada Basin.

The long section 3 extending from the Kara Sea to the Canada Basin was 
finished at 84°30'N, 138°25'W on 7 September, the 1991 North Pole day for 
Oden and Polarstern. The upper layers of the section showed the different 
freshwater contributions to the Arctic Ocean. In the Nansen Basin the low 
salinity upper layer was dominated by the seasonal melting. Below this thin 
melt water layer lies a slightly more saline mixed layer that was formed 
through ice melting on top of warm Atlantic water north of Svalbard in 
winter. In the Amundsen Basin a freshwater input from the Siberian shelf seas 
could be distinguished. This low salinity shelf water is formed through 
mixing between the river runoff and the Barents Sea inflow branch and is 
advected into the deep Arctic Ocean basins mainly across the shelf break in 
the eastern Laptev Sea and western East Siberian Sea. Because it is much less 
saline, the shelf water flows above the winter mixed layer that is advected 
in the boundary current from the Nansen Basin. The latter then appears as a 
halocline between shelf water and the Atlantic layer. Over the Lomonosov 
Ridge and in the Makarov Basin the input from the shelves was more 
pronounced, but the vertical structure was similar as in the Amundsen Basin. 
Closer to the Alpha Ridge and in the Canada Basin different contributions of 
the Pacific inflow were encountered and the freshwater content in the upper 
layer increased dramatically, as it became thicker and less saline.

Section 4 (Fig. 4.1.5) extended from the Canada Basin across the Mendeleyev 
Ridge and the Makarov Basin and across the Lomonosov Ridge into the Amundsen 
Basin. The Pacific water then gradually disappeared and the Makarov Basin was 
here mostly influenced by the Siberian shelf outflow. This makes sense, 
having a stronger signal of Pacific water closer to the North American 
continent, as it moves towards the Canadian Arctic Archipelago and Fram 
Strait, while the influence of the Siberian shelves remains closer to the 
Eurasian continent.

The Atlantic water temperature and salinity on section 4 closer to Siberia 
were similar to those observed on section 3. However, the intermediate water 
at the Mendeleyev Ridge was somewhat colder and less saline than that in the 
central Makarov Basin suggesting an inflow of Barents Sea branch water that 
first had crossed the Lomonosov Ridge close to the continental slope and then 
been injected into the Makarov Basin at the Mendeleyev Ridge. This inflow 
seemed to cross over to the Alpha Ridge providing the less saline, colder 
intermediate water observed there on section 3 (see above). At the Lomonosov 
Ridge the salinity and temperature of the intermediate water was even lower 
and had a similar shape as the intermediate water in the Amundsen Basin. The 
deep water, however, was warmer and more saline throughout the Makarov Basin 
up to the Lomonosov Ridge. On this second crossing of the Lomonosov Ridge the 
difference in temperature between the two cores of Atlantic water was equal 
to that observed on section 3. This indicates that the water flows in 
different directions on the two sides of the ridge, towards Greenland on the 
Amundsen Basin side and towards Siberia on the Makarov Basin side and that 
little interaction between the two streams takes place across the ridge. By 
contrast the intermediate water of the Barents Sea branch at the Amundsen 
Basin side of the ridge appeared to cross the ridge at several locations and 
return towards Siberia.

Section 5 (Fig. 4.1.6) started at the Gakkel Ridge and led southward onto the 
Laptev Sea shelf. Little change in properties of the water masses occurred on 
the first and central part of the section. The water column was characterised 
by a cooled and diluted Atlantic layer with temperature slightly above 1°C 
and salinity below 34.9. A depression in the Gakkel Ridge, reaching below 
5,000 m was sampled but nothing spectacular was observed there in the deep 
and bottom layers. At the continental slope the Atlantic water was above 2°C 
and more saline, 34.93 - 34.94, but the Atlantic core never reached the high 
temperatures and salinities observed on sections 1, 2 and 3. The Atlantic 
core displayed several intrusions and no undisturbed Fram Strait branch 
Atlantic water mass was seen.

The salinity of the Barents Sea branch intermediate water was at minimum over 
the central Gakkel Ridge and increased closer to the continental slope as the 
Atlantic water characteristics became more influenced by the Fram Strait 
inflow branch. Closer to the slope, where the temperature and salinity of the 
Atlantic layer again decreased the salinity minimum below the Atlantic core 
again became stronger, although not as prominent as farther into the basin. 
Between 1,000 and 2,000 m at the slope a warmer, more saline water mass was 
encountered. Such water mass was not present on section 3 and does not appear 
to originate from the Barents Sea branch, at least not without diapycnal 
mixing with some other water masses. At the deeper part of the slope, close 
to the bottom the salinity and at stations deeper than 2,500 m also the 
temperature increased towards the bottom. This could indicate shelf-slope 
convection with entrainment of intermediate water.

On the upper part of the slope the salinity increased from north to south in 
the upper layers showing that the Fram Strait branch winter mixed layer was 
replaced by the winter mixed layer from the Kara Sea associated with the 
Barents Sea branch. The temperature has increased as compared to section 3 
indicating interaction with the underlying Atlantic water, probably due to 
enhanced mechanical mixing at the continental slope.

On the shallow Laptev Sea the bottom water retained a fairly high salinity, 
33 - 34, and at the southernmost station at 46 m water depth, the salinity at 
the bottom was still close to 33. The upper 10 - 20 m, however, were 
characterised by low salinity, either due to ice melt or due to river runoff 
or a combination of both. On all stations except the last the bottom water 
was cold, below -1.5°C, while the upper low salinity surface layer had 
temperatures close to or above 0°C. The sampling of the last station, 411, 
was finished at 3 am on the 24 September at 75°12'N, 121°22'E.


Summary

The extremely light ice conditions allowed for the study of a much larger 
part of the Arctic Ocean than was initially planned. However, in the 
ice-covered area the ice melt of about 1.5 m was not extraordinarily large 
but in the ice-free part a substantially higher melting was observed.

The Barents Sea branch was characterised by a strong salinity minimum in the 
intermediate layer below the Atlantic water. However, the salinity changed 
strongly in space suggesting that the inflow was very variable in time and 
that the salinity of the Barents Sea inflow might have decreased in recent 
years.

The Atlantic water of the Fram Strait branch became much cooler and fresher 
between section 3 and section 5, i.e. between the eastern Kara Sea slope and 
the western Laptev Sea slope. This transformation might partly be due to 
mixing between the two branches, partly be due to a recirculation of the 
branches in the Nansen Basin. This is consistent with the intrusions and 
interleaving seen in the colder, less saline Atlantic layer offshore of the 
Fram Strait branch on sections 1, 2 and 3.

The salinity of the winter mixed layer in the Nansen Basin was low, 34.1 - 
34.2 as compared with the more commonly encountered 34.3-34.4. This could be 
due to the fact that the Atlantic water that entered through Fram Strait has 
become warmer in recent years. By contrast the Barents Sea and the Kara Sea 
winter mixed layer had salinities around or above 34.5.


Fig. 4.1.2a: Distribution of potential temperature (degC) along section 1 
             (see Fig. 4.1.1 for location)

Fig. 4.1.2b: Distribution of salinity (psu) along section 1 (see Fig. 4.1.1 
             for location)

Fig. 4.1.2c: Distribution of oxygen (µmol/kg) along section 1 (see Fig. 4.1.1 
             for location)

Fig. 4.1.3a: Distribution of potential temperature (degC) along section 2 
             (see Fig. 4.1.1 for location

Fig. 4.1.3b: Distribution of salinity (psu) along section 2 (see Fig. 4.1.1 
             for location)

Fig. 4.1.3c: Distribution of oxygen (µmol/kg) along section 2 (see Fig. 4.1.1 
             for location)

Fig. 4.1.4a: Distribution of potential temperature (degC) along section 3 
             (see Fig. 4.1.1 for location)

Fig. 4.1.4b: Distribution of salinity (psu) along section 3 (see Fig. 4.1.1 
             for location)

Fig. 4.1.4c: Distribution of oxygen (µmol/kg) along section 3 (see Fig. 
             4.1.1 for location)

Fig. 4.1.5a: Distribution of potential temperature (degC) along section 4 
             (see Fig. 4.1.1 for location)

Fig. 4.1.5b: Distribution of salinity (psu) along section 4 (see Fig. 4.1.1 
             for location)

Fig. 4.1.5c: Distribution of oxygen (µmol/kg) along section 4 (see Fig. 4.1.1 
             for location)

Fig. 4.1.6a: Distribution of potential temperature (degC) along section 5 
             (see Fig. 4.1.1 for location)

Fig. 4.1.6b: Distribution of salinity (psu) along section 5 (see Fig. 4.1.1 
             for location)

Fig. 4.1.6c: Distribution of oxygen (µmol/kg) along section 5 (see Fig. 4.1.1 
             for location)



4.2  XCTD observation
     Takashi Kikuchi 
     Japan Agency for Marine-Earth Science and Technology


Background

Boundary currents along the continental slope and the ridges of the Arctic 
Ocean carry the changing signals advected from the Atlantic and they form 
part of the origins of the North Atlantic Deep Water. They are very narrow 
and therefore need hydrographic observations with high resolution which - in 
accordance with an economic use of ship time - can be achieved with 
expendable CTDs.

The eXpendable Conductivity, Temperature, and Depth data acquisition and 
processing equipment (XCTD) can measure temperature and conductivity (i.e., 
salinity) from sea surface to 1,100 m depth in only five minutes. The XCTD 
system is manufactured by The Tsurumi-Seiki Co., LTD, Japan. It mainly 
consists of XCTD probe, launcher, digital converter and personal computer for 
data processing. The XCTD probe can be launched from the ship or on the ice 
into water and sinks down with constant velocity measuring temperature and 
conductivity. Accuracy of XCTD temperature should be better than 0.01°C. 
However, please note that XCTD salinity has large error rates of more than 
0.03 before calibration. By using closest or interpolated CTD data the XCTD 
salinity data should be calibrated if possible.

XCTD operation has some advantages compared with CTD observation although the 
accuracies in temperature and salinity is not better than CTD observation. 
One is that we can save ship time for the observation. XCTD observation down 
to 1,100 meter takes only 5 minutes. Another big advantage is that it is 
portable. The XCTD system with battery pack is less than 10 kg in weight. 
This allowed us to operate from helicopter. Hence, we used XCTDs between CTD 
stations to increase the spatial resolution and we extended some of the CTD 
sections by a line of helicopter-borne XCTDs.

Work at sea

89 XCTD casts at 84 stations were collected during this cruise (Table 4.2.1). 
Some XCTD stopped data sampling before reaching the bottom or full 
measurement depth (1,100 m) due to sea ice, software trouble, and so on. 24 
of 89 XCTDs were conducted by helicopter operation (Fig. 4.2.1), and others 
were launched from the ship (Fig. 4.2.2). Figure 4.2.3 shows map of the XCTD 
observation sites during this cruise as well as CTD observation sites. The 
map shows that we did XCTD observations mainly focused on across the European 
continental slope regions and major ridges of the Arctic Ocean with high 
resolution of 5~10 nm in space. Helicopter XCTD operations were conducted 
along 65°E across the Eurasian continental slope on 16 and 17 August, 85.7N 
over the Alpha Ridge on 5 September, and 140W toward the Canadian Basin on 7 
September.

After the observation, salinity calibrations of each XCTD data were carried 
out by using closest or spatially-interpolated CTD data. I used the closest 
or interpolated CTD data to check salinity values at the Atlantic water depth 
and at 1,000 meter depth. As results, the calibrated XCTD salinity data had 
conservative accuracies better than 0.01 psu. Figure 4.2.4 shows 
temperature-salinity diagram and temperature profiles of all of calibrated 
XCTD data which was submitted as the final version.

Preliminary results

Figure 4.2.5 shows some results of XCTD observations over Alpha Ridge and in 
the Canadian Basin (XCTD no. 46-62). The XCTDs were conducted by helicopter 
operation as extension lines of the section 3. Pacific water (PW) which is 
defined as temperature maximum around 32.7-33.0 psu in salinity can be found 
especially in the northern Canadian Basin. We can see the same signal over 
the Alpha Ridge but weak. In the Makarov Basin such PW signal are much 
smaller or not found. On the other hand, the lowest Atlantic water 
temperature observed during this cruise was found here (-0.38°C).


Tab. 4.2.1: Summary of XCTD data sampling date and location

Sta. No.  XCTD    Date      Time   Latitude     Longitude    Bottom  Measurement
           No.                                               Depth     depth
--------  ----  ---------  -----  ----------  -------------  ------  -----------
PS701240    1   2007/8/1   21:00  81 30 44.1   33 59 89.7 E    192      192
PS70/241    2   2007/8/1   22:02  81 34 58.6   34 0  4.9  E    230      230
PS70/242    3   2007/8/1   23:09  81 39 58.9   34 0  14.1 E    713      349
            4   2007/8/1   23:22  81 40 6.1    34 0  27.4 E    713      713
PS70/244    5   2007/8/2    6:18  81 44 54.7   33 59 33.3 E   1650     1100
PS70/245    6   2007/8/2    7:30  81 49 52.9   34 0  13.6 E   2000     1100
PS70/247    7   2007/8/2   12:03  81 55 18.4   33 59 0.0  E   2100     1100
PS70/249    8   2007/8/2   23:48  81 59 56.1   33 58 13.2 E   2240     1100
PS70/250    9   2007/8/3    6:38  82 4  56.8   34 0  22.2 E   2300     1100
PS70/251   10   2007/8/3    7:48  82 9  55.8   33 59 51.8 E   2530     1100
PS70/252   11   2007/8/3   12:20  82 15 18.0   34 1  38.2 E   2687     1100
PS70/253   12   2007/8/3   16:28  82 19 54.9   34 0  21.4 E   2800     1100
PS70/254   13   2007/8/3   18:10  82 25 2.0    34 2  28.1 E   3052     1100
PS70/255   14   2007/8/4    4:17  82 31 0.5    33 53 50.2 E   3083     1100
PS70/266   15   2007/8/14   8:36  83 6  46.6   61 41 42.8 E   2769     1100
PS70/268   16   2007/8/14  22:30  82 48 43.0   60 51 13.5 E   1668     1100
by Heli.   17   2007/8/16  18:02  82 40 23.3   65 38 1.6  E   1426     1100
by Heli.   18   2007/8/16  18:21  82 50 19.4   65 35 23.7 E   2024     1100
by Heli.   19   2007/8/16  18:39  83 0  6.0    65 29 10.2 E   2415     1100
by Heli.   20   2007/8/16  18:57  83 10 29.3   65 22 48.9 E   2656     1100
by Heli.   21   2007/8/16  19:17  83 20 6.1    65 16 35.4 E   2889     1100
by Heli.   22   2007/8/17  12:30  81 59 53.7   65 58 19.1 E    538      538
by Heli.   23   2007/8/17  12:56  82 9  55.7   65 55 37.1 E    560      560
by Heli.   24   2007/8/17  13:11  82 19 44.4   65 49 42.6 E    691      691
by Heli.   25   2007/8/17  13:29  82 30 13.3   65 45 34.7 E   1102     1100
PS70/304   26   2007/8/25  21:23  85 18 31.9   90 14 45.1 E   3655     1100
PS70/309   27   2007/8/28  12:25  87 4  10.3  104 37 51.0 E   4269     1100
PS70/311   28   2007/8/29   6:23  87 49 21.7  113 12 56.0 E   4339      588
           29   2007/8/29   6:33  87 49 21.7  113 12 56.0 E   4339     1100
PS70/312   30   2007/8/29  20:55  88 10 13.0  119 46 3.3  E   3442     1100
PS70/313   31   2007/8/29  22:45  88 9  9.8   124 49 21.5 E   2892     1100
PS70/315   32   2007/8/30   7:25  88 11 3.0   135 1  14.7 E   1325      628
PS70/316   33   2007/8/30  10:38  88 11 6.6   139 33 35.8 E   1257     1100
PS70/323   34   2007/9/1    2:28  88 6  2.3   154 39 7.3  E   3496      484
           35   2007/9/1    2:33  88 5  59.3  154 41 54.2 E   3496     1100
PS70/325   36   2007/9/1   16:22  88 2  54.4  165 6  49.5 E   3901     1100
PS70/326   37   2007/9/1   21:29  88 2  16.3  169 59 0.5  E   3851     1100
PS70/327   38   2007/9/2    5:45  87 57 7.2   179 59 1.2  E   3484     1100
by Heli.   39   2007/9/2   16:25  86 19 47.5  178 57 47.2 W   3813     1100
PS70/329   40   2007/9/3   13:15  87 44 36.6  162 37 58.4 W   2898     1100
PS70/332   41   2007/9/4    2:05  87 18 37.4  150 8  52.0 W   3081     1100
PS70/334   42   2007/9/4   22:46  86 42 21.1  142 23 14.1 W   2786      227
           43   2007/9/4   22:55  86 42 19.6  142 22 43.1 W   2786     1100
PS70/336   44   2007/9/5    9:04  86 8  22.5  137 29 44.1 W   2172     1100
PS70/337   45   2007/9/5   12:08  85 56 33.5  136 22 16.2 W   1933     1100
by Heli.   46   2007/9/5   13:19  85 42 47.9  132 26 42.5 W   1486     1100
by Heli.   47   2007/9/5   13:38  85 41 29.5  129 57 39.0 W   1414     1100
by Heli.   48   2007/9/5   13:56  85 42 0.0   127 30 11.2 W   1356     1100
by Heli.   49   2007/9/5   14:18  85 42 28.0  125 5  49.1 W   1624     1100
by Heli.   50   2007/9/5   14:39  85 42 13.0  122 35 9.5  W   1401     1100
by Heli.   51   2007/9/5   14:57  85 41 46.0  120 2  34.8 W   1209     1100
by Heli.   52   2007/9/5   15:17  85 41 52.4  117 35 10.6 W   1642     1100
PS70/339   53   2007/9/6   13:37  85 23 19.5  136 17 2.1  W   1858      121
           54   2007/9/6   13:41  85 23 19.5  136 17 2.1  W   1858      563
PS70/341   55   2007/9/7    0:03  84 47 0.1   137 59 14.1 W   1881     1100
by Heli.   56   2007/9/7   13:24  84 17 39.1  139 59 42.0 W   2268     1100
by Heli.   57   2007/9/7   13:45  84 4  57.0  139 55 46.1 W   2375     1100
by Heli.   58   2007/9/7   14:06  83 52 29.8  140 1  6.2  W   2414     1100
by Heli.   59   2007/9/7   14:25  83 40 53.2  139 59 51.8 W   2587     1100
by Heli.   60   2007/9/7   14:46  83 27 41.8  139 59 8.4  W   2779     1100
by Heli.   61   2007/9/7   15:09  83 15 5.5   139 59 24.6 W   2907     1100
by Heli.   62   2007/9/7   15:40  83 3  19.9  139 57 6.1  W   2996     1100
PS70/344   63   2007/9/8    2:16  84 36 36.9  141 40 53.7 W   1991      917
PS70/347   64   2007/9/8   20:35  84 52 32.6  154 8  33.8 W   2216     1100
PS70/348   65   2007/9/9    0:39  84 58 44.4  158 42 12.6 W   1997     1100
PS70/350   66   2007/9/9   15:40  85 22 26.3  167 12 6.7  W   1900     1100
PS70/353   67   2007/9/11  15:06  86 35 43.9  162 12 47.7 E   3848     1100
PS70/354   68   2007/9/11  17:19  86 33 51.2  159 43 5.5  E   3758     1100
PS70/355   69   2007/9/11  19:10  86 31 39.6  157 17 5.8  E   2762     1100
PS70/356   70   2007/9/11  20:28  86 31 18.2  155 30 1.4  E   1622      833
PS70/357   71   2007/9/11  21:38  86 30 45.0  153 44 55.1 E   1288     1100
PS70/359   72   2007/9/12   9:09  86 28 15.8  149 17 44.7 E   1184     1100
PS70/360   73   2007/9/12  10:55  86 26 18.6  146 47 47.8 E   1101     1100
PS70/361   74   2007/9/12  12:43  86 25 36.2  144 4  25.0 E    871      871
PS70/366   75   2007/9/15   9:50  86 3  31.5  119 18 28.0 E   4245     1100
PS70/367   76   2007/9/15  17:10  85 39 50.0  112 21 32.1 E   4060     1100
PS70/368   77   2007/9/16   1:09  85 10 59.0  106 59 36.8 E   3953     1100
PS70/369   78   2007/9/16   3:36  85 0  59.0  105 26 5.7  E   3985     1100
PS70/370   79   2007/9/16   8:15  84 50 36.3  103 59 30.9 E   3901      772
PS70/387   80   2007/9/21   5:21  78 38 13.1  124 35 51.2 E   2731      871
PS70/388   81   2007/9/21   6:19  78 29 58.0  124 35 31.4 E   2570     1100
PS70/390   82   2007/9/21  15:15  78 14 19.1  124 22 31.8 E   2339     1100
PS70/392   83   2007/9/21  18:28  78 1  46.8  124 6  52.3 E   2143      517
PS70/393   84   2007/9/21  19:08  77 56 36.3  124 1  37.0 E   2079     1100
PS70/395   85   2007/9/21  21:46  77 46 30.1  123 47 51.6 E   1871     1100
PS70/396   86   2007/9/21  22:10  77 42 35.0  123 42 21.1 E   1801     1100
PS70/398   87   2007/9/22   4:53  77 33 17.2  123 34 52.2 E   1516     1100
PS70/399   88   2007/9/22   5:28  77 28 48.3  123 29 36.8 E   1335     1100
PS70/400   89   2007/9/22  12:19  77 22 14.6  123 25 13.7 E   1034     1034



Fig. 4.2.1: XCTD observation on the ice by helicopter operation at XCTD no. 
            18 on August 17

Fig. 4.2.2: XCTD observation from the Polarstern at PS70/268 (XCTD no. 16) on 
            August 14

Fig. 4.2.3: Map of XCTD observation site (white) with CTD location (black) on 
            the IBCAO chart

Fig. 4.2.4: (Left) Potential temperature-salinity diagram and (Right) 
            potential temperature profiles of all 89 XCTD data

Fig. 4.2.5: Results of XCTD no. 46 to 62 which were conducted by helicopter 
            operation. (Left, top) Potential temperature section. (Left, 
            bottom) Bottom topography over the XCTD station. (Right, top) 
            Potential temperature-salinity diagram between 31 and 34.5 psu in 
            salinity range. (Right, top) Potential temperature-salinity 
            diagram between 34.65 and 34.95 psu in salinity range.



4.3  Deployment of ice-tethered buoys

     Takashi Kikuchi(2), Sebastian   (1)Alfred-Wegener-lnstitut
     Mechler(3), Sergey Fisarev(4),  (2)Japan Agency for Marine Earth Science 
     Benjamin Rabe(1)                   and Technology
                                     (3)Optimare
                                     (4)P.P. Shirshov Institute of Oceanology


In order to obtain year-round measurements of temperature, salinity, velocity 
and under-ice turbulence, ice-tethered platforms (ITPs) with various 
instrumentation were deployed. They consist of a sub-ice sensor system that 
is connected by a cable to a surface unit that transmits the data to shore 
via satellite. Since they drift with the host ice floe they have the 
potential to provide observations over a substantial region of the Arctic 
Ocean.

  • 3 ITPs (Ice-Tethered Profilers) equipped with Seabird CTDs that will 
      sample temperature and salinity profiles once per day between the 
      surface and 800 m water depth,
  • 1 ITAC (Ice-tethered Acoustic Current profiler) consisting of a RDI ADCP 
      (75 kHz, Long Ranger) that measures the velocity profile of the upper 
      500 m once per day,
  • 1 OFB (Ocean Flux Buoy, from Tim Stanton, Naval Postgraduate School) that 
      measures turbulent fluxes of heat and salt immediately below the ice.

These platforms contribute to the "International Arctic Ocean Observation 
System" (iAOOS) that aims at a persistent observation network.

In total, seven different types of ice buoy systems were deployed. Here we 
report on the ocean measurement systems and refer to chapter 3 for 
ice-related measurements.

Three Ice Tethered Profilers (ITP) measure thrice daily 
temperature/salinity/depth profiles with 1 m vertical resolution between 8 
and 760 m using a profiling CTD unit (Seabird Electronics, Inc. model 41CP) 
on a wire tether and an inductive modem to communicate the data to a surface 
unit (SU). The ITP SU records GPS position and relays all data via an Iridium 
satellite modem connection to a server at Woods Whole Oceanographic 
Institution (WHOI) in Woods Hole (Massachusetts, USA). The ITPs are 
manufactured by WHOI with a profiler from McLane Research Laboratories 
(Falmouth, Massachusetts, USA).

A system similar to the ITP, a Polar Ocean Profiling System (POPS) 
manufactured by MetOcean Data Systems (Dartmouth, Nova Scotia, Canada) 
financed by the Japan Agency for Marine-Earth Science and Technology 
(JAMSTEC, Tokyo, Japan) was deployed, but unfortunately had to be recovered 
after on-site tests failed.

One Arctic Ocean Flux Buoy (AOFB) from Tim Stanton at the Naval Postgraduate 
School in Monterey (California, USA), equipped with one set of temperature, 
salinity and depth sensors and an FSI current meter, measures small scale 
fluctuations in the surface layer a few meters below the ice. In addition, a 
RDI 300 kHz Acoustic Doppler Current Profiler measures velocity profiles of 
the top 80 m of the water column. Similar to the ITP, an SU relays the 
information to the Naval Postgraduate School via Iridium. A wind generator 
provides additional power to extend the operating life of the buoy beyond the 
capacity of the buoy batteries.

The prototype of a buoy newly developed by Optimare Sensorsysteme AG 
(Bremerhaven, Germany) in collaboration with the Alfred Wegener Institute in 
Bremerhaven (Germany) was deployed for the first time: An Ice Tethered 
Acoustic Current profiler (ITAC), measuring ocean current velocity profiles 
from 2 m under the ice to a depth of around 500 m, incorporates an ADCP 
mounted (initially) 50 cm under the ice floe. The ADCP is rigidly connected 
via a stainless steel pole with a wooden beam on the surface. A cable 
provides the electrical connection to a SU with a GPS receiver and an Iridium 
modem. To allow the recording of the ADCP orientation even in regions of low 
horizontal magnetic field strength, a 2nd GPS is positioned about 98 m away 
in line with the wooden beam and the ITAC SU. Data are relayed daily via the 
Iridium Short Burst Data (SBD) message service to an e-mail address at 
Optimare; all ITAC SU data, which also include temperature and horizontal 
tilt measurements, are also relayed via the ARGOS system once a week. The 
communication is bi-directional and also allows setting of data sampling 
parameters via SBD messages, both for the ADCP and the ITAC SU (e.g. GPS 
sampling rate).

The buoys were deployed indifferent combinations along the cruise track (Fig. 
4.3.1). The weak and thin sea ice cover posed a severe challenge to find 
suitable ice floes. Generally, if an appropriate floe could not be easily 
found from the bridge of Polarstern, one or more survey flights were 
conducted by helicopter, landing on potentially usable floes and drilling a 
few holes with a 2' electrically powered ice auger to test ice thickness. 
Once a floe was identified, an initial survey of the ice thickness was 
carried out by the sea ice physics group to provide guidance in finding a 
suitable deployment site. The survey was performed with the EM31 canoe in 
conjunction with 2' drilling and visual observation of ice surface features, 
such as melt-ponds and ridges. For the deployments, ice holes were drilled, 
and the system was lowered using tripods with chain hoists. Since the ITAC 
required a hole of 60 cm the mechanical drilling required considerable effort 
and time. For future deployments, a hot-water system, similar to that 
employed by WHOI, should be used. Maps of three different buoy deployment 
sites, where ITPs were deployed, show the distribution of sea ice thickness 
in relation to the buoy locations (Fig. 4.3.2). In particular the large 
'Super Station' presented a compromise between availability of suitable ice 
floes as well as sufficiently thick but uniform areas and the large number of 
buoys that needed to be deployed with certain minimum spacing. The required 
thickness of around 2 to 2.5 m and expected long-term stability of the ice 
could only be found in areas close to some old ridge systems as the remainder 
of the floe was largely covered by refrozen melt-ponds.


Fig. 4.3.1: Cruise track with ocean ice-buoy deployment locations: The 
            squares represent ITP#16 and ITP#15, each deployed with a surface 
            meteorological buoy; the circle marks the 'Super Buoy Station', 
            including ITP#12, ITAC, AOFB and an Ice-Mass-balance Buoy 
            (IMB) and a webcam. Contours represent ocean bottom topography 
            from the IBCAO dataset (in m).

Fig. 4.3.2: Ice-floe survey map of 'Super Station' with buoy locations. 
            Shades represent EM31 sea ice thickness in m (courtesy of the AWI 
            Sea ice Physics group).


Due to the real-time transmission and processing of the buoy data we can 
present first results of the some of the deployed systems. The ITP#12 
surveyed, as part of the 'Super Buoy Station', around the Lomonosov Ridge 
between 86° 40' to 87° 24' N (see Fig. 4.3.3), on the edge between the 
Amundsen and Makarov Basins. The temperature section against time (Fig. 
4.3.4) shows occasional drops in the maximum in the Atlantic water layer, 
around 300 dbar. This is representative of the margin between the two basins 
and is in agreement with previous findings that the Atlantic water layer is 
generally colder in the Makarov Basin than in the Amundsen Basin, e.g. during 
the Oden 91 expedition (Anderson et al., 1994). These intrusions are also 
evident around 300 dbar in both the salinity profiles (Fig. 4.3.5) and the 
temperature profiles (Fig. 4.3.6).


References

Anderson, L.G., Björk, C., Holby, 0., Jones, E.P., Kattner, C., Koltermann, 
    K.-P., Liljeblad, B., Lindegren, R., Rudels, B. and Swift, J.H. 1994. 
    Water masses and circulation in the Eurasian Basin: Results from the Oden 
    91 Expedition, J. Geophys. Res., 99, p. 3273-3283.


Fig. 4.3.3: Drift track of ITP#12 ('Super Station' ice floe) during September 
            2007

Fig. 4.3.4: ITP#12 potential temperature, referenced to 0 dbar vs. pressure 
            and time for the drift track shown in figure 4.3.3

Fig. 4.3.5: Selected profiles of ITP#12 in the vicinity of the Lomonosov 
            Ridge: Salinity.

Fig. 4.3.6: Selected profiles of ITP#12 in the vicinity of the Lomonosov 
            Ridge: Potential temperature, referenced to 0 dbar




5.  GEOTRACES

Background and general objectives

The availability of trace elements and isotopes (TEI) in the ocean controls 
and limits marine productivity, their present distributions reflect ocean 
circulation and they can be used to identify the sources and sinks of metals 
and other matter. TEI records in sediments allow the reconstruction of past 
climatic conditions and of past changes in ocean circulation. Our knowledge 
of tracer distribution is still strongly based on the successful GEOSECS 
expeditions in the 1970s. Since then, there have been major improvements in 
sampling, analytical, and modelling techniques. The international programme 
GEOTRACES has been initiated to make use of these developments and make major 
advances in our understanding of TEI cycling in the coming decade. One major 
aim of international GEOTRACES (http://www.geotraces.org) is:

"To determine global ocean distributions of selected trace elements and 
isotopes, including their concentration, chemical speciation, and physical 
form, and to evaluate the sources, sinks, and internal cycling of these 
species to characterise more completely the physical, chemical and biological 
processes regulating their distributions".

The International PolarYear(IPY) is an excellent opportunity to study Trace 
Elements and Isotopes in theArcticand Antarctic Oceans. An international 
suite of vertical sections in the polar oceans is integrated in the IPY 
project No. 35 (http://www.ipy.org/development/eoi/proposal-details.php?id=35) 
entitled: "International Polar Year GEOTRACES: An international study of the 
biogeochemical cycles of Trace Elements and Isotopes in the Arctic and Southern 
Oceans". ARK-XXII/2 was the first expedition carried out in the context of this 
IPY-GEOTRACES. A second Polarstern expedition is scheduled in the Southern Ocean 
(ANT-XXIV/3; 2008).

The present expedition allowed the parallel sampling of a wide spectrum of 
tracers. Most of the key parameters mentioned in the GEOTRACES science plan 
are covered at least at some large stations. Sampling included tracers for 
river and shelf inputs, plankton production and particle rain, anthropogenic 
inputs and water mass circulation.

Intercomparison/Intercalibration

Intercalibration activities are ongoing for trace metals (SAFe) and nutrients 
(see report below). Special GEOTRACES intercalibration expeditions devoted to 
other parameters are scheduled for the summer of 2008. In order to allow a 
proper intercomparison of data acquired during our expedition with data 
produced after these intercalibration expeditions, we have collected sets of 
samples for some groups of tracers to be distributed among participating 
laboratories. Details of the sampling for Th/Pa, REE and Barium are given in 
the respective subprojects.


Data Management

All data of Isotopes and Trace Metals will be reported into the worldwide 
database of the GEOTRACES programme.


Work at sea

It was our objective to make parallel analyses of a wide spectrum of tracers 
at the same stations. This strategy is essential if we want to interpret the 
signals of the various tracers in a common context. Backbone of the tracer 
studies was the sampling with the new ultra clean CTD/Rosette of NIOZ (see 
below), with additional casts of the conventional rosette for less 
contamination-prone parameters. While some tracers could be determined at all 
sampled stations and depths, others could only be analysed at selected 
stations because of the large water volumes required or analytical 
constraints. The resulting sampling programme is presented in Table 5.1.


Tab. 5.1: GEOTRACES: Sampling list of trace elements, isotopes and supporting 
          parameters             

                      s                                              b
                     l                  C                           P
                    a                  I                           0                   I
                   t      s           D                                               9
                  e      t   P   O   C                           1
                 M  k   n   M   8   3                           2                   2       N       i   a
                   l   e   S   1   1           e               /                   1       O       S   -
              e   A   i   D   a   a           B   h       h   o   a   a   a   c   s       D       B   r
             c   /   r       t   t           /   T       T   P   R   R   R   A   C   c           /   o
            a   C   t   4   l   l       E   d   4   P   /   0   4   8   6   7   /   T   C       C   l   P
           r   I   u   H   e   e   a   E   N   3   S   a   1   2   2   2   2   u   9   O       O   h   E
Station   T   D   N   C   d   d   B   R   e   2   I   P   2   2   2   2   2   P   9   D   #   P   C   T
         --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- ---
                      2,
          1)      2)  3)  2)  2)              4)  5)  6)      7)  8)  9)         10) 11)
         --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- ---
PS70/228  X   X   X   X   X       X   X       X
PS70/229          X                                               S
PS70/232          X                                                           S       X
PS70/236  X   X   X   X   X       X   X       X      PD   X   S   S           X       X       X   X   X
PS70/237  X   X   X   X   X       X   X       X           X   S   S           S       X       X   X   X
PS70/239  X   X   X       X       X   X       X   X  PD   X   S   X           S       X       X   X   X
PS70/240          X   X   X                                                           X
PS70/241          X                                               S
PS70/243          X           X               X                                       X           X
PS70/246  X   X   X   X       X   X   X       X                   S
PS70/248          X                                       S       S                           S       S
PS70/255  X   X   X       X       X   X       X   X   T   X       X   X       S       X       X   X   X
PS70/257          X   X   X                   X           S   S   S   S               X   B   S       S
PS70/258  X   X   X               X   X                   S                                   S       S
PS70/260  X   X   X   X   X       X   X       X   X  PD   X       X   S   X   X               X   X   X
PS70/261  X   X   X   X   X       X   X       X           S   S   S   S                       S       S
PS70/263          X   X   X   X                           S   S                       X       S       S
PS70/264          X   X   X                   X               S   S   S       S       X
PS70/265          X                                       S                                   S       S
PS70/266  X   X   X   X   X   X   X   X       X   X   T   X   S   X   S   X   S   X   X   F   X   X   X
PS70/267          X       X   X                                                   X
PS70/268  X   X   X   X   X                   X           S   S                   X           S       S
PS70/271          X   X   X       S   S   B   X           S   S   S           X       X       S       S
PS70/272  X   X   X   X   X       S   S   X               S   X               S               S       S
PS70/273          X       X   X                                               X       X
PS70/274          X   X   X   X                           S   S                       X       S       S
PS70/276  X   X   X       X       X   X   X   X   X  PD   X   S   X   S       X   X   X       X   X   X
PS70/277          X       X   X                               X
PS70/279  X   X   X   X   X   X   X   X   X   X   X  PD   X   S   X   S       X   X   X   F   X   X   X
PS70/280          X   X   X                                                           X
PS70/283          X       X   X                                                       X
PS70/284          X   X   X   X                               S
PS70/285  X   X   X   X   X       S   S       X           S   S                       X   F   S       S
PS70/286          X       X   X                                                   X
PS70/288          X       X   X                               S
PS70/289          X       X                                                       X
PS70/290          X       X                               S   S                   X   X       S       S
PS70/291  X   X   X   X   X       S   S                                       S
PS70/292          X       X                                       S   S               X
PS70/294          X       B                                   S                       X
PS70/295  X   X   X   X   X                   X           S                           X   F   S       S
PS70/298          X                                                           S       X   F
PS70/299  X   X   X   X   X       X   X                   S   S                               S       S
PS70/301  X   X   X   X           S   S       X   X   T   X   S   X   S   X   S       X   F   X   X   X
PS70/302  X   X   X       X                                                           X   F
PS70/303          X                                       S   S                               S       S
PS70/305          X                                           S
PS70/306  X   X   X   X           S   S       X           S       S           S       X   F   S       S
PS70/307          X       X                                                           X
PS70/308          X   X                                                       S       X   F
PS70/309  X   X   X   X   X       X   X       X   X  PD   X   S   X   S   X   S       X   FB  X   X   X
PS70/310  X   X   X   X                                                       X
PS70/312          X       X                               S   S               S       X   F   S       S
PS70/314          X   X                                                       S       X   F
PS70/316  X   X   X       X       X   X       X               S               S       X   F
PS70/319  X   X   X                                                                   X   FB
PS70/320          X                                       S                           X   F   S       S
PS70/321          X   X                                                       X
PS70/322          X       X                                   S   S   S               X   F
PS70/324          X                                                           S       X   F
PS70/326  X   X   X   X                                       S               S       X   F
PS70/328  X   X   X   X   X       X  XI       X   X  PDI  X   S   X   S   X   X   X   X   FB  X   X   X
PS70/331          X   X   X                                                           X   F
PS70/333  X   X   X       X                   X           S   S   S           S               S       S
PS70/335          X   X   X                                   S               S       X   F 
PS70/338  X   X   X   X   X       X   X       X   X       X   S   X   S       S       X   FB  X   X   X
PS70/340          X   X                                       S                       X   F       X
PS70/342  X   X   X   X   X       X   X       X       T   X   X   S   S       X       X   FB  X   X   X
PS70/345          X   X   X                                                   S       X   F
PS70/346          X   X                                                               X
PS70/349  X   X   X   X   X       X   X       X           S   S               S       X   FB  S       S
PS70/351          X       X                                                   S       X
PS70/352  X   X   X   X   X       I           X           S   S   S   S      12)  X   X   F   X   X   S
PS70/358          X   X   X                   X   X       S   S   X           S       X   F   X   X   S
PS70/362          X   X                                                               X   F 
PS70/363  X   X   X       X                           T   S   S   S   S   X   S               X   X   S
PS70/365          X                                                           S       X   F
PS70/371  X   X   X       X       X   X       X           S   S               S       X       S       S
PS70/372  X   X   X                                                           S
PS70/373  X   X   X       X                               S                           X   F   S       S
PS70/375          X                                                           S       X
PS70/377          X       X                               S   S               S               S       S
PS70/379  X   X   X                           X           S                   S               S       S
PS70/381          X                                                           S       X
PS70/382  X   X   X       X       X   X                   S                   S       X       S       S
PS70/383          X       X                   X           S   S               S       X   F   S       S
PS70/384          X                                       S                           X   F   S       S
PS70/385  X   X   X       X       X   X       X       T   X   S   X   S       S       X   B   X   X   X
PS70/387          X       X                               S                       X   X   F   S       S
PS70/389  X   X   X       X       X   X                       S               S
PS70/391          X       X                               S                       X   X   F   S       S
PS70/394          X       X                                                   S
PS70/397          X       X                                                       X
PS70/400  X   X   X       X   X   X   X   X   X   X   PD  X   S   X   X       X   X   X   F   X   X   X
PS70/401          X       X   X                                                   X
PS70/403          X   X   X                   X                                       X
PS70/404          X       X   X                           S                   S               S       S
PS70/405          X   X   X                                                           X   F
PS70/407  X   X   X   X   X       X   X   X   X   X       X   X   X   S       S       X   F   X   X   X
PS70/409          X   X   X                               S   X               S       X   B   S       S
PS70/411  X   X   X   X   X   X   X   X   X   X   X   PD  X   X   X   S       X       X   FB      X   X


 1) for details see table GEOTRACES-2                 X  depth profile
 2) more stations were sampled in-between             S  surface water only
 3) upper 200 m of water column                       B  bottom sample
 4) samples were collected from                       F  fluorescence maximum
    0, 25, 50, 75, 100, 150 and 200 m.                I  Intercalibration
 5) ISP: in-situ pumps for Radium and 
         size-fractionated POC/234Th
         down to 1,000 m water depth
 6) PD   particulate and dissolved 230Th and 231Pa
    T    total 230Th and 231Pa
 7) short-lived Ra isotopes with RaDeCC technique
 8) adsorption on MnO2 cartridges
 9) BaSO4 precipitation
10) Sampling in the Atlantic Layer
11) including optics/fluorescence spectra
12) Pu surface sample; Cs and 129I profile





5.1  A- trace elements

     Karel Bakker, Lorendz Boom, Maarten Kiunder, Rob Middag, Sven Ober, 
     Charles-Edouard Thuroczy and Patrick Laan
     Royal Netherlands Institute of Sea Research


Objectives

The distribution and biological availability of Fe is strongly controlled by 
its physicalchemical speciation within seawater, where colloids and 
Fe-organic complexes are dominant actors. The external sources of Fe into the 
oceans are either from above (dust) or from below (sediments) and will be 
constrained by Al and Mn for aeolian dust input and sedimentary redox cycling 
sources, respectively. The Fe enhances phytoplankton growth, which in turn 
strongly controls the biological pump for uptake of CO2 from the atmosphere 
into polar oceans. The increasing CO2 in polar ocean waters may affect 
phytoplankton ecophysiology, with key link of metal Fe in the overall 
photosynthetic apparatus.



Work at Sea

The Ultra Clean CTD system (UCC)
Sven Ober, Patrick Laan, Lorendz Boom
Royal Netherlands Institute of Sea Research

During the cruise a special CTD-system was used to sample for trace-elements 
and isotopes. This CTD-system consists of 3 major modules: a winch with a 
superaramide CTD-cable, a box-shaped titanium CTD-frame and a clean air 
container that is designed to hold the CTD-frame in order to enable 
subsampling and filtration under clean air conditions. The CTD-frame is made 
of pure titanium and was equipped with a Seabird SBE9+ CTD underwater unit, a 
SBE3 thermometer, a SBE4 conductivity sensor, a SBE5 underwater pump, a SBE43 
DO-sensor, a Chelsea MK-111 fluorometer, a Seapoint OBS and a special 
sampling-system. This sampling-system consists of a Multivalve hydraulic 
multiplexer and 24 GoFlo sampling bottles each with its own hydraulic release 
unit. (De Baar et al., 2007; Ober et al., 2002).

In addition to the above mentioned sensors a Dr. Haardt fluorometer type 
BackScat 1 for detecting yellow substance was mounted from station 266, cast 
1. From station 371, cast 2 the Aquatracka fluorometer was dismounted and a 
WetLabs C-Star transmissometer was mounted instead.


Tab. 5.2: Deployment list of Ultra-Clean CTD with sampling list of trace 
          metals

Station     Position      Position     Sample  DFe  DAI  DMn  Li   DIC/ Fe    Nut  Sil
            Lat           Lon          depth                  bra  Alk  lig   rie  ver
                                       [m]                    ry        ands  nts
----------  ------------  -----------  -----  ----  ---  ---  ---- ---  ----  ---  ---
PS70/228-1  75° 0.03'N    34° 0.00'E    100    yes  yes  yes  yes  yes        yes
PS70/236-1  77° 30.05'N   33° 59.20'E   150    yes  yes  yes  yes  yes        yes
PS70/237-1  78° 59.83'N   33° 59.39'E   225    yes  yes  yes  yes  yes        yes
PS70/239-1  80° 59.68'N   33° 59.80'E   175    yes  yes  yes  yes  yes        yes
PS70/246-1  81° 52.27'N   34° 0.68'E   1750    yes  yes  yes  yes  yes        yes
PS70/255-1  82° 30.25'N   33° 57.01'E  2950    yes  yes  yes  yes  yes        yes
PS70/258-1  83° 59.94'N   33° 59.81'E  3990    yes  yes  yes  yes  yes        yes
PS70/260-2  84° 29.38'N   36° 8.29'E   3935    yes  yes  yes  yes  yes        yes
PS70/260-4  84° 29.51'N   36° 6.15'E   3935    yes  yes  yes            yes   yes       charly
                                                                                        cast
PS70/261-1  84° 38.27'N   60° 55.24'E  3700    yes  yes  yes  yes  yes        yes
PS70/266-1  83° 8.09'N    61° 46.17'E  2950    yes  yes  yes  yes  yes        yes
PS70/268-1  82° 48.48'N   60° 48.35'E  1500    yes  yes  yes  yes  yes        yes
PS70/272-l  82° 15.15'N   61° 59.67'E   210    yes  yes  yes  yes  yes        yes
PS70/276-1  82° 5.05'N    68° 57.50'E   650    yes  yes  yes  yes  yes        yes
PS70/279-2  81° 14.68'N   86° 12.49'E   315    yes  yes  yes  yes  yes        yes
PS70/279-6  81° 12.20'N   86° 18.47'E   300    yes  yes  yes  yes  yes        yes
PS70/285-2  82° 8.47'N    86° 20.20'E   680    yes  yes  yes  yes  yes        yes
PS70/291-1  82° 42.71'N   86° 16.33'E  2200    yes  yes  yes  yes  yes        yes
PS70/295-1  83° 16.14'N   86° 18.04'E  3200    yes  yes  yes  yes  yes        yes
PS70/299-1  84° 3.02'N    89° 3.42'E   3550    yes  yes  yes  yes  yes        yes
PS70/301-2  84° 34.30'N   89° 50.89'E  3650    yes  yes  yes  yes  yes  yes   yes
PS70/302-1  84° 53.23'N   90° 5.94'E   3650    yes  yes  yes  yes  yes        yes
PS70/306-1  85° 55.42'N   91° 10.79'E  3700    yes  yes  yes  yes  yes        yes
PS70/309-2  87° 2.71'N   104° 50.31'E  1500    yes  yes  yes  yes  yes  yes   yes       combi
                                                                                        cast
PS70/309-4  87° 1.94'N   104° 51.24'E  4325    yes  yes  yes  yes  yes        yes
PS70/310-1  87° 39.81'N  111° 57.18'E  4250    yes  yes  yes  yes  yes        yes
PS70/316-1  88° 10.75'N  139° 36.22'E  1250    yes  yes  yes  yes  yes        yes
PS70/319-1  88° 40.05'N  153° 42.58'E  2650    yes  yes  yes  yes  yes        yes
PS70/326-1  88° 1.85'N   169° 59.63'E  3900    yes  yes  yes  yes  yes        yes
PS70/328-2  87° 49.60'N  170° 24.44'W  3900    yes  yes  yes  yes  yes        yes
PS70/328-4  87° 49.50'N  170° 21.17'W  2000    yes  yes  yes  yes  yes        yes
PS70/333-1  87° 1.64'N   146° 23.42'W  3200    yes  yes  yes  yes  yes        yes
PS70/338-2  85° 42.20'N  135° 2.09'W   1475    yes  yes  yes  yes  yes        yes
PS70/342-1  84° 29.97'N  138° 24.75'W  2200    yes  yes  yes  yes  yes        yes
PS70/349-1  85° 3.92'N   164° 29.82'W  1950    yes  yes  yes  yes  yes        yes
PS70/352-2  86° 38.52'N  177° 32.69'E  3900    yes  yes  yes  yes  yes        yes
PS70/363-5  86° 27.95'N  134° 55.46'E  3850    yes  yes  yes  yes  yes        yes
PS70/371-2  84° 39.32'N  102° 43.99'E  4050    yes  yes  yes  yes  yes        yes
PS70/372-1  84° 19.84'N  107° 22.44'E  4060    yes  yes  yes  yes  yes        yes
PS70/373-2  84° 11.79'N  108° 56.93'E  4050    yes  yes  yes  yes  yes        yes
PS70/379-1  82° 51.57'N  117° 51.13'E  4352    yes  yes  yes  yes  yes        yes
PS70/382-1  81° 21.45'N  120° 43.15'E  5200    yes  yes  yes  yes  yes        yes
PS70/385-1  79° 20.88'N  124° 20.83'E  3425    yes  yes  yes  yes  yes        yes  yes  
PS70/389-1  78° 21.30'N  124° 31.30'E  2500    yes  yes  yes  yes  yes        yes
PS70/400-1  77° 22.94'N  123° 24.80'E  1091    yes  yes  yes  yes  yes        yes
PS70/407-1  76° 10.83'N  122° 7.86'E     55    yes  yes  yes  yes  yes  yes   yes
PS70/411-1  75° 12.03'N  121° 21.61'E    35    yes  yes  yes  yes  yes        yes


In total 49 casts were carried out with the UCC-system including 2 test casts 
(Table 5.2).

Throughout the whole cruise the system worked very reliably, although some 
small technical problems occurred. The conductivity-sensor had to be 
exchanged for a spare because it appeared to have a slightly shifted 
calibration (although it was calibrated recently) and the OBS had to be 
exchanged, because this sensor apparently did not survive the pressure during 
the deepest cast of the cruise (5,220 m) although this sensor was rated up to 
6,000 m. The Dissolved Oxygen sensor showed erratic values at depths over 
about 2,000 m. This problem was solved by exchanging the cable between the 
sensor and the underwater unit. At steep gradients some salinity-spiking was 
observed. Possible cause is the changed duct of the CTD. The longer tubes 
slow down the flow of the water and therefore the standard timing of the 
sensors is not optimal. This will be become clear during postprocessing of 
the data in the near future. In case the retuning of timing of the sensors 
will not solve the spiking the most probable cause is a disturbed flow near 
the sensors. Another, more free-flow, location for the sensors in the frame 
must be considered.

The hydraulic bottle control system worked perfectly (100%) and the GOFLO 
samplers worked almost perfectly (99% based on nutrient data). Prior to the 
cruise the edges of the holes in the top and bottom closing spheres of each 
Go-FLO sampler were made less sharp and prior to each cast all the spheres 
were sprayed with Teflon spray. These efforts clearly paid off.

Highest priority was given to the sampling of complete vertical profiles 
throughout the complete (4 - 5 km depth) water column at deep water stations 
in the different central Arctic Ocean basins. The sampling depth differed 
from 33 meter as the shallowest station, in the Laptev Sea, up to 5,220 meter 
for the deepest station at the south of the Gakkel Ridge. From these 49 casts 
27 were deeper than 2,000 meter.


References

De Baar, H.J.W., K.R. Timmermans, P. Laan, H.H. De Porto, S. Ober, J.J. Blom, 
    M.C. Bakker, J. Schilling, G. Sarthou, M.G. Smit and M. Klunder (2008) 
    Titan: A new facility for ultraclean sampling of trace elements and 
    isotopes in the deep oceans in the international Geotraces programme, 
    Marine Chemistry, in press.

Ober, S., Groenewegen, R.L., Boekel, H.J., Keijzer, E.J.H., Derksen, J.D.J., 
    Laan, M., 2002. A new way of oceanographic watersampling. Abstract of 
    presentation at Inmartech, 8 October 2002, Yokusuku, Japan. 
    http://www.jamstec.go.jp/jamstec-e/whatsnew/inmartech2002/programme.pdf.



Results

After recovery the complete frame with its 24 samplers was placed inside its 
home laboratory. This is the ultraclean laboratory container NIOZ-7 placed on 
the aft-deck (Arbeitsdeck). The seawater was processed and either filtered, 
over filtration cartridges by pressurizing each sampler with nitrogen gas 
from cylinders, or unfiltered and collected in pre-cleaned bottles for the 
analyses of dissolved Fe, Al, Mn and alkalinity and dissolved inorganic 
carbon and brought to the different analists for analysis.

Besides this, the sampling for a variety of the GEOTRACES community was done 
like Rare Earth Elements, Ba, Uranium and Thorium. Also an extra litre bottle 
was sampled for analyses of various trace elements like Fe, Co, Ni, Cu, Zn, 
Ag, Cd and Pb on ICPMS in the home laboratory.

Titan proved to be extremely robust and easy to handle. The possibility to 
sample 24 depths in combination with the direct CTD data was extremely 
convenient and improved the trace metal data quality compared to the 
traditional way of trace metal sampling. With this system we are able to 
sample every small anomaly present. For the trace metals extreme caution in 
sample handling was done. All sample bottles were rinsed 5 times with a small 
amount of the sample. For the determination of the dissolved fractions the 
seawater was filtered using a Sartobran 300 cellulose acetate filter from 
Sartorius. This filter contains a 0.45 µm front filter and a 0.2 µm end 
filter. For this filtering we used pressurized N2 gas from a bottle. An 
overpressure of less then 1 bar was sufficient.

After sampling the samples were double bagged to avoid any air borne 
contamination and transported to another clean room container where the 
samples were acidified to pH 1.8 with Seastar 12M baseline grade hydrochloric 
acid. An overall success rate of 99% was achieved for the GoFlo (General 
Oceanics) bottles based on the nutrient data.

Three deep stations were sampled for ID ICPMS for Fe for inter-calibration 
purposes of the at sea method.


Subproject Al: Dissolved iron
               Maarten Klunder
               Royal Netherlands Institute of Sea Research

Dissolved iron was measured directly on board by Flow Injection Analysis 
(FIA) after De Jong et al. 1998 in a cleanroom container. In a continuous FIA 
system the acidified pH 1.8, filtered (0.2 µm) seawater is buffered to pH 
4.0. The iron is concentrated on a column which contains the column material 
aminodiacetid acid (IDA). This material binds only transition metals and not 
the interfering salts. After washing off the column with ultra pure water 
(MQ) the column is eluted with diluted acid. After mixing with luminol, 
peroxide and ammonium the oxidation of luminal with peroxide is catalysed by 
iron and a blue light is produced and detected with a photon counter. The 
amount of iron is calculated using a standard calibration line, where a known 
amount of iron is added to low iron containing seawater. Using this 
calibration line a number of counts per nM iron is obtained.

All 47 stations and corresponding depths have been analyzed on board. The 
values of DFe measured varied from 0.18 nM to 10.4 nM. The standard deviation 
varied between 0% and 7% (exceptional), but was generally lower than 5%. 
The standard deviation of the values is determined of a duplicate measurement 
of the same sample bottle. To correct for contamination during the process or 
in the sample bottle a duplicate sample was taken of every station depth. The 
daily consistency of the system was verified using a drift standard. 
Regularly a certified SAFe standard for the long term consistency and 
absolute accuracy was measured. Our average value of the certified SAFe (D2) 
standard was 0.93 (+1- 0.07) nM (n=24), well within range of the 0.91 (+1- 
0.17) nM certified value (Johnson et al., 2007).

Next to the 47 stations also the amount of dissolved iron in the 1,000 kDa 
filtered fraction was measured for five casts. The corresponding 0.2 µm 
filtered fraction of the same cast was also measured. The 1,000 kDa filtered 
fraction generally contained a lower amount of dissolved iron.


Preliminary results

The preliminary data shows that the values for dissolved iron in the Nansen 
basin are according to the ranges found for the North Atlantic Deep Waters, 
0.6 - 0.7 nM. (Martin et al, 1993). In the Amundsen and Makarov Basins the 
values appear to decrease to ranges of 0.4 - 0.6 nM. The stations west of the 
Gakkel Ridge appear to have the common Fe-distribution in the surface waters, 
with a minimum value at the chlorophyll maximum depth (varying from 100 to 25 
meter). After passing the Gakkel Ridge an increase in the surface waters (0 - 
100 m) was found, of up to 3 nM at maximum. This is expected to be caused by 
river input. On the northernmost station of the Gakkel Ridge an increase of 
dissolved iron was found at around a depth of 2,500 -3,000 metres, of which a 
trace was to be seen into the Amundsen Basin.


Fig. 5.1: Depth profile of dissolved iron in the Amundsen Basin



References

Johnson et al., 2007. Developing standards for dissolved iron in Seawater. 
    Eos, Vol 88, n. 11.

Martin J.H., Fitzwater, S.E., Gordon, R.M., Hunter, C.N., and Tanner, S.J. 
    (1993), Iron, primary production, and carbon- nitrogen flux studies 
    during the JGOFS North Atlantic Bloom Experiment. Deep sea Research II. 
    40, 115-134.

De Jong, J.T.M, den Das, J., Bathmann, U., Stoll, M. H.C., Kattner, G., 
    Nolting, R.F., and de Baar, H.J.W. (1998). Dissolved iron at subnanomolar 
    levels in the Southern Ocean as determined by shipboard analysis. 
    Analytica Chimica Acta, 377, 113-124.




Subproject A2: Physical and chemical speciation of dissolved iron
               Charles-Edouard Thuroczy 
               Royal Netherlands Institute of Sea Research

To understand the distribution of the iron in the seawater the concentration 
of natural tigands binding iron and their strength (conditional stability 
constant) are measured after a size fractionation of the seawater: 3 classes 
of size are studied here: unfiltered water, 0.2 µm filtrated water and < 
1,000 kDa ultra-filtrated water.

Four deep stations have been sampled in order to characterize and show any 
difference between 3 basins in the Arctic Ocean: Nansen basins (2,950 m and 
3,935 m), Amundsen Basin (4,500 m) and Makarov Basin (3,900 m). 3 shallow 
stations have also been sampled to show the influence of the river inputs on 
these basins: Barents Sea (175 m), Kara Sea (310 m) and Laptev Sea (55 m). 
Four of these stations will be used later on to study the kinetic exchange 
between the different forms of iron.

In a clean-room container the filtered seawater is ultra-filtrated as to 
isolate the colloids size class of dissolved iron (< 1,000 kDa). The 
ultrafiltrate comprising the 'truly dissolved' fraction, as well as the 
dissolved organic complexed fraction were analysed for Fe, Mn and Al.

The natural ligand concentration is measured by doing a complexing ligand 
titration with addition of iron (between 0 to 8 nM of Fe added). The 
competing ligand 'TAC' is used (2-(2-Thiazolylazo)-p-cresol) and the complex 
(TAC)2-Fe is directly measured by cathodic stripping voltammetry (CSV). The 
electrical signal recorded with this method (nA) is converted as a 
concentration (nM), then the ligand concentration is calculated by knowing 
the dissolved iron concentration:

                  [Fe-l] = [Fe(added)] + [dFe] - [(TAC)2-Fe].





Subproject A3: Dissolved Al and Mn as source tracers for iron
               Rob Middag
               Royal Netherlands Institute of Sea Research

Dissolved Al and dissolved Mn were measured directly using shipboard FIA 
measurements. In a continuous FIA system, the acidified pH 1.8, filtered (0.2 
µm) seawater is buffered to pH 4.8 and 8.5 for Al and Mn, respectively. The 
metals are concentrated on a column which contains the column material 
aminodiacetid acid (IDA). This material binds only transition metals and not 
the interfering salts. After washing off the column with ultra pure water 
(MQ) the column is eluted with diluted acid.

The Al is determined using lumogallion after Resing et al. 1994. Lumogallion 
is a fluorometric agent and reacts with aluminium. The change in the 
fluorescence detected by a fluorometer is used as a measure for the dissolved 
Al concentration.

In order to verify the consistency of the analysis, every day a sample was 
measured from a 25 litre tank that was filled in the beginning of the cruise. 
The average value was 4.98 nM with a standard deviation of 3.16% (0.157 nM). 
Also a duplicate sample was taken every cast and this sample was analysed 
with the samples of the next cast to further check for inter daily variation. 
Since the sample was from a different station every time one cannot calculate 
a standard deviation like for the samples from the 25 litre tank. However, 
one can calculate the percentage deviation for every duplicate and calculate 
the average from the absolute deviation percentages. This was 3.57%.

The Mn is detected using the chemoluminescence method of Doi et al. 2004. The 
oxidation of luminol by hydrogen peroxide produces a blue light. This 
oxidation reaction is catalyzed by manganese and the increase in the 
production of blue light is detected by a photon counter and used as a 
measure for the dissolved Mn concentration.

Also for Mn similar consistency checks as for Al have been performed. The 
average value for the 25 litre tank was 0.998 nM with a standard deviation of 
2.19% (0.022 nM). The average absolute deviation percentage between the 
duplicates was 2.84%. Furthermore seawater was analysed that was collected on 
a NIOZ cruise at two depths in the Atlantic Ocean. One depth per day was 
analysed, so a specific depth was analysed every other day. There was some 
variability between the different bottles of the same depth, but the 
deviation within a bottle was similar to that of the 25 litre tank.

All 47 stations and corresponding depths have been analyzed on board. The 
daily consistency of the system was verified using a so-called drift standard 
and regularly a certified SAFe standard was run for long term consistency and 
absolute accuracy.


Preliminary results

The preliminary data shows that the values in the surface for dissolved 
aluminium in all the basins are extremely low and increasing with depth. This 
indicates that no dust input occurred throughout the entire cruise.

For dissolved manganese the surface values were relatively high most likely 
caused by input from the shelf and/or by rivers. The values in the deep 
waters were much lower than expected.

On the northern most station of the Gakkel Ridge an increase of dissolved 
manganese was found at around a depth of 2,500 - 3,000 m, of which a trace 
was to be seen into the Amundsen Basin. For dissolved aluminium no increase 
was found at these depths.


Fig. 5.2: Depth profiles of dissolved manganese and aluminium in the Amundsen 
          Basin



References

Doi, T., Obata, H., Maruo, M., 2004. Shipboard analysis of picomolar levels 
    of manganese in seawater by chelating resin concentration and 
    chemiluminescence detection. Anal Bioanal Chem 37, 1288-1293.

Resing, J., Measures, C.I., 1994. Fluorimetric determination of Al in 
    seawater by FIA with in-line preconcentration. Anal. Chem. 66, 4105-4111.




Subproject A4: Alkalinity and Dissolved Inorganic Carbon
               Sven Ober 
               Royal Netherlands Institute of Sea Research


For unravelling the geochemical cycles of the ocean it is necessary to 
measure the CO2-system. Therefore high-precision measurements were made of 
the total dissolved inorganic carbon (DIC) content and total alkalinity (TA). 
Alk/DIC-samples were drawn from the GoFlo-samplers of the UCC-system prior to 
the sub-sampling for the other parameters.

Analysis of DIC was performed using the "coulometric method" (Johnson, 1993; 
Dickson, 1993). The Total Alkalinity (TA) analysis was performed with the 
'standard' titration, using curve fits along modified Gran plots (Gran, 1952; 
Bradshaw 1981; Dickson, 1993). Both analyses were performed using a single 
integrated system: the VINDTA system (Versatile Instrument for Determination 
of Titration Alkalinity, MARIANDA: Marine Analytics and Data, Kiel, Germany), 
which also controlled the coulometer. Analyses accuracy was assured through 
the use of certified reference material (CRM, supplied by Dr. A. Dickson, 
Scripps Institute of Oceanography).

For the coulometric determination of DIC (a slight adaptation of the method 
described by Johnson 1993), an accurately known amount of sample (~2O ml) is 
dispensed with an automated, thermostated pipette into a stripper vessel. The 
sample is acidified here, converting all carbonate and bicarbonate species 
into CO2(aq). The evolving CO2 is rapidly removed from the sample by using N2 
as a carrier gas. The CO2enriched N2 stream is led through the solution in 
the coulometric cell, which absorbs the CO2 and becomes more transparent. The 
coulometer subsequently electrically titrates the solution back to its 
original opacity. The required amount of charge is a direct and linear 
measure of the amount of CO2 absorbed. With knowledge of the sample's volume 
and density, the concentration of DIC (the sum of CO2-species in the sample) 
is easily calculated. Total alkalinity is mathematically derived from a 
fourth-order curve fit along a modified Gran plot of electrode potential 
versus volume of acid added to an accurately known amount of sample (~1OO 
ml), dispensed with an automated thermostated pipette. Titration is performed 
in a thermostated cell. Samples are brought to the calibration temperature of 
the pipettes.

Before samples were analyzed the VINDTA was stabilized by running 4 to 6 
dummy samples. As soon as the VINDTA was stable a CRM was applied (in 
duplicate). After a batch of samples was measured a second CRM was applied to 
check for drift of the VINDTA. In cases the batch was big an extra CRM was 
applied after about half of the batch in order to obtain an extra reference 
point. In most cases the VINDTA did not drift significantly. A final 
estimation of the data quality has to be done after a careful re-evaluation 
of all the measurements at NIOZ.



References

Bradshaw, Alvin L.; Brewer, Peter G.; Shafer, Deborah K.; Williams, Robert 
    T.;1981; Measurements of total carbon dioxide and alkalinity by 
    potentiometric titration in the GEOSECS programme; Earth and Planetary 
    Science Letters, Volume 55, Issue 1, p.99-115.

DOE (1994) Handbook of methods for the analysis of the various parameters of 
    the carbon dioxide system in sea water. Version 2, A. G. Dickson & C. 
    Goyet, eds. ORNL/CD IAC-74.

Gran, G. (1952) Determination of the equivalence point in potentiometric 
    titrations, Analyst, 77, 661.

Johnson K. M. (1); Wills K. D. (1); Butler D. B.; Johnson W. K. ; Wong C. S. 
    (1993) Coulometric total carbon dioxide analysis for marine studies: 
    maximizing the performance of an automated gas extraction system and 
    coulometric detector Marine Chemistry, 44, 167-187.




Subproject A5: Nutrients
               Karel Bakker
               Royal Netherlands Institute of Sea Research


Equipment and methods

Nutrients were analysed in a thermostated laboratory container with a 
Technicon TRAACS 800, continuous flow auto-analyser. The sample rate was set 
at 60 samples per hour, measuring about 4,500 samples during the cruise. 
Measurements were made simultaneously on four channels: phosphate, silicate, 
nitrate and nitrite together, and nitrite separately. All measurements were 
calibrated with standards diluted in low nutrient seawater LNSW, and LNSW was 
used as wash water between the samples.

The colorimetric methods used are as follows

Phosphate: Ortho-phosphate is measured by formation of a blue reduced 
Molybdophosphate-complex at pH 0.9-1.1. Potassium Anti monyltartrate used as 
the catalyst and ascorbic acid as a reducing agent. The absorbency is 
measured at 880 nm. (Murphy, J. & Riley, 1962).

Silicate: Measured as a blue reduced Silicomolybdenium-complex at 800 nm. 
Ascorbic acid is used as reducing agent and oxalic acid is used to prevent 
interference of phosphate. (Strickland and Parsons, 1972).

Nitrite: Diazotation of nitrite with sulfanylamide and N-(1-naphtyl)-ethylene 
diammonium dichloride to form a pink dye measured at 550 nm.

Nitrate and Nitrite (here called Nox): Nitrate is first reduced in a 
copperized cadmiumcoil using imidazole as buffer and is then measured as 
nitrite at 550 nm. (K. Grasshoff et al, 1983).


Sample handling

The samples were collected in 100 ml high-density polyethylene sample 
bottles, after first being rinsed three times with a small amount of the 
sample, taken directly from the CTD-rosette bottles. The samples were kept 
cool and dark, stored in a refrigerator and analysed normally within 10 hours 
and within 16 hours as a maximum. Analyses were carried out using 
high-density polyethylene "pony-vials" with a volume of 6 ml, they were 
rinsed three times before filling with the samples. For duplicate analysis 
purposes between runs, the deepest sample at every station was capped in a 
pony-vial to be measured for a second time during the next run. To avoid 
evaporation during the runs, all vials including the calibration standards 
used were sealed with "parafilm" under tension, so that a sharpened sample 
needle easily penetrated through leaving a small hole in the film.


Calibration and Standards

Nutrient primary stock standards were prepared at the home laboratory, NIOZ.
Phosphate: by weighing Potassium dihydrogen phosphate in a calibrated 
volumetric PP flask set to 1mM PO4.

Silicate: for silicate a certified standard (Merck) was diluted until 1.78 mM 
Si (stored at room temperature in an 100% humidified box).
Nitrate: weighing in Potassium nitrate set to 10mM NO3.
Nitrite: weighing in Sodium nitrite set to 1mM NO2.

The calibration standards were prepared daily by diluting the separate stock 
standards, using three electronic pipettes, into four volumetric 100 ml PP 
flasks (calibrated at the lab) filled with low nutrient sea water LNSW. The 
blank values of the LNSW were measured on board and added to the calibration 
values to get the absolute nutrient values.

Cocktail standard

This standard acts as a lab reference and its use is described under "quality 
control". It is made in the laboratory containing phosphate, silicate and 
nitrate in a solution containing 40 mg Hg2Cl2 per litre as a preservative. 
Every time it was used it was diluted 250 times with the same pipette, and 
the same volumetric flask.

Quality Control

Our standards have already been proven by inter calibration exercises like 
ICES and Quasimeme, and last year the RMNS exercise organised from Michio 
Aoyama MRI/Japan, to be within the best obtainable limits to the mean of the 
better laboratories. To gain some accuracy the Cocktail standard is monitored 
now since 1997, showing between run reproducibility better than 1.5%, but 
typically 0.7% of its average value.


                            average value  SD       N
                     -----  -------------  -------  --
                     PO4     086 µM        0.008µM  74
                     SiO2   13.5 µM        0.054µM  74
                     NO3    13.9 µM        0.091µM  52


The advantage of a cocktail standard is like using a reference standard with 
three nutrients mixed into one bulk, giving for each run a quite good 
overview of how the instrument is performing. It also provides a methodology 
to correct data from run to run for producing better isoline-plots from 
station to station along horizontal surfaces within the ocean.

In preceding cruises, especially in an area like the Weddell Sea, where 
nutrient gradients in deep water are very small, back-correction (implying a 
factor in each run to multiply with, for gaining the average cocktail value 
after the whole transect in each run) with use of the cocktail is absolutely 
necessary to be able to discern the small true differences between samples.

Others have reported the use of a real reference standard supplied from deep 
water (2,000 m) but this turns out to be not stable over a period longer than 
three weeks. However during the second transect of the current cruise, the 
cocktail-based data produced was well within expected performance, so 
back-correcting afterwards is not necessary.

During the cruise, a graph was made for all the runs with a listing of the 
cocktail values. So bad runs were easily recognised if a value was not within 
the alarmsettings of +1- 1.5% (this was typically better than +1- 1%). 
Deviations beyond the +1- 1.5% verification setting, did upon further 
verification, usually show up as irregularities of the analyser instrument 
(as noisy peaks, or gain calculation problems etc.), upon which the given 
samples were then re-analysed.


Statistics

For most of the nutrient parameters in this area it was not interesting to 
calculate the mean detection limit MDL. The exception was NO2, which showed a 
few small detectable peaks at the surface layer, and for the rest of the 
profile values around or below detection limits smaller then 0.01 µM. In the 
same statistical run the MDL was calculated as well as the standard deviation 
on standards at two levels.

Mean Detection Limits (calculated as 6 x S.D. of the sampled baseline water)


                             µM     Used measuring ranges µM
                    -------  -----  -------------------------
                    PO4      0.01   1.50 *
                    SiO2     0.01   18.0 *
                    NO3+NO2  0.03   21.0
                    NO2      0.007  1.00


* for SiO2 the preset range of the instrument was raised in the most Eastern 
  part of the cruise region, to higher range of 31 µM, and similarly for PO4 
  to higher range of 2 µM. This was necessary because of the highly 
  nutrient-enriched waters derived from the Pacific Ocean at a depth around 
  70 - 125 m.

Reproducibility: of 5 sample bottles at two levels given with coefficient of 
variation%

                level I   Std dev.  Cv%    level II   Std dev.  Cv%
       -------  --------  --------  ----   ---------  --------  -----
       PO4      0.193 µM  0.002     0.85   0.96 µM    0.002     0.17
       SiO2     2.504 µM  0.001     0.39   8.772 µM   0.021     0.24
       NO3+NO2  0.312 µM  0.004     1.43   13.648 µM  0.029     0.22
       NO2*     0.010 µM  0.001     0.11*  0.41 µM    0.002     0.20*

* For NO2 the% listed is the percentage of the full scale value due to the 
  low natural concentration in the seawater being only lower than 40% of full 
  range!


In order to obtain better values, an attempt was made to scale in the range 
for the nutrients to be measured such that the maximum was always at a level 
of 60 - 90% of full scale.


Cross-runs statistics

In order to obtain cross-run statistical values, analyses were carried out 
twice on the same sample from the bottle closed at the bottom layer in the 
first run, and in the consecutive run. This provides the possibility to 
estimate the precision from station to station in a horizontal way. It is 
well known that the reproducibility within one calibrated run for an auto 
analyser is much better than measurements made across several runs, with each 
run having its own calibration settings. Analysis of these (cross runs) 
duplicate samples shows that the absolute differences are for

      PO4      to be s.d.0.015µM   (avg. level 0.9µM PO4 n=23)
      SiO2     to be s.d.0.l3lµM   (avg. level 8.l7µM SiO2 n=23)
      NO3+NO2  to be s.d.0.175µM   (avg. level 12.70µM NO3+NO2 n=23)

In the raw data set of the first transect, due to the improvement in 
temperature stability during following transects those values will improve 
especially for PO4 to better than 0.01 µM.

Nevertheless, for our cocktail standard measured in every run, the resulting 
values remained stable for all nutrients during the cruise. In the future it 
would be highly advisable to produce and distribute a certified nutrient 
reference material, like the standard seawater for salinity, DIC, DOC. Such 
approach is now being pursued in the international community, and very likely 
would greatly improve the true accuracy, hence much improved compatibility of 
data better comparison between various laboratories and cruises.


Problems

Temperature stability

Temperature stability of the laboratory container in the first week, using an 
airconditioning unit just diagonal opposite the analyser, gave a data offset 
been seen in recording the cocktail standard in a plot of +1 0.02 µM PO4. 
Just by placing a kind of sieve curtain between the air-conditioner and the 
analyser to lead the cold air not massively, but gently towards the other 
half of the container where the analyser is placed, largely solved this 
problem. This curtain improved the temperature stability within 1°C instead 
of 2°C, and reduced the cross-run offset in the cocktail for PO4 to +1-0.01 
µM on a value of 0.86 µM PO4.

Evaporation during analysis

After the first two transects, I noticed that evaporation of sample water in 
the sampler-tubes can effect the data depending on the length of the run and 
the volume in the tube; evaporation was about 1.6% per day, so 0.1% within a 
run from start to end (measured relative humidity in the lab-container was 
around 23%!!). It was clear that all sample tubes in the sampler should be 
covered with parafilm, although there is routinely made a gain-drift control 
assuming that the drift for all samples tubes is linear.


Fig. 5.3: Dissolved silicate values obtained during the third transect



References

Murphy, J. & Riley, J.P., 1962. A modified single solution method for the 
    determination of phosphate in natural waters. Analytica chim. Acta 27, 
    31-36

Strickland and Parsons, 1972: A practical handbook of sea water analysis, J. 
    Fish. Res. Bd. Canada. 167: 311 pp.

Grasshoff K. et al, 1983: Methods of seawater analysis. Verlag Chemie GmbH, 
    Weinheim. 419 pp.




Subproject A6: Involvement of Co, Ni, Cu, Zn, Ag, Cd in biological cycles

Sample bottles of one litre each were filled with filtered seawater for 
measurements afterwards in the home laboratory of Co, Ni, Cu, Zn, Cd as well 
as dissolved Fe. Latter dissolved Fe as a duplication hence 
confirmation/verification of the direct shipboard detection. The home 
laboratory measurement of this suite of trace metals will be done by 
High-Resolution Inductively Coupled Plasma Mass Spectrometry (HR-ICP-MS) with 
preceding in-line column pre-concentration of the metal elements from 
seawater. Another set of small 60 ml bottles was collected and stored for 
measurements afterwards of dissolved silver Ag in the laboratory of 
collaborator Dr. Eric Achterberg, National Oceanography Centre, Southampton, 
UK.



5.2  B- natural and anthropogenic radionuclides

Subproject B1: 234Th as tracer of export production of POC
               Michiel Rutgers van der Loeff(1),   (1)Alfred-Wegener-lnstitut
               Ingrid Voege(1),                    (2)Xiamen University, China
               Pinghe Cai(2), Kate Lepore(3)       (3)University College of Dublin


Objectives
1) To acquire accurate estimates of upper ocean POC export fluxes in the 
   Arctic Ocean;
2) to infer the export fluxes of some particle-reactive elements/compounds 
   (i.e., Fe, Al, Mn, Cu, Cd, Ni, Zn, and Ag) that will be measured by other 
   researchers in the same regions (see section on trace metals above; and
3) to carry out the intercomparison of POC export studies between 234Th/238U 
   and 210Po/210Pb methods.

Work at sea
1. A total of 38 depth profiles of total 234Th in 4-l water samples were 
   collected with the CTD.

2. In-situ pumps were deployed at 14 stations (see Table GEOTRACES-5.1) down 
   to 1,000 m depth. The pumps at 50, 100, 150, and 300 m depth were equipped 
   with multiple filter heads enabling the successive filtration over 100 µm, 
   53 µm screens and a 1 µ precombusted QMA filter in order to determine the 
   POC/234Th ratio in those size fractions. MnO2 cartridges mounted behind 
   these filters and in two more pumps deployed at 500 and 1,000 m depth will 
   be used for the analysis of the 228Ra/226Ra ratio (see subproject 4).

3. The AWI-Isitec developed automated 234Th analyser "Nick" was used 
   throughout the expedition. The performance of the system and the yield of 
   the automated MnO2 precipitation was judged using deep-water samples and 
   by comparison with the 4-l technique.

Preliminary results

The depth profiles of total 234Th show that the 234Th deficit is only 
confined over the shelf and in the upper 25 m of the deep basin of the Arctic 
Ocean (see Fig. 5.4 below). This indicates that little export of POC has 
occurred over the time scale of 234Th in all areas aside from the shelf 
region.


Fig. 5.4: Depth profiles of total 234Th over the shelf (a), slope (b) and 
          deep basin (c, d) of the Arctic Ocean. Note that the recovery for  
          234Th is yet to be determined.


The automated 234Th analyser ("Nick") yielded a semi-continuous series of 
samples of suspended matter. Beta counting yielded the surface distribution 
of particulate 234Th (Fig. 5.5) which is closely related to the suspended 
load. Suspended load was very low in the Nansen Basin and in the eastern and 
southern Makarov Basin. Higher loads were observed in shelf regions but also 
over the Lomonosov and Alpha Ridge.

Samples obtained in this way could in principle also be used to obtain a 
distribution of e.g. POC or chlorophyll-a.


Fig. 5.5: Map of the distribution of particulate 234Th (expressed as 
          234Th/238U ratio) in surface water over the entire cruise


The performance of the automated precipitation was improved by adjusting the 
addition of chemicals and the time allowed for the formation of the MnO2 
precipitate. The recovery remained somewhat lower than for the usual 
small-volume technique, but sufficiently reproducible to allow the monitoring 
of the 234Th depletion in the surface water (Fig. 5.6). Final results can 
only be given after the yield determinations of the 4-l technique through the 
recovery of the added 230Th spike.


Fig. 5.6: The distribution of particulate and total 234Th/238U ratio in 
          surface waters on the last transect along the Gakkel Ridge towards 
          the Laptev shelf, as measured with the automated Th analyser. 
          Particulate Th, related to overall particulate load in the surface 
          water, goes through a minimum in the deep Amundsen Basin. Whereas 
          there is some minor depletion of total 234Th in the deep basin, a 
          consistent and large depletion signifying an export of particles to 
          the seafloor is only found on the shelf, south of 77°N.



Subproject B2: Tracing flux of particulate organic matter with Polonium-210 
               and Lead-210
               Oliver Lechtenfeld
               Alfred-Wegener-Institut

Background

Polonium-210 (210Po, 138 days half life) and Lead-210 (210Pb, 22.3 years half 
life) are produced by stepwise radioactive decay of Uranium-238 in seawater. 
These particlereactive radionuclides 210Po and 210Pb are present in seawater 
in dissolved form and adsorbed onto particles. Following adsorption onto 
particle surfaces, 210Po especially is transported into the interior of cells 
where it binds to proteins. In this way, 210Po also accumulates in the food 
chain. 210Po is therefore considered to be a good tracer for POC, and traces 
particle export over a timescale of months. 210Pb adsorbs preferably onto 
structural components of cells, biogenic silica and lithogenic particles, and 
is therefore a better tracer for more rapidly sinking matter. In combination 
with Thorium-234 the 210Po-234Th tracer pair can be used to distinguish POC 
and silica flux. This work is coordinated by Jana Friedrich, AWI.


Objectives

Our goals were (1) to get a better insight into Po binding sites in organic 
matter and (2) to trace pathways of particulate and dissolved matter leaving 
the Siberian Shelf.

(1) We investigate to which extent TEP can play a role in extending 210Po as 
    a proxy for POC and whether TEP and 210Po/210Pb data can be related. This 
    work is done in collaboration with Maya Robert and Uta Passow (AWI). We 
    further look into the distribution of 210Po on POC and in different 
    plankton assemblages.

(2) The pathways of particulate and dissolved matter were followed by the 
    combined use of 210Po and 234Th as a tracer pair (and perhaps 210Pb) for 
    particle flux (collaboration with Pinghe Cai). This information gathered 
    from water column will be complemented with the results of the 
    210Po-210Pb study in sea ice (subproject 5, Patricia Camara) to provide a 
    more thorough picture of particle transport from the shelf to the open 
    sea and from surface to depth.

Work at sea

Depth profiles of 210Po/210Pb were sampled at 17 stations (see Table 5.1). In 
the upper approximately 500 m of the water column the particulate matter was 
collected separately by filtration; below this depth total activities were 
determined. At the same stations samples were taken for the analysis of POC 
and TEP. At a further 34 stations (Table 5.1) these same parameters were 
sampled in the surface water from the ship's seawater line.

Expected results

The study of 234Th (subproject 1) showed that very little export had occurred 
on the time scale of 234Th decay in all areas apart from the continental 
shelf. All the more interesting will it be to see whether any vertical 
particle transport can be observed on the somewhat longer time scales 
reflected by the isotopes 210Po and 210Pb.



Subproject B3: Pan-Arctic Investigation of Modern and Past Changes in      
               Boundary Scavenging using 231Pa and 230Th
               Kate Lepore
               University College of Dublin

Background

The distributions of 231Pa and 230Th in the Arctic Ocean are primarily 
determined by patterns in particle flux and boundary scavenging. The 
activities of these radionuclides in marine sediments can be used to 
investigate past changes in the magnitude and spatial distribution of 
particle flux. In addition, changes in the water column distribution of these 
isotopes are indicative of changes in the rate of water mass ventilation as 
well as particle flux.

In this stud', water samples were collected for the analysis of total, 
dissolved, and particulate 231Pa and 230Th activities. In addition, sediment 
cores were sampled for the analysis of surface sediment and down-core 
activity ratios of 231Pa and 230Th. As part of the GEOTRACES programme, 
intercalibration samples were collected at one station in the Makarov Basin 
for the analysis of total 231Pa and 230Th at participating laboratories.


Work at sea

Water samples were collected along all transects (Fig. 5.4). Depth profiles 
of samples for analysis of total 231Pa and 230Th were collected at 6 
stations, and samples for particulate and dissolved 231Pa and 230Th were 
collected at 9 stations. These stations were distributed across the basins 
and margins visited during ARK-XXII/2.

Sediment samples were collected everywhere samples were taken for dissolved 
and particulate 231Pa and 230Th. In addition, subsections of the box cores 
were taken throughout the shelf and slope of the Kara Sea, and the Nansen, 
Amundsen, and Makarov Basins (Fig. 5.5).

Water samples were collected at 2,000 m in the Makarov Basin (station 328) 
for the intercalibration of total 231Pa and 230Th measurements. 1 L,2 L, and 
10 L samples were collected and acidified with ultra clean nitric acid for 
independent analyses at ten oceanographic institutes.

Preliminary Results;

231Pa and 230Th analyses are performed in an ultra clean laboratory using 
mass spectrometry techniques. Therefore, no analyses were performed at sea.


Fig 5.7. Map of water column 231Pa and 230Th samples. Circled stations are 
         those where dissolved and particulate samples were collected.

Fig 5.8. Map of sediment samples collected for 231Pa and 230Th analysis.



Subproject B4: Radium isotopes and 227Ac
               Michiel Rutgers van der Loeff, Lars Gremlowski
               Alfred-Wegener-Institut


Background

Four radium isotopes are supplied to the ocean by contact with the continent 
or deep-sea)-sediments: 223Ra, (half-life 11.4 d); 224Ra (3.7 d), 226Ra 
(1620 y) and 228Ra (5.8 y). The distribution of these isotopes in seawater 
has been shown to be most helpful to evaluate shelf-basin exchange and water 
residence times. The distribution of 228Ra can add information on 
circulation time to that obtained of the other tracers of fresh water and 
continental inputs (like δ18O, fluorescence, Ba, nutrients). Like Ra 
isotopes, 227Ac is a released from sea sediments, but its main source is in 
deep-sea sediments. This tracer is therefore especially useful to study deep 
water mixing and ventilation.

Cooperation partners: Claudia Hanfland (AWI), Brad Moran (URI).


Work at sea

222Rn was monitored semi-continuously by gas-water exchange in the ship's 
seawater supply using a RAD7 system.

Large volume surface water samples were collected for radium isotopes using 
the Polarstern's seawater intake, filtered through a 1 µm cartridge filter. 
For 228Ra/226Ra, 1 - 2 m3 of filtrate were passed over MnO2-coated 
polypropylene cartridges. The isotope ratio will be quantified in the home 
laboratory by Soxhlet leaching and subsequent gamma spectroscopy. For the 
quantification of 226Ra Radium was coprecipitated on BaSO4 from 20-l samples 
(19 surface water samples and 2 depth profiles; see Table 5.1). 226Ra in 
other samples will be interpolated from a relationship we expect to derive 
between 226Ra and dissolved silicate (determined in the same samples by 
Karel Bakker, NIOZ).

For short-lived radium isotopes, the filtrate was transferred to 300-l tanks, 
positioned in the fish lab of Polarstern. Each sample was pumped at <1 1/min 
using an electric in-situ aquarium pump (in each drum) through 
MnO2-impregnated acrylic fiber to scavenge radium isotopes. Occasionally 
subsurface samples of 50 - 150 L were obtained with the regular CTD (stations 
272, 277, 342, 407, 409, 411; Table 5.1 GEOTRACES). Fibers were partially 
dried using compressed air, and short-lived 223Ra and 224Ra measured using 
RaDeCC detectors. Longer-lived 228Ra will be measured on the fibers by gamma 
counting 228Ra/226Ra ratio in the shore-based lab and/or by recounting the 
224Ra activity after ingrowth of 228Th.

Further subsurface samples were obtained with in-situ pumps. Whenever these 
were deployed (subproject 1; 14 stations, see Table 5.1 GEOTRACES) 1 or 2 
MnO2-coated cartridges were fitted behind the filtration unit to absorb Ra 
and Th isotopes. These pump casts will provide 228Ra activities at 50, 100, 
150, 300, 500 and 1,000 m depth.

In all major basins 227Ac profiles were sampled. At stations 260, 266, 301 
(Nansen Basin), 309, 363 (Amundsen Basin), and 328 (Makarov Basin) 50-l 
samples were collected at 3 - 5 depths in parallel with sampling of 231Pa 
(subproject 3, Kate Lepore), an essential parameter to determine which part 
of the 227Ac activity is supported by its parent 231Pa. The Ac was 
coprecipitated on MnO2, the filters will be analysed in Bremerhaven.

Preliminary Results 

Radium isotopes

The activity of 224Ra in surface water, as measured with the RaDeCC system, 
showed a dramatic increase from the Nansen Basin towards the Alpha Ridge. It 
is to be expected that 224Ra reflects the distribution of 228Ra. Although 
the intermediate nuclide 228Th is highly particle reactive, the Th export 
from the surface water is so low (subproject 1 that it is not to be expected 
that 228Th becomes strongly depleted relative to 228Ra, and the short lived 
daughter of 228Th must then be in near-equilibrium with 228Ra. We therefore 
consider the distribution measured on board as representing 228Ra. The 
salient results are

1- Even at the easternmost station in the Canada Basin there was no 
   indication yet of a decline of 228Ra, so we may not even have reached the 
   maximum 228Ra activity. The maximum 228Ra is much further east than 
   found in 1991, more like the situation from 1994 as described by Smith et 
   al., 2003.
2- The activity found in the easternmost station is far higher than the 
   activities found on the Laptev shelf. Contrary to our expectations, the 
   Laptev shelf cannot be the source of the high 228Ra activities in the 
   central Arctic. The source of those activities must be found in the Bering 
   StraitlChukchi Sea area.
3- recent external inputs of 224Ra can be expected from shelf sediments. The 
   usual procedure to determine excess 224Ra (not supported by 228Ra and 
   228Th) is to recount the samples after several weeks (224Ra half life is 
   3.5 days). According to this procedure, there would be excess 224Ra 
   throughout the surface Arctic, because part of the supporting 228Th is 
   removed by filtration.


222Rn

We had expected that this isotope might be measurable if there had been 
sufficient release from the seafloor and exchange with surface waters on the 
shallow shelf, but even in the shallow Laptev Sea stations 222Rn did not 
reach the detection limit of the RAD7 system (order of 4 Bq.m-3).



Subproject B5: Radionuclides as tracers of the role of sea ice in the 
               transport, dispersion and accumulation of particulate matter 
               and associated species in the Arctic Ocean
               Patricia Cámara Mor
               Universitat Autànoma de Barcelona


Background and objectives

The Arctic Ocean is covered by sea ice, perennial or seasonal. The sea ice 
plays an important role in the global and regional climate system and also in 
the oceanic circulation. It controls and influences in the surface heat, 
momentum and salt balance, energy balance between ocean-atmosphere through 
the albedo, light penetration important for biological productivity. There is 
a link between ocean-iceatmosphere, the sea ice being an indicator of 
climatic change.

Sea ice is formed during winter mainly on the shallow continental shelves 
(<50 m depth), especially the Laptev, Kara and Barents Seas (Aagaard et al., 
1981; Colony and Thorndike, 1985; Reimnitz et al., 1992; Dethleff et al., 
1998; 2005; Eicken et al., 2005). The sea ice drifts mainly with the 
Transpolar Drift (TPD) over the Euroasian Basin and the anticyclonic Beaufort 
Gyre in the Canadian Basin (Barrie et al., 1998).

The origin, drifting patterns and fate of sea ice drive the transport of sea 
ice sediments (SIS) in the Arctic Ocean. The entrainment and enrichment of 
sediments or particles take place during sea ice formation and by deposition 
during the transit (Kempema et al., 1989; Barrier et al., 1989; Reimnitz et 
al., 1992; Rigor and Colony, 1997). The sediments entrained in sea ice are 
transported for long distances across the Arctic Ocean and they are released 
during transit across the Arctic Ocean and in ablation areas, namely the Fram 
Strait, the Greenland shelves and the Barents Sea (Pfirman et al., 1990; 
Hebbeln and Wefer, 1991; Parlov and Pfirman, 1995), linking the marginal seas 
(especially Laptev and Kara Seas) to the deep basins and the central Arctic 
to the lower latitudes. Indeed, the sediment discharge significantly 
increases the sedimentation rate in the ablation areas, and even in the open 
Arctic Ocean (Nürnberg et al., 1994; Hebbeln, 2000; Dethleff, 2005).

Sea ice also transports many chemical species. Since the 1990s, a number of 
studies have pointed out that sea ice plays a major role in the transport and 
redistribution of SIS and associated chemical species, including 
radionuclides (Pfirman et al., 1995; Meese et al., 1997; Landa et al., 1998; 
Cooper et al., 1998, 2000; Masque et al., 2003, 2007). Understanding the 
processes involved in the incorporation, transport, accumulation and 
redistribution of radionuclides from different sources is essential to assess 
their fate.

Radionuclides tend to be absorbed and/or associated in sediments, especially 
in the fine-grained particles. Once incorporated, radionuclides can be 
transported and redistributed across the Arctic Ocean associated with sea 
ice-sediments. The radionuclide content in SIS depends on a number of 
parameters; mineralogy, sediment size, chemistry properties, conditions of 
sea ice formation and there are many theories trying to explain the 
radionuclide activity in SIS.

Some processes during sea ice transit may enhance the concentrations of both 
sediments and radionuclides: sea ice tends to redistribute the sediments on 
the surface through freezing/melting cycles. During transit the formation of 
cryoconite holes takes place. Some effects of its formation are the 
aggregation of sediment particles and potential enhanced scavenging of 
particle reactive species present in sea ice (Nürnberg et al., 1994). 
Furthermore, chemical species and particles may be incorporated through 
deposition of aerosols from Arctic haze, snow, rain and dry deposition. 
Another source of radionuclides is the scavenging by SIS from the ocean 
surface.

The fact that sea ice is an important agent for the transport and dispersion 
of sediments and dissolved or particulate chemical species in the Arctic 
Ocean has several implications: i) its relative importance in the global 
ocean in terms of fluxes in the Arctic and export to the Atlantic Ocean; ii) 
the basic mechanisms that regulate the incorporation of tracer element and 
isotopes (TEIs) species to ice and, particularly, to entrained sediments; and 
iii) the final fate of sediments and TEIs by release during transit and, 
especially, in main ablation areas such as the Fram Strait. These processes 
are modulated by the mechanisms that regulate the TEIs-sediments-icewater 
interactions, within a frame of changing conditions in the Arctic Ocean. 
Radionuclides, both of natural and artificial origin are a group of relevant 
TEIs in the GEOTRACES programme, as potential chronometers of environmental 
processes and because of the knowledge of their source terms.

Our objectives are focused on the study of the role of sea ice in the 
transport of radionuclides (namely 210Pb, 210Po, 7Be, 230Th, 231Pa, 234Th, 
239%, 240%, 137Cs, Ra, 129I) in the Arctic, with special emphasis on those 
associated to sea ice sediments, with the aim of using them as tracers of 
several processes. This includes: i) investigating what is the actual 
importance of the interaction with sea water in respect to scavenging of 
particle reactive radionuclides from sea water by sea ice sediments in 
comparison to atmospheric inputs during transit and concentrations in 
sediments in sea ice formation areas. ii) Potential use of some of the 
radionuclides as tracers of transit times of sea ice. iii) Estimate the 
actual balance of these isotopes in the Arctic, and in particular in respect 
to understanding the importance of release of sea ice sediments and 
associated radionuclides.

Work at sea

During this cruise samples of sea ice, atmosphere and sediments were taken 
(Table 5.3). Further sampling and analysis of water samples is described in 
subproject 8 (anthropogenic radionuclides). With respect to sea ice, we 
obtained samples of sea ice (20), sub-ice water (11), whole ice cores (14), 
melt ponds (10), sea ice sediments (9) and ice cores sliced every 10 cm (11). 
Atmosphere samples consisted of aerosol filters (18) and precipitation (wet, 
dry and weekly, 18). Samples will be further processed at the home laboratory 
by ion column chemistry and pre0pared for isotope and activity 
determinations. The measurement of 7Be and 210Pb/210Po from sea ice, water 
sub ice and whole core will be analysed via gamma and alpha detectors 
respectively in the University Autonomous of Barcelona (UAB).

In addition, 29 sediment cores were taken at selected sites for the analysis 
of the distribution of man-made radionuclides in the sediment.


Sea ice samples

Sea ice samples consist in sea ice, sub-ice water, whole cores and melt 
points. In order to obtain 100 l of water once sea ice or sea ice cores were 
melted, we collected up to 6 30 l-barrels or drilled from 6 to 10 sea ice 
core depending on the thickness of the ice in the case of sea ice samples and 
whole core, respectively. Once on board, the samples were transferred into 
120 l barrels and left to melt completely. Further processing was identical 
to the procedure described for water samples, although in this case 209Po 
spike and stable Pb as chemical carrier were added.

In addition, sea ice cores were drilled and sliced every 10 cm in order to 
analyse 210Pb and 210Po. Each section was transferred to a plastic container 
until melting, subsequently the salinity was measured and the sample was 
filtered. The filtrate was acidified with 32% HCl. 209Po and stable Pb as 
chemical yield tracers and an iron carrier solution (FeCl3)2 were added under 
constant stirring. After waiting 12 hours to reach chemical equilibrium, the 
pH was then raised to 9 with NaOH in order to precipitate the iron as 
Fe(OH)3. This scavenges the polonium and lead isotopes onto the precipitate. 
After that the precipitate was stored in small plastic bottles for further 
processing at the laboratory.

Sea ice sediments were sampled when dirty ice was detected from the vessel 
bridge and flights on helicopter. Once "dirty ice" was observed, 
approximately 20 kg of dirty ice were collected from the upper surface ice 
floes and from ridges by using stainless steel shovels or/and ice-hammer in 
order to obtain blocs of turbid sea ice. Onboard, sea ice samples were thawed 
and sea ice sediments were isolated from the supernatant liquid by careful 
decantation. Afterwards, sea ice sediments were kept frozen (-20°C) in 
plastic bags until their analysis in the laboratory.


Sediment cores

29 sediment cores were collected with the multicorer or boxcorer, depending 
on the station. The top first 5 cm were cut in slices every 0,5 cm, sections 
from 5 to 21 cm were cut every 2 cm and the bottom part every 5 cm. Samples 
were stored in the freezer at -20°C.


Atmosphere samples

Atmosphere samples consist in aerosols and precipitation (wet and dry). 18 
aerosol filters were collected by using an aerosol pump placed on top of the 
vessel (Peildeck). Filters for aerosols were changed every 24 hours during 
transit sections. Precipitation samples comprise 7 total (wet and dry), 8 wet 
and 6 weekly. Ad-hoc precipitation collectors were placed on the top of the 
Polarstern (Peildeck). For wet precipitation, the collector was opened upon 
start of rain or snow and closed at its end; for wet plus dry precipitation, 
the collector was constantly opened and the sample was changed once the wet 
event ended; the weekly samples remained opened for a week. Precipitation 
samples were transferred to plastic bottles and were acidified with 32% HCl. 
209Po, stable Pb and Be spikes as chemical yield tracers and an iron carrier 
solution (FeCl3)2 were added under constant stirring. After waiting 12 hours 
in order to get a chemical equilibrium, the pH was then raised to 9 with NaOH 
in order to precipitate the iron as Fe(OH)3. The precipitate was transferred 
to small bottles and stored for further processing at home.


Absorption/desorption experiment

In order to investigate the role that sea ice sediments play in the transport 
of 230Th and 231Pa and the implications it may have on the use of the Pa/Th 
proxy in the Arctic, an absorption/desorption experiment was carried out on 
board. It was replayed three times. The experiment consisted in adding into 
plastic bottles with filtered sea water some grams of sea ice sediments. 
These were kept in suspension using a motor which stirred the water. After 
different times (1 h, 2 h, 4 h, 6 h, 12 h, 24 h,48 h, 72 h, 96 and 120 h), a 
bottle was filtered to isolate the sediments from the water. Water was 
acidified with 32% HCl and sediment filters were kept frozen (20°C) in 
plastic bags until their analysis at home.



References

Aagaard K., 1981. On the deep circulation of the Arctic Ocean. Deep-Sea 
    Research, 82: 251-268.

Barnes P., Reimnitz E., Fox D.H., 1982. Ice rafting of fine grained sediment, 
    a sorting and transport mechanism, Beaufort Sea., Alaska. Journal of 
    Sedimentary Petrology, 52(2), 493-501.

Barrie L., Falck E., Gregor D., Iverson T., Loeng H., Macdonald R., 1998. The 
    influence of physical and chemistry processes on contraminant transport 
    into and within the Arctic. In: Gregor D., Barrie L., Loeng H., editors. 
    The AMAP assessment, 1998, 25-116 pp.

Colony R., Thorndike A.S., 1985. Sea ice motion as a drunkard's walk. Journal 
    of Geophysical Research, 90, 965-974.

Cooper L.W., Kelley J.M., Bond L.A., Orlandini K.A., Grebmeier J.M., 2000. 
    Sources of the transuranic elements plutonium and neptunium in Arctic 
    marine sediments. Marine Chemistry, 69, 253-276.

Cooper L.W., Larsen, I.L., Beasley, T.M., Dolvin, S.S., Grebmeier, J.M., 
    Kelley, J.M., Scott, M. and Jonhson-Pyrtle, A., 1998. The distribution of 
    radiocesium and plutonium in Sea Ice-entrained Arctic sediments in 
    relation to potential Sources and sinks. Journal of Environmental 
    Radioactivity, 39 (3), 279-303.

Dethleff D., 2005. Entrainment and export of Laptev Sea ice sediments, 
    Siberian Arctic. Journal of Geophysical Research, 110, C07009, doi: 10.1 
    029/2004JC002740.

Dethleff D., Loewe P., Kleine E., 1998. The Laptev Sea flaw lead - Detailed 
    investigation on ice formation and export during 1991/92 winter season. 
    Cold Region. Science Technology, 27 (3), 225-243.

Eicken H., Gradinger, R., Gaylord, A., Mahoney, A., Rigor, I., Melling, H., 
    2005. Sediment transport by sea ice in the Chukchi and Beaufort Seas: 
    Increasing importance due to changing ice conditions?. Deep-Sea Research 
    II, 52, 3281-3302.

Hebbeln D., 2000. Flux of ice-rafted detritus from sea ice in the Fram 
    Strait. Deep-Sea Reseach II, 47, 1773-1790.

Hebbein D., Wefer G., 1991. Effects of ice coverage and ice-rafted material 
    on sedimentation in the Fram Strait. Nature, 350, 409-411. Kempema E.W., 
    Reimnitz E., Barnes P.W., 1989. Sea ice sediment entrainment and rafting 
    in the Arctic. Journal of Sedimentary Petrology, 59, 308-317.

Landa E., Reimnitz E., Beals D., Pochkowski J., Rigor I., 1998. Transport of 
    137Cs and 239'240Pu with ice-rafted debris in the Arctic Ocean. Artic, 
    51, 27-39.

Masque P., Cochran J.K., Hebbeln D., Hirschberg D.J., Dethleff D., Winkler 
    A., 2003. The role of sea ice in the fate of contaminants in the Arctic 
    Ocean: Plutonium atom ratios in the Fram Strait. Environmental Science 
    and Technology, 37, 4848-4864.

Masque P., Cochran J.K., Hirschberg D.J., Dethleff D., Hebbeln D., Winkler A. 
    and Pfirman 5., 2007. Radionuclides in Arctic sea ice: tracers of 
    sources, fates and ice transit time scales. Deep-Sea Research I, doi:1 
    0.101 6/j.dsr.2007.04.01 6

Meese D.A., Reimnitz E., Tucker W.B.III, Cow A.J., Bischof J., Darby D., 
    1997. Evidence for radionuclide transport by sea ice. Science of Total 
    Environment, 202267.

Nürnberg D., Wollenburg I., Dethleff D., Eicken H., Kassens H., Letzig T., 
    Reimnitz E., Thiede J., 1994. Sediments in Arctic Sea ice: implications 
    for entrainment, transport and release. Marine Geology, 119, 185 -214.

Pavlov V., Pfirman S., 1995. Hydrographic structure and variability of Kara 
    Sea:imlications for pollutant distribution. Deep-Sea Research II, 42 (6), 
    1369-1390.

Pfirman S., Lange M.A., Wollenburg I., Schlosser P., 1990. Sea ice 
    characteristics and the role of sediment inclusions in deep-sea 
    deposition: Arctic-Antarctic comparisons. In: U. Bleil and J. Thiede 
    (eds.), Geological History of the Polar Oceans: Arctic versus Antarctic. 
    Kluwer Academic Publishers, pp. 187-211.

Pfirman S.L., Eicken H., Bauch D., Weeks W.F., 1995. The potential transport 
    of pollutants by Arctic Sea ice. Science of the Total Environment, 159, 
    129-146.

Reimnitz E., Marincovich L., McCormick M., Briggs W., 1992. Suspension 
    freezing of bottom sediment and biota in the Northwest Passage and 
    implications for Arctic Oceam sedimentation. Canadian Journal of Earth 
    Science, 29, 693-703.

Rigor I., Colony R., 1997. Sea ice production and transport of pollutants in 
    Laptev Sea, 1979-1993. The Science of the Total Environment, 202, 89-110.

Smith, J.N., Moran, S.B., Macdonald, Robie W., 2003. Shelf-basin interactions 
    in the Arctic Ocean based on 210Pb and Ra isotope tracer distributions. 
    Deep-Sea Research, Part 1, 50(3): 397-416.



Tab. 5. 3: Summary of sea ice and sediment samples

Station                  sea ice  whole  section  melt    water    sea ice   sediment
                         sample   core    core    ponds  sub ice  sediments    core
-----------------------  -------  -----  -------  -----  -------  ---------  --------
PS70/239                                                                        X
PS70/348                    X       X       X       X       X
PS70/255
PS70/357                    X       X       X  
PS70/259
PS70/260                    X       X       X       X       X                   X
PS70/363                                                                        X
PS70/264                    X       X       X       X       X
PS70/365                                                                        X
PS70/266                                                                        X
PS70/369                                                                        X
PS70/370                                                                        X
PS70/271                    X       X       X       X       X
PS70/374                                                                        X
PS70/276                                                                        X
PS70/377                                                                        X
PS70/279                                                                        X
PS70/382                                                                        X
PS70/285                    X       X       X               X                   X
PS70/394                                                                        X
PS70/399                                                                        X
PS70/301                    X       X       X       X       X                   X
PS70/306                                                                        X
PS70/308
PS70/309                    X       X       X       X       X                   X
PS70/316                                                                        X
PS70/319                                                                        X
PS70/322                    X       X               X       X
PS70/328                    X       X               X                           X
PS70/333                                                                        X
PS70/338                    X       X       X       X                           X
PS70/342                    X                       X                X          X
PS70/352                    X       X               X       X
PS70/358                                                                        X
PS70/363                    X       X       X               X
PS70/365
PS70/371                    X       X       X               X
PS70/383                                                                        X
PS70/384                                                                        X
PS70/385                                                                        X
Helistation  
  82°47,50N-33°58,37E       X  
Helistation  
  83°59,58N-34°23,12E                                                X
Helistation  
  83°59,61N-34°01,58E                                                X
Helistation  
  82°30,13N-65°45,35E       X  
Helistation  
  83°36,269N-60°23,930E     X                                        X
Helistation  
  83°25,52N-61°59,16E                                                X
Helistation  
  85°08,7N-60'48,9E                                                  X
Helistation  
  83°17,75N-86°11,32E       X  
Helistation  
  85°33,88N-90°26,36E       X  
Helistation  
  87°29,89N-109'33,33E      X  
Helistation  
  84°27,O1N-148'25,70W                                               X
Helistation  
  84°15,65N-108°44,76E                                               X
Helistation  
  82°12,91N-108°54,95E                                               X
     





Subproject B6: Rare Earth Elements, 10Be and the isotopic composition of
               Nd (eNd)
               Sabine Mertineit, Michiel Rutgers van der Loeff
               Alfred-Wegener-Institut



Objectives

Rare Earth Elements (REE) and the isotopic composition of Nd are important 
water mass tracers. The varying REE-pattern and isotopic signature of Nd is 
transferred to the ocean via processes such as riverine inputs, dust inputs, 
or leaching of shelf sediments and ice drifted sediments. In addition to 
selective weathering, elemental fractionation may also occur during aqueous 
transport, where natural particles and colloids are of great importance. The 
REE concentrations coupled with the Nd isotopic ratios are powerful tracers 
to investigate scavenging processes and to predict the fate of elements 
brought from the continent. The REE's residence times on the order of 1,000 
years make them ideal tracers for water masses as it allows for long distance 
transport while preventing complete homogenisation. Especially Nd isotopes 
are useful in paleoceanography, as their isotopic signature is preserved in 
ferromanganese nodules, foraminifera, and Fe-Mn oxide coatings of sediments.

Beryllium isotopes (cosmogenic 10Be and lithogenic 9Be) give additional 
information on inputs and water mass sources. Cooperation partners: Per 
Andersson (National Museum of Natural History, Stockholm, Sweden) and Martin 
Frank (IFM-GEOMAR, Kiel).


Work at sea

Samples have been collected for REE in dissolved and particulate form and in 
sea ice sediments.

For dissolved REE, 1-l seawater samples were collected using the ultra-clean 
CTDRosette. The samples passed an in-line 0.2µ cellulose acetate filter (see 
trace metals UCC procedure). They were acidified with 1 ml of distilled 
nitric acid (HNO3).

For the purpose of intercalibration/intercomparison with other laboratories, 
20 filtered samples were taken from the ultra-clean CTD as well as from the 
AWI CTD-Rosette at 2,000 m in the Makarov Basin (sta 328). These samples were 
acidified in the same way and will be distributed to other laboratories.

Particulate REE in surface waters was collected by the ship's seawater 
membrane pump (Klaus pump with inlet at 11 m depth) and a continuous flow 
centrifuge. 2,000 6,800 l of seawater were centrifuged at a rate of about 500 
- 1,000 l per hour at 16,000g.

Sea ice sediments were sampled by coring. Cores were sliced in 10 cm pieces 
and melted aboard in parallel with similar experiments on radionuclides 
(subproject 5). It was important to prevent contamination thus the ice cores 
were drilled by hand and the location for drilling on the floe was always 
chosen upwind so that the fumes of other working groups with motor drills or 
from the ship were blown in the counter direction. Furthermore all the 
working tools and containers for the ice slices were cleaned thoroughly 
before every sampling. Because of the different interests of all the groups 
on the ice it was not possible to start coring as the very first one without 
any other people on the floe. But at one ice station (sta. 352) it was 
examined if and how much the fumes of motor drills and helicopters could 
contaminate the ice cores. Therefore at the very beginning of the ice station 
one ice core was drilled by hand where no other group was working on the ice 
and some time later a second core was drilled by hand under the usual 
conditions as described above.


Tab. 5.4: Time and location of ice core stations where samples were collected 
          for REE studies

          number   stations     date    time         coordinates
          ------  ----------  --------  -----  ----------------------
            1     PS 70/248   02.08.07  13:13  81°56.71'N 34°2.16'E
            2     P570/260    07.08.07  09:27  84°29.37'N 36°8.29'E
            3     PS70/264    12.08.07  11:07  83°39.16'N 60°25.59'E
            4     PS70/271/1  15.08.07  12:58  82°30.18'N 60°47.67'E
            5     PS70/285/1  20.08.07  12:54  82°8.59'N  86°19.08'E
            6     PS70/301/1  24.08.07  09:25  84°34.77'N 89°50.26'E
            7     PS70/309/1  27.08.07  21:49  87°2.74'N  104°47.78'E
            8     PS70/328    02.09.07  12:00  87°49.79'N 170°33.37'W
            9     PS70/352/1  10.09.07  12:07  86°38.27'N 177°33.49'E


Fig. 5.9: Location of REE water column sampling (triangles), Ice stations 
          (circles) and locations where dirty ice was sampled by helicopter 
          (stars)


After melting, subsamples were taken for δ18O and Ba. The remainder was 
filtered and the particulate matter was stored at -20°C for later analysis of 
REE.

When sea ice sediments come into contact with seawater, desorption and 
dissolution of REE may cause changes in the REE pattern and isotopic 
composition of Nd in the seawater. This process was investigated 
experimentally. Dirty ice was sampled on three stations visited by 
helicopter:
                   #     date         coordinates
                   -  ---------  -----------------------
                   1  06 Aug 07  83°59.61'N,  43°01.58'E
                   2  08 Sep 07  84°27.01'N, 148°25.70'W
                   3  17 Sep 07  84°15.65'N, 108°44.76'E

The interaction of REE between these ice-rafted sediments and filtered 
seawater was studied by one exchange experiment (with sample #1) on board. 
After melting a large (approx 80-l) dirty-ice sample, the water was separated 
by decantation and centrifugation and about l g of the remaining sediments 
was added to freshly filtered sea water. 24 hours later it was filtered. In 
parallel, 10-l of dirty ice was melted, filtered and the filtrate was 
collected. All filtrate samples were acidified with 8 ml of distilled nitric 
acid (HNO3).

The REE composition will be compared to that of the same water with no dirty 
ice added to test the contribution of sea ice sediments to the isotopic 
signature of Nd and the REE-pattern. Further experiments with ice-rafted and 
centrifuged sediments will be carried out at AWI.

Fifteen 10-litre samples of seawater were collected from surface water and 
CTD/Rosette casts in the Russian shelf region for ENd and Be analyses (Table 
5.1). These samples were filtered and acidified for Neodymium isotopic 
analysis in Stockholm. A parallel 1-l aliquot of these filtered samples was 
acidified for 10Be and 9Be analysis in Kiel (IFM-GEOMAR).



Subproject B7: Barium as river water tracer
               Sabine Mertineit, Michiel Rutgers van der Loeff
               Alfred-Wegener-Institut


Objectives

Barium has been shown to be a powerful tracer to distinguish river water 
masses from Eurasian and American origin (Taylor et al., 2003). In 
combination with other tracers of fresh water components (salinity, δ18O, 
fluorescence, 228Ra, nutrients) it will be possible to obtain an overview of 
the distribution of the various fresh water sources during our expedition as 
a contribution to the EU programme DAMOCLES.


Work at sea

15 ml unfiltered seawater samples for Ba analyses were collected throughout 
the expedition from the ultra-clean CTD (Table 5.1). The samples were 
acidified with 30 µl distilled HCl and will be analysed with isotope dilution 
ICPMS at AWI In order to determine whether ultra clean sampling is necessary 
for this element, parallel samples were taken on the entire depth profile 
using the conventional AN Rosette at station 266.

For the purpose of intercalibration/intercomparison with other laboratories, 
20 samples were taken from a single bottle of the ultra-clean CTD as well as 
from the AWI CTD at 2,000 m in the Makarov Basin (sta 352). These samples 
will be distributed to other laboratories.


Reference

Taylor J. R., Falkner K. K., Schauer U., and Meredith M. (2003) Quantitative 
    considerations of dissolved barium as a tracer in the Arctic Ocean. J. 
    Geophys. Res. 108(C12), 3374, doi: 10.1 029/2002JC001 635.



Subproject B8: Anthropogenic radionuclides 129I, 99Tc, 137Cs and Pu
               Andreas Wisotzki(1), Patricia  (1)Alfred-Wegener-Institut
               Camara(2), Kate Lepore(3)      (2)Universitat Autànoma de  
                                                 Barcelona 
                                              (3)University College of Dublin

Objectives

The main source of 129I and 99Tc is their release by reprocessing plants of 
nuclear fuel. The sources (Sellafield, La Hague) are well known. 137Cs and Pu 
isotopes have also been released by fallout, and the isotopic composition of 
Pu differs among the various sources. The distribution of these tracers is 
used to model the Arctic circulation and especially the relative 
contributions of Atlantic and Pacific water masses. Cooperation partners: 
Michael Karcher (AWI/OASYS), Claudia Hanfland (AWI), Pere Masque (UAB), Brad 
Moran (URI), John Smith (BIO).

Work at sea

Plutonium isotopes and Cesium-137 are present at ultra trace levels, which 
complicates their measurement and forces us to collect large volume samples. 
Large volume (100 I) sea water samples were collected by using Niskin bottles 
mounted on the CTD rosette, comprising 56 surface water samples and 11 water 
profiles of 6 - 8 depths (Table 5.1). The 100-l water samples were 
transferred into a plastic barrel and acidified with 32% HCl. 242Pu and 
stable Be spikes as chemical yield tracers, and an iron carrier solution 
(FeCl3)2 were added under constant stirring. After waiting 12 hours in order 
to get a chemical equilibrium, the pH was then made 9 with NaOH in order to 
precipitate the iron as Fe(OH)3. This scavenges beryllium and plutonium onto 
the precipitate while cesium stays in solution. After that the supernatant 
was transferred into a second barrel while the precipitate was stored in 
small plastic bottles for further processing in the home laboratory. By 
addition of 65% HNO3, the pH was lowered again and subsequently stable Cs 
was added as a chemical yield tracer. Then, a pre-weighed sample of ammonium 
molybdophosphate was added while stirring thoroughly. This produced a yellow 
precipitate that scavenges Cs. The precipitate was left to settle and then 
transferred into smaller bottles. The isotopic composition of Pu will be 
determined by AMS at the University of Sevilla, Spain, and 137Cs activities 
will be analyzed via gamma-counting at UAB. In one case (station 352) 137Cs 
was collected with KCFC resin (to be analysed at BIO).

129I was sampled along with 137Cs and Pu isotopes (Table GEOTRACES-1). 1-l 
samples for 129I were collected from the CTD-rosette and stored with no 
special requirements (e.g., at room-temperature) until shore-based analysis. 
99Tc was sampled especially in the Atlantic layer (Table 5.1). For 99Tc, 5-I 
samples were collected from the CTD-rosette and stored in plastic canisters 
with no special requirements.



5.3   C-related parameters

Subproject Cl: Dissolved organic matter (DOM) in the Arctic Ocean
               Sally Walker
               Texas A&M University at Galveston


Recently a project was funded by the NSF (USA) to investigate the formation 
of the Arctic Halocline by using organic tracers and in-situ fluorescence. We 
found elevated levels of terrestrial DOM and associated fluorescence in 
halocline waters which is very intriguing and suggests the involvement of 
river water in halocline formation. Unfortunately, we also found that the 
fluorescence signal can not be attributed to the terrestrial source alone, 
rather there seem to be several sources involved. The in-situ probe is very 
general and does not allow any further distinction of the fluorescence 
signal. With the samples we hope to unravel some of these different sources 
by using a new method combining high resolution spectrofluorometry and 
parallel factor analysis. With this new method we will be in a much better 
position to determine source waters involved in halocline formation which is 
critical for our understanding of climate change the Arctic Ocean system.


Work at sea

With a fluorescence sensor mounted on the CTD unit, a full water depth 
profile of fluorescence was obtained at every station. This gave us real time 
information on the horizontal and vertical distribution of fluorescence which 
is caused by certain organic compounds dissolved in seawater. Based on the 
fluorescence signal we selected our sampling location and depth and collected 
between 1 and 20 l of seawater at several depth levels. We collected samples 
from 70 stations throughout the cruise as seen in figure 5.10. Each sample 
was split into subsamples to determine the following parameters: dissolved 
organic carbon (DOC), dissolved organic nitrogen (DON), optical properties 
(absorbance and fluorescence), carbohydrates, lignin phenols, and nitrogen 
isotopes (nitrate). Water samples were immediately filtered, frozen and will 
be transported to the home laboratory for further analyses which will take 1 
to 2 years to complete.


Preliminary Results

Depth distribution of in-situ fluorescence in the Arctic Ocean reveals a 
subsurface maximum of fluorescence throughout most of the basin, most likely 
dominated by river CDOM. The maximum fluorescence in the Canadian Basin 
appears to be associated with halocline waters while the maximum fluorescence 
in the Eurasian Basin appears to be associated with the transpolar drift 
surface waters. To confirm a common origin of fluorescence in these two 
basins, we will pair them with molecular level analyses, 3-dimensional 
fluorescence coupled with parallel factor analysis.


Fig 5.10: Map of stations collected during ARK-XXll/2 for DOM analyses.



Subproject C2: Phyto- and Protozooplankton ecology in the water column
               Lilith Kuckero
               Alfred-Wegener-Institut


Objectives

Since the 1990s phyto- and protozooplankton ecological investigations on 
biomass, productivity and related biochemical parameters as chlorophyll a, 
particulate organic carbon! nitrogen (POC!PON), carbonate, and biogenic 
silica have been carried out in Arctic waters mainly in the Fram Strait area. 
During the years 1993-1996 sampling was also conducted in the Amundsen and 
Nansen Basins. During the present cruise, about 10-15 years later, the same 
investigations will be done for comparison with the old data to understand 
eventual changes due to a changing environment. Specific questions addressed: 
Are there regional differences in the seasonal distribution patterns of 
phytoplankton and protozooplankton, POC/PON, carbonate and biogenic silica in 
the ice-covered Arctic ocean? What is the influence of the respective abiotic 
factors? Which are the most remarkable processes within it for the pelagic 
food web? What changes can we measure?


Work at sea

Water was sampled with the CTD-Rosette at different depths from surface to 
bottom for filtration and measurements of Chlorophyll a, POC/PON and biogenic 
silica at home. Also water samples from surface down to 100 m depth have been 
taken (from the CTD) for species abundances on several station during the 
cruise. These samples were fixed with buffered formaldehyde. Additionally 
water was sampled by Apstein-net (a hand-operated net) at several stations to 
obtain an overview of the biodiversity.


Expected results

A quantitative and qualitative analysis of the fixed samples will give an 
overview of the regional and seasonal distribution of the phyto- and 
protozooplankton species and their abundances. At station PS70/411 a high 
abundance of several Chaetoceros species was observed in contrast to the 
species abundance observed at other stations.



5.4  Coupling of methane and DMSP cycles in the marginal ice zone and on 
     polar shelves
     Ellen Damm, Ingrid Voege
     Alfred-Wegener-Institut


Background

Recent change in the Arctic may have profound effect on natural 
biogeochemical cycles in seawater. Especial feedback effects to pathways of 
climatically relevant biogases like methane and DMS will loom large in the 
equation of change. The recent marine methane cycle is influenced by 
microbial induced in situ production, which creates a methane surplus 
relative to the atmospheric equilibrium concentration in ocean surface water. 
However the potential of the upper ocean methane cycle remains underestimated 
because in situ production is masked by a simultaneous and nearly equals in 
situ oxidation. Hence the carbon isotopic ratio of methane will be used to 
trace the in situ production and the subsequent consumption processes, which 
provides insights into the recent methane cycle.

A principal pathway by which methane is readily formed is the methylothrophic 
methanogenesis. However, direct evidence of this role of methylated 
substrates in aerobic seawater is still lacking. An abundant methylated 
substrate in the surface ocean is dimethylsulfoniopropionate (DMSP). Large 
amounts are produced annually by marine phytoplankton and its turnover plays 
a significant role in carbon and sulfur cycling in the surface ocean. 
Cleavage of DMSP leads to formation of dimethylsulfide (DMS). DMS partly 
escapes to the atmosphere while bacteria will oxidize large amounts of DMS in 
the water column before it can be released to the atmosphere. Anaerobic 
metabolism of DMS may result in the production of methane. Here we focus on 
the coupling of methane production/consumption cycle with DMSP turnover in 
polar water during phytoplankton bloom especially in the marginal ice zone.


Work at sea

Methane concentrations were measured at 50 stations in the Barents Sea, Kara 
Sea, Laptev Sea and in the central Arctic Ocean. Water samples were collected 
in Niskin bottles mounted on a rosette sampler from 200 m depths up to the 
surface. The dissolved gases were immediately extracted from the water and 
were analysed for methane by gas chromatograph equipped with a flame 
ionization detector (FID) on board ship. Gas samples were stored for 
investigations of the δ13CCH4 values in the home laboratory. Furthermore at 
each station samples for the analyses of DMSP (p), DMSP (d), were taken, 
which will be analyzed in the home lab.


Preliminary results

Methane in situ production occurs during the summer phytoplankton bloom in 
surface water indicated by increased methane concentrations. The highest 
methane concentration is detected at the marginal ice zone shown along a 
transect running from an open water region up to under the ice figure 5.11.

In the shallow shelf area of the Laptev Sea unusually extended methane 
anomaly is indicated by methane concentrations up to 300 nM, which exceeds 
the equilibrium concentration with the atmosphere up to hundredfold. Here the 
strong water stratification and freshwater inflows primarily restrict the sea 
to air flux of methane during summer and consequently favour the accumulation 
of methane in the water column. Conspicuous is the missing methane oxidation, 
which should be under consideration for deeper understanding of the recently 
marine methane cycle.


Fig. 5.11: Methane concentration in polar surface water





6.  MARINE BIOLOGY


6.1  Zooplankton investigations
     Kristina Barz(1),                   (1)Alfred-Wegener-Institut
     Adrian Basilico(1),                 (2)Shirshov institute of Oceanology
     Ksenia Kosobokova(2),               (3)Laboratoire d'Oceanographie et de
     Antoine Nowaczyk(3)                    Biogeochemie


Background and objectives

Biological investigations in the Greenland Sea and Eurasian Basin in the 
1990s demonstrated that the composition and distribution of pelagic fauna in 
the Arctic Ocean is strongly affected by the inflow of Atlantic water (Hirche 
& Mumm, 1992, Mumm, 1993; Kosobokova, Hirche, 2000). This inflow advects 
North-Atlantic zooplankton populations from the Greenland Sea via the Fram 
Strait and from the Barents Sea shelf into the Eurasian Basin (Hirche & Mumm, 
1992; Kosobokova & Hirche, 2000). The conditions to which these populations 
are physiologically adapted are quite different from the conditions in the 
Arctic Ocean, so that their survival in the Arctic largely depends on their 
tolerance to Arctic conditions. While many species die off shortly after 
entering the Arctic Ocean, others survive due to their starvation potential 
or even continue their development for some time. Consequently, the Arctic 
Ocean is a large sink for organic carbon produced in the North Atlantic, with 
the region of sedimentation dependent on transport velocity and survival 
time.

During the 1990s, various observations indicated that the circulation of 
Atlanticderived water in the Arctic Ocean had changed considerably. In the 
Eurasian Basin the Atlantic layer had become warmer and saltier (Schauer et 
al., 2004) and the boundary between the Atlantic and Pacific waters moved 
into the Canada Basin to an extent not previously observed (McLaughlin et 
al., 2002). These changes may have strong consequences for the pelagic 
ecosystems and hence sequestration of carbon and biogeochemical cycles in the 
Arctic Ocean. An increase of advection of Atlantic populations may 
significantly increase the sedimentation of advected biogenic material. 
Further warming could favour the survival of the highly productive Atlantic 
communities, which finally could replace the Arctic fauna characterized by 
low biomass and low production (Hirche & Mumm, 1992; Kosobokova & Hirche, 
2000).

In order to understand the processes and factors regulating advection of 
Atlantic zooplankton in the western Arctic and trophodynamic processes in the 
Arctic pelagic communities, the zooplankton work focused on the following 
aspects:
of preferential grazing

  • composition and spatial distribution of zooplankton on transects 
    perpendicular to the continental slope and across the ridges in relation 
    to the Atlantic water inflow
  • starvation potential of C. finmarchicus as an estimate of the tolerance 
    of this Atlantic copepod to the Arctic conditions - low temperature and 
    poor food availability
  • analyses of organic carbon content, C/N ratio, stable isotopes ratio, dry 
    mass, and lipid composition to understand the life strategies and 
    trophodynamic relationships in the Arctic pelagic food web
  • reproductive biology and life strategies of mesopelagic copepods to 
    understand adaptations of the deep-water plankters to short pulsed flux 
    of organic matter down to the Arctic deep sea
  • role of small plankton animals (1 - 2 mm) in the Arctic food web and 
    their adaptations to the environment



References

Hirche, HJ & Mumm, N (1992). Distribution of dominant copepods in the Nansen
    Basin, Arctic Ocean, in summer. Deep-Sea Res. 39 Suppl. 2: S485-S505.

Kosobokova, KN & Hirche, HJ. (2000). Zooplankton distribution across the 
    Lomonosov Ridge, Arctic Ocean: species inventory, biomass and vertical 
    structure. Deep-Sea Res 147: 2029-2060.

McLaughlin, F., Carmack, E., MacDonald, R.W., Weaver, A.J. & Smith, J. 
    (2002). The Canada Basin 1989-1995: Upstream events and farfield effects 
    of the Barents Sea, J. Geophys. Res. 107, doi:10.1029/2001JC000904.

Mumm, N. (1993). Composition and distribution of mesozooplankton in the 
    Nansen Basin, Arctic Ocean, during summer. Polar Biol. 13: 451-461.

Schauer, U., E. Fahrbach, S. Osterhus, and G. Rohardt (2004), Arctic warming 
    through the Fram Strait: Oceanic heat transport from 3 years of 
    measurements, Journal of Geophysical Research, 109(C06026), 
    10.1029/2003JC001823.



6.1.2  Zooplankton sampling
       Kristina Barz(1),                 (1)Alfred-Wegener-Institut
       Adrian Basilico(1),               (2)Shirshov institute of Oceanology
       Ksenia Kosobokova(2),             (3)Laboratoire d'Oceanographie et de
       Antoine Nowaczyk(3)                  Biogeochemie


For the investigation of the species composition and distribution, 
zooplankton were collected by a multiple closing net (XXL multi-net, 0.5 µm2 
mouth opening, 150 µm mesh size, Hydrobios, Kiel), which provided stratified 
sampling of the entire water column from the surface to the bottom. Multi-net 
sampling was carried out on five transects. Three of them extended seaward 
from the outer continental shelf over the continental slope into the deep 
Nansen and Makarov Basins north of the Barents, western Kara and Laptev Seas 
(Transects A, B, E). Transect C extended from the north-eastern continental 
slope of the Kara Sea into the deep Nansen Basin, crossed the Nansen-Gakkel 
Ridge, deep Amundsen Basin, the Lomonosov Ridge at latitude 88°N, and then 
extended into the deep Makarov Basin up to the Alfa-Mendeleyev Ridge. 
Transect D extended from the deep Makarov Basin across the Lomonosov Ridge to 
the Nansen-Gakkel Ridge along 85 - 86°N. Between three and five multi-net 
stations were taken on transects A, B, D, and E, and fifteen stations along 
transect C.

From four to eight layers were sampled in the shelf region, and nine depth 
layers at all stations off the shelf.

In total, 30 multi-net stations were taken: four in the shelf region, seven 
in the slope region and 19 in the deep basins (eleven of them in the area 
deeper than 3,000 m, and two deeper than 4,000 m, Fig. 6.1). The samples were 
preserved in 4% boraxbuffered formaldehyde for further processing.


Fig. 6.1.1: Zooplankton station locations. 1 - Multi-net; 2 - Bongo net, 
            100 µm and 300 µm mesh size; 3 - Bongo net, 300 µm and 500 µm 
            mesh size


For the study of life strategies and trophodynamics of the Arctic pelagic 
food web, zooplankton were collected by Bongo nets towed vertically from the 
upper 100 m and 1,500 m. The catches of the 100 and 300 µm mesh size nets 
were used for experiments with live small plankton organisms (grazing, 
excretion, gut evacuation rate, oxygen consumption). The catches of 300 and 
500 µm mesh size nets were used to collect larger animals, e.g., Calanus 
finmarchicus, for starvation experiments, and C. finmarchicus and other 
copepods for analyses of carbon, C/N ratio, lipid composition, stable isotope 
ratios, dry weight, and egg production experiments. In total, Bongo nets were 
taken at 54 stations (Fig. 6.1).



6.1.2  Starvation experiments
       Kristina Barz(1),                  (1)Alfred-Wegener-Institut
       Adrian Basilico(1),                (2)Shirshov institute of Oceanology
       Ksenia Kosobokova(2),


Starvation experiments were carried out with Calanus finmarchicus adult 
females and copepodite stage V (CV). Three experiments were set up: (1) with 
animals collected on the shelf of the Barents Sea (st. 232), (2) with animals 
from the northern slope of the Barents Sea influenced by the Atlantic water 
inflow (st. 243), and (3) with animals collected west of the crest of the 
Lomonosov Ridge at 88° N in the back flow of Atlantic water (st. 312). C. 
finmarchicus were sorted out of the Bongo net catches immediately after 
capture. Five replicates of CVs were set up in the first experiment, 3 of 
adult females and 2 of CVs in the second, and 2 replicates of adult females 
and 2 of CVs in the third experiment. Each replicate contained 50 specimens. 
Animals were placed in five 2,000 ml Plexiglas insets with 300 - 500 µm mesh 
false bottoms suspended in 3,000 ml TPX jars. The jars were filled with 
filtered sea water and kept at dim light at temperatures between -1.0 and 
+0.5°C. The condition of animals was checked every second day, dead specimens 
were removed, counted and preserved in 4% formaldehyde. The water was 
exchanged on the third day after set up of each experiment, and every second 
week thereafter.

After 30 days of starvation, 10 CV specimens were removed from the first 
experiment at random and stored for Corg content analysis. From the second 
and third experiments, 8 females and 8 CV specimens were removed after 30 
days for the same purpose. After 60 days of starvation the same amount of 
animals were again removed for Corg analyses from each pool of experimental 
animals.

Preliminary results

The first two experiments were set up on 2 and 3 August, and thus lasted for 
nearly two months until the end of the cruise, while the third experiment was 
set up at the very end of August and lasted 5 weeks. In all experiments, the 
number of animals decreased by the end of observational period, although the 
patterns of decrease differed between the experiments and between stages 
within the same experiment.

In Exp.1 with CVs from the northern Barents Sea there was no mortality within 
the first 9 days after setup (Fig. 6.1.2 a,). During the following two months 
abundance was decreasing at a rate of ca. 4 in day-1. After 2 months only ca. 
10% of the animals survived (Fig. 6.1.2). Some of them moulted to adult 
females, which might have caused high mortality.

In Exp.2 with animals collected north of Svalbard, the first dead females 
were observed on day 9, and the first dead CVs on day 17 (Fig. 6.1.3 a, b). 
The mortality rates were lower than in Exp.1. In CVs a rate of ca. 0.6 md 
day-1 was observed, while in females it was ca. 1.3 md day-1. By the end of 
the observational period, ca. 70% of CVs and 45% of females had survived.

In Exp.3 during the first 12 days a few CVs and females died (Fig. 6.1.4 a, 
b), however, mortality rate was quite low. Within the next 25 days of 
observations the mortality rate in CVs averaged 0.7, and in females 1.6 md 
day-I. The mortality rates were higher in adult females compared to CVs in 
both Exp.1 and Exp.2. The experiments will be continued at AWI


Fig. 6.1.2 a, b: Calanus finmarchicus, starvation experiment 1, st. 232. 
                 a - number of alive specimens, 
                 b - proportion (%) of alive specimens

Fig. 6.1.3 a, b: Calanus finmarchicus, starvation experiment 2, st. 243. 
                 a - number of alive specimens,
                 b - proportion (%) of alive specimens

Fig. 6.1.4 a, b: Calanus finmarchicus, starvation experiment 3, st. 312. 
                 a - number of alive specimens, 
                 b - proportion (%) of alive specimens



6.1.3  Organic carbon content, C/N ratio, stable isotope ratios and dry mass 
       measurements
       Kristina Barz(1),                 (1)Alfredwegener.l nstitut
       Ksenia Kosobokova(2)              (2)Shirshov institute of Oceanology


Adult animals, eggs and copepodite stages II-V of 29 copepod species, and 
eggs and embryos of a chaetognath Eukrohnia hamata were collected for 
measurements of organic carbon content, C/N ratio, dry mass and lipid 
composition. Live specimens of target species were sorted from Bongo net 
catches immediately after collection, pre-sorted to species and stage level 
and kept in filtered sea water for 24 hours.

For carbon content, C/N ratio, stable isotope ratios and dry mass 
measurements, 1 to 5 specimens of large animals (5 - 10 mm) and 20 to 100 
smaller ones (2 - 5 mm) with empty guts where used for each replicate. The 
specimens were shortly rinsed in distilled water and placed in tin caps. For 
dry weight measurements, animals were placed in pre-weighted aluminum caps. 
Whenever possible, 2 - 3 replicates of each stage of each particular species 
were prepared for each analysis. The samples were stored at -80°C.

For lipid composition analyses animals were transferred to 
Dichlormethane-Methanol (2:1) and stored at -20°C.



6.1.4  Reproductive biology of pelagic copepods
       Adrian Basilico(1),                (1)Alfred-Wegener-Institut
       Ksenia Kosobokova(2)               (2)Shirshov institute of Oceanology


Egg production experiments were carried out with the interzonal copepod 
Calanus finmarchicus, C. glacialis, and the mesopelagic Paraeuchaeta 
glacialis, Gaetanus tenuispinus, and G. brevispinus. Clutch size was also 
assessed for the mesopelagic species P. barbata, P. polaris, Chiridius 
obtusifrons, Augaptilus glacialis. and Euaugaptilus hyperboreus.

Calanus finmarchicus and C. glacialis used for egg production experiments 
were sorted from the 0 - 1,500 m Bongo net catches immediately after capture. 
25 - 30 single females of each species were placed in cell wells filled with 
filtered sea-water from the 0 - 100 m water layer. Additionally, 25 - 30 
females were pooled in 150 ml Plexiglas insets with a mesh (300 - 500 µm) 
false bottom to separate eggs from females. These were then suspended in 250 
ml TPX jars containing filtered sea water. Egg production during the first 24 
hours was used as a measure of the actual rate in the field (in-situ egg 
production rate).

Mature females of mesopelagic species were also sorted immediately after 
capture for egg production studies. Single females of the free-spawning 
Gaetanus tenuispinus, and G. brevispinus were placed in cell wells filled 
with pre-screened seawater collected from below 300 m. They were incubated at 
dim light at temperatures between -1.0∞ and + 0.5∞C. The females were checked 
every 12 h for egg production. Eggs were counted and removed; subsequently, 
egg size was measured and egg morphology was studied under a 
stereomicroscope. Females of egg-brooding Paraeuchaeta barbata, P. polaris, 
Chiridius obtusifrons, Augaptilus glacialis, and Euaugaptilus hyperboreus 
were also sorted immediately after capture, and the number of eggs in the egg 
sacs was counted after preservation in 4% borax1buffered formaldehyde. The 
formaldehyde-preserved multi-net catches were additionally used to enumerate 
and measure eggs in the egg sacs of these egg-brooding species. The clutch 
size was estimated as the number of eggs produced during one spawning event 
by a free-spawning or egg-brooding female. Female prosome and total length 
were measured under a stereomicroscope at 25 x magnification.

Egg-bearing females of P. glacialis and females with developing eggs in the 
ovaries were sorted alive and incubated in 250 ml TPX jars at -1.0∞ to + 
0.5∞C to assess the duration of egg development and spawning intervals. 
Females were checked daily for egg production. Hatching nauplii were counted 
and removed.


Preliminary results

Egg production experiments with Calanus glacialis were carried out at 9 
stations. Spawning females were observed on the slope north of Franz Josef 
Land and on the northern Kara Sea shelf only (sts. 268, 271, 279). Egg 
production rate (EPR) was generally low and varied from 0.6 to 4.5 eggs fem-1 
d-1. The only exception was st. 271, where EPR reached 13.5 eggs fem-1 d-1. 
Individual clutch size varied from 16 to 19 eggs fem-1. No egg production was 
observed on the Barents Sea shelf (st. 236) or in the ice-covered deep basins 
(st. 257, 260, 264, 338, 371).

Egg production experiments with C. finmarchicus were carried out at 4 
stations (sts. 243, 268, 271, 279). Egg-laying was observed at the only 
station 271 north of the Franz Josef Land. Average EPR for the pooled 84 
females was 1.95 eggs fem-1 d-1, individual clutch size varied from 21 to 31 
eggs fem-1.

Observations on egg production of mesopelagic copepods were carried out at 18 
stations. In total, successful reproduction was observed in 62 specimens of 8 
species.



6.1.5  LOKI
       Adrian Basilico
       Alfred-Wegener-Institut


LOKI

The newly developed system FLOKI (Flowthrough Onsight Key species 
Investigation) was tested on the cruise for the first time in the field. This 
system produces high-resolution pictures of objects in the size spectrum 
between 200 µm and nearly 2 cm. It is the aim of this system to store images 
of live organisms, which are easier identified than preserved material, and 
thus to facilitate zooplankton counting. FLOKI consists of a digital camera 
with 4 megapixels resolution, which acquires pictures at a rate of 15 
images/sec from particles passing through a glass cuvette of 2*1*0.4 cm 
(length, width, height). The sample is pumped through the cuvette at a rate 
of ca. 100 ml min-. Each object of a certain light intensity (region of 
interest) is cut out and stored separately on a hard disk. In the laboratory, 
the images will be screened and sorted by an image analysis programme.

15 Bongo net (100 µm, 300 µm) samples were processed by FLOKI during the 
cruise. After initial technical problems the instrument worked well, and more 
than half a million pictures were taken and sorted (Fig. 6.1.5). The test 
revealed some technical problems, that will need amendment. Thus, air bubbles 
in and steam on the cuvette were probably the result of too large temperature 
differences between sample and working location of FLOKI. The pumping speed 
will need to be increased for fast swimming large animals as found in the 
Arctic, cuvette size and correspondent tubing should be customisable to 
organism size. The results of the FLOKI counts will be compared with manual 
counts.



6.1.6  Role of small plankton and trophodynamic relationships in the Arctic 
       pelagic food web
       Antoine Nowaczyk 
       Laboratoire d'oceanographie et de Biogeochemie


Among the copepods present in the Arctic Ocean, the cosmopolitan and small
Oithona similis, the Atlantic Calanus finmarchicus, and the Arctic Calanus 
glacia/is, Calanus hyperboreus and Metridia longa are the major objects of 
the study. O. similis is one of the most abundant zooplanktonic organisms in 
all oceans. The other chosen species strongly dominate the zooplankton 
biomass in the Arctic Ocean. In order to quantify their grazing impact on 
lower trophic levels and to study the mechanisms of adaptation of small and 
large-sized organisms to the conditions in the Arctic pelagic ecosystem, 
process studies were performed.

From the 100 µm mesh size Bongo net tows (0 - 100 m), samples were collected 
at 19 stations (Table 6.1), quickly filtered onto GF/F filters, frozen in 
liquid nitrogen then transferred to the -80°C freezer. Instantaneous 
measurements of feeding through pigment gut contents, amino and fatty acids 
will be assessed from these samples at a specific level. Upon return to the 
lab, pigment gut contents will be performed by fluorimetry according to Dam 
and Peterson. Amino acids and fatty acids composition will be measured by gas 
chromatography. The composition of those compounds will also give a good 
indication of the recent feeding history of the individuals studied (most 
common species). The data will be used to assess the grazing impact of each 
species and the daily removal of phytoplankton.

Feeding experiments were run throughout the cruise at 16 stations (Table 
6.1). Live individuals (i.e. moving under the dissecting microscope) of 
Oithona sp., C. finmarchicus, C. glacialis, C. hyperboreus and M. longa were 
transferred to incubation bottles containing water collected at the 
chlorophyll max. Feeding rates will be assessed through measurements of 
pigment composition between control and experimental bottles. HPLC analysis 
will allow the determination of preferential grazing of taxonomical groups of 
phytoplankton (CHEMTAX). Water was also preserved (acid lugol) for 
microscopic analysis.

Excretion experiments were run at 16 stations. In order to measure the 
ammonium and phosphate excretion of the main species of copepods (same as for 
the feeding experiments), individuals were incubated in filtered 0.2 µm sea 
water. At the end of the experiment, samples were frozen at -20°C and will be 
analyzed upon return to the lab using a technicon autoanalyser and low 
phosphate method (MAGIC method: Rimmelin, P. & T. Moutin. 2005. Re-
examination of the MAGIC method to determine low orthophosphate concentration 
in seawater. Analytica Chimica Acta, 548(1-2), 174-182). To estimate the 
diurnal variability of excretion rate, a 48h experiment was carried out at 
station PS 70/379, water sub samples were taken at 2, 5, 9, 13, 17, 21, 24, 
36 and 48 hours.

Gut evacuation rates experiments were carried out at 4 stations in order to 
estimate phytoplankton ingestion rates from gut contents.

Individuals collected at each station as well as those from the experiments 
were measured under a dissecting microscope and frozen at -80°C for 
subsequent measurements of dry weight and C/N/P contents.


Table 6.1: Experiments

Stations    Specific    Excretion    Feeding     Feeding    Abundance   Gut
           Instantane-  experiment  experiment  experiment             evacu-
            ous rates                             large                ation
          Gut contents                            volume                rate
           Amino acid
           Fatty acid
             C/N/P
            contents
--------  ------------  ----------  ----------  ----------  ---------  ------
70/232         x                        x
70/236         x             x
70/243         x             x          x           x
70/257         x             x          x           x
70/260         x             x          x
70/264         x             x          x           x
70/268         x             x          x
70/279         x             x          x           x
70/290         x             x          x
70/301         x             x          x
70/307         x             x          x
70/312         x             x          x           x           x
70/322         x             x                                  x
70/328         x             x          x                       x
70/338         x             x          x           x           x         x
70/352         x             x          x                       x
70/362         x                                                          x
70/371         x                                                          x
70/385         x                                                          x
-----------------------------------------------------------------------------
TOTAL         19             15         15          6           5         4



Reference

Rimmelin, P. & T. Moutin. 2005. Re-examination of the MAGIC method to 
    determine low orthophosphate concentration in seawater. Analytica Chimica 
    Acta, 548(1-2), 174-182.



6.2  Biodiversity of polar deep-sea eukaryotic microbiota - molecular versus 
     morphological approach
     Beatrice Lecroq
     University of Geneva Sciences III


Background

Over the past few years, cultivation-independent identification of microbial 
organisms by PCR amplification and sequencing of ribosomal RNA genes revealed 
a huge diversity of microbiota in environmental samples. Many new species and 
higher-level lineages have been discovered. However, the microbial diversity 
in the polar deepsea benthos is still largely unexplored. Foraminifera and 
the closely related large testate (shell-bearing) gromiids, are a significant 
and often visually conspicuous component of the deep-sea and high-latitude 
benthic fauna. In addition to the geologically important and well-known 
calcareous foraminifera, deep-sea and highlatitude assemblages include 
substantial numbers of soft-shelled, mostly single chambered species, most of 
which are undescribed. Our previous studies of polar Foraminifera revealed 
exceptionally high morphological and molecular diversity of some 
monothalamous (single-chambered) morphospecies, particularly abundant in 
high-latitude settings (Gooday et al., 1996, Pawlowski et at., 2002, 
Pawlowski et al., 2005). On the other hand, we found very weak genetic 
differentiation between some common Arctic and Antarctic deep-sea calcareous 
species (Pawlowski et al., 2007).

Objectives

The main objective of this project was to examine the diversity of microbial 
eukaryotes (protists) in deep-sea Arctic sediments by using environmental DNA 
approach. The project is focus on two important groups of marine protists: 
Foraminifera and Cercozoa. Both groups belong to the recently established 
supergroup of Rhizaria (Nikolaev et al., 2004).

Using material collected during this cruise, we aim to

(1) obtain rDNA sequence data for broad taxon sampling of Arctic Foraminifera 
    and Cercozoa;
(2) establish their phylogenetic relationships with Antarctic and other 
    deep-sea species;
(3) describe new species based on their morphological and genetic 
    characteristics.

Methods

The foraminifera were isolated from surface sediment samples (1 - 2 cm) 
collected by multicorer (usually 1 or 2 cores were examined for each site) 
and boxcorer (about 100 cm2 were collected for each corer). 5 ml of the 
surface sediment was collected and immediately deep frozen in the liquid 
nitrogen. These samples will be used for total DNA extraction and study of 
microbial eukaryote diversity. The rest of sediment was sieved through the 
sieves of 0.5 mm, 0.125 mm and 0.063 mm. The sieving took place in the cooler 
container and the samples were stored at 3°C. The samples were examined at 
dissecting microscopes and all living foraminifera were isolated. The 
isolated foraminifera were identified, photographed with digital camera 
mounted on the microscope and later either immediately processed for DNA 
extraction or frozen in the liquid nitrogen. Some species have been fixed in 
the formalin for further morphological study.

Preliminary Results

During the course of the expedition a total of 30 stations were sampled, 
yielding 608 DNA extracts, 30 samples of frozen sediment for total 
environmental DNA analysis and 31 preserved specimens in formalin (Table 
6.2).

A total of 84 morphospecies regrouped under 50 Genera were identified (Fig. 
6.2.1). Regarding the stations, samples showed great differences in abundance 
and composition of living specimens. Further investigations might be able to 
precise whether the depth could also be involved in those diversities 
differences. The number of specimens sampled, isolated, extracted and fixed 
fully satisfy the quantitative objective of this campaign. This will enable a 
successful comparison of this material with specimens previously sampled in 
other parts of the world.


Tab. 6.2: Summary of the samples collected in each station

                                                   DNA
                      Frozen      Specimens    extractions    
          Stations  sedimental    preserved   in Guanidine   Genera
                     samples    in Formaline  or AP1 buffer
          --------  ----------  ------------  -------------  ------
            239-6        5             0            57         19
            260-5        1             0             1          1
            263-2        4            14            24          8
            265-1        3             0            13          6
            265-2        2             1             9          6
            266-7        2             0            53          3
            271-5        3             0            73         13
            276-7        3             0            16         10
            277-2        3             0            14          7
            279-12       3             0             1          1
            282-2        3             1             7         39
            285-6        2             0             9         20
            285-7        3             0            22         10
            294-2        3             8            22          4
            294-3        2             0             3          2
            299-2        3             0             3          2
            301-11       3             0             1          1
            306-2        2             1            22         10
            309-8        3             6            21          6
            316-5        2             0            14          5
            316-6        2             0            20          8
            319-2        2             0             7         24
            330-1        3             0            71          4
            333-4        3             0            28          4
            338-9        3             0            16          5
            342-10       3             0            11          5
            352-3        2             0            14          7
            358-3        3             0            38         16
            384-1        2             0             9          7
            385-9        2             0             9          5
            Total       80            31           608         50



Fig. 6.2.1: Number of speicmens per Genera and per station (with depths of 
            the stations)



References

Gooday, A.J., Bowser, S.S. and Bernhard, J.M., (1996) Benthic foraminiferal 
    assemblages in Explorers Cove, Antarctica: A shallow-water site with 
    deep-sea characteristics. Prog. Oceanog. 37: 117-166.

Nikolaev, S.I., Berney, C., Fahrni, J., Bolivar, I., Polet, S., Mylnikov, 
    A.P., Aleshin, V.V., Petrov, N.B., Pawlowski, J. (2004) The twilight of 
    Heliozoa and rise of Rhizaria: an emerging supergroup of amoeboid 
    eukaryotes. Proc. Natl. Acad. Sci. USA 101:8066- 8071.

Pawlowski, J., Fahrni, J.F., Brykczynska, U., Habura, A., Bowser, S.S. (2002) 
    Molecular data reveal high taxonomic diversity of allogromiid 
    Foraminifera in Explorers Cove (McMurdo Sound, Antarctica). Polar Biology 
    25: 96-105.

Pawlowski, J., Fahrni, J.F., Guiard, J., Konlan, K., Hardecker, J., Habura, 
    A., Bowser, S.S. (2005) Allogromiid foraminifera and gromiids from under 
    the Ross Ice Shelf: morphological and molecular diversity. Polar Biology 
    28:514-522

Pawlowski J., Fahrni J., Lecroq B., Longet D., Cornelius N., Excoffier L., 
    Cedhagen T., Gooday A.J. (2007) Bipolar gene flow in deep-sea 
    foraminifera. Molecular Ecology. 16:4089-96.



Acknowledgements

The author would like to thank the captain Stefan Schwarze and his crew for 
the optimal working conditions they provided at sea. I am also extremely 
grateful to the Alfred-Wegener-Institut and the members which have supervised 
this cruise. Finally I wish to say a very special thank to Robert Spielhagen, 
Anna Schmidt, Kristin Daniel and Norbert Lensch for their great help in 
corers manipulation and to Antoine Nowaczyk and Charles-Edouard Thuroczy for 
their rich collaboration in the laboratory.





7.  MARINE GEOLOGY
    Kristin Danie(l), Catalina Gebhardt(1),  (1)Alfred-Wegener-Institut
    Norbert Lensch(1), Valery Rusakov(3),    (2)IFM-GEOMAR
    Anna Schmidt(2), Pjotr Semenov(4),       (3)Vernadsky Institut of 
    Robert Spielhagen(2)                        Geochemistry and
                                                Analytical Chemistry
                                             (4)VNIIO All-Russia Research 
                                                Institute for Geology and 
                                                Mineral Resources of the
                                                World Ocean


Background and goals

Paleoenvironmental changes in and around the Arctic Ocean have been 
fundamental in the last 1 million years and especially in the last 200,000 
years. Terrestrial and marine research has revealed a high variability of ice 
sheets on the circum-Arctic continents and equally important changes in the 
ice coverage and Atlantic Water inflow in the Arctic Ocean and the 
river-runoff. Arctic river run-off is of vital importance for the Arctic low 
saline surface water outflow and the formation of the Arctic sea ice cover, 
both a prerequisite for the maintenance of strong oceanographic contrasts in 
the Nordic Seas and the deepwater renewal, which runs the thermohaline 
convection (THC), as well as for the sediment and chemical budgets of the 
Arctic Ocean. While there is some evidence for Holocene variability of 
river-runoff from earlier studies in the southern Kara Sea (SIRRO Project), 
knowledge for changes in the deglaciation, the Weichselian, and the last 
interglacial ("Eemian") are scarce. It was unknown prior to cruise ARK-XXI 
1/2 whether the river run-off across the Kara Sea shelf was blocked by the 
northern Eurasian ice sheet during part of the last glacial maximum (LGM). 
For earlier glacial episodes, the exact glacial limits of ice sheets on the 
Kara Sea are also elusive. The position of the ice front and its variability 
must have exerted enormous influence on the sediment export to the Arctic 
Eurasian Basin (debris flows, turbidites etc.). It was one of the goals to 
retrieve sediment cores from which the glacial history of the northern Kara 
Sea and its major outlets (the St. Anna and Voronin troughs) can be 
reconstructed at utmost high time resolution.

In the last two decades a number of sediment cores had been investigated 
which gave valuable information on the history of the Late Quaternary Arctic 
Ocean, its water masses, and its ice coverage. Most of these cores stemmed 
from the eastern Arctic Ocean (15°W-75°E) and the Lomonosov Ridge relatively 
close to the Pole. Marine geological information from the Siberian sector of 
the eastern Arctic Ocean and from the area beyond the Lomonosov Ridge was 
still rather scarce. The second major goal of cruise ARK-XXII/2 was therefore 
to increase the archive of sediment cores available for paleoenvironmental 
studies of the central Arctic Ocean with special emphasis of the Amerasian 
Basin.



7.1   Parasound sediment echosounding
      Wiebke Nehmiz FIELAX


Scientific objectives, technical settings and operation conditions

One of the fixed sensor installations onboard Polarstern is the sediment 
echosounder PARASOUND (Atlas Hydrographics, Bremen). The system provides 
digital, high resolution information on the sediment coverage and the 
internal structure of the sediments. This can be used to interpret the 
sedimentary environments and their changes in space and time. During 
ARK-XXII/2, the aim of PARASOUND profiling was to select coring locations for 
gravity cores, box cores, multi corer and kastenlot cores.

In order to obtain bottom and sub-bottom reflection patterns, the echosounder 
uses the so-called parametric effect: PARASOUND radiates two primary 
frequencies in the range of 18 to 23.5 kHz that generate a secondary pulse of 
lower frequency which provides the signal. This parametric frequency is the 
difference frequency of the two primary waves transmitted. The parametric 
frequency can be chosen between 2.5 and 5.5 kHz and is adjusted by varying 
the primary frequencies. During this cruise, the primary low frequency was 
set to 18 kHz and the primary high frequency was set to 22 kHz, so the signal 
had a frequency of 4 kHz. Due to its low secondary frequency and a small 
emitting angle of 2 degrees, PARASOUND achieves high resolution of the 
sediment structures and penetrating depths of around 70 meters with a 
possible vertical resolution of ca. 30 cm.

During ARK-XXII/2 only the operation mode Single Pulse was used with the 
registration of the full trace length including the water column. This mode 
transmits a beam and waits to send the next one until having stored the 
reflected signal. The former PAR mode of the DS2-System is now called Pulse 
Train and was not available on this trip yet. The parametric pulse length was 
always set to 2. Partly, heavy ice conditions affected the quality of the 
data negatively causing noisy records with some traces missing.

The PARASOUND system onboard Polarstern

The Atlas Parasound is a permanently installed hull mounted system on 
Polarstern.

This highest level narrow-beam parametric sub-bottom profiler system for deep 
sea applications was upgraded to system version P-70 in May 2007. After the 
sea acceptance test was not passed during the leg ARK-XXII/la, some parts of 
the system were not running stable, but nevertheless useful sediment 
information to take geological samples were acquired.

The hardware and software configuration using two Windows-based PCs for 
system control and data management, respectively, offers a lot of functions 
via both Operator and Data Storage PC. Data recording is done by the software 
ATLAS PARASTORE-3 (Atlas Hydrographics, Bremen). This software provides a 
user-friendly graphical interface and has been designed to acquire, 
visualize, process, store, convert, quality control, replay and print data 
from the profiles of the system. One additional PC is set up in the 
"Windenleitstand" for watch keeping purposes and as a backup system.

During ARK-XXII/2, this was used once for data storage during a break down of 
the other two PCs.

Additionally to the features that were used on this cruise, the new system 
and the PARASTORE-3 software offer many more possibilities to adjust 
parameters and operate in different modes of sounding. Most of these are not 
used for standard procedures. On this cruise only settings were used that are 
comparable to the old system to be able to reproduce the standard of former 
cruises and to be able to compare the data sets acquired with the systems of 
different age.

Due to problems with the online-print function of the ATLAS PARASTORE-3 
software (usually the software is supposed to be able to acquire, store and 
print the data simultaneously, but until now the online-print function has 
caused a complete crash of the software if the other two functions are used 
at the same time), the final visualisation of the PS3 data was done with the 
software SeNT (Hanno von Lom, Universität Bremen). The graphics were saved in 
the GIF format and printed out later than the data was acquired.

Data Management

The capabilities of ATLAS PARASTORE-3 in acquiring the PARASOUND data are 
two-fold: (i) simultaneously as raw data from the parametric channel and the 
NBS channel for full trace length including the water column, and (ii) 
selected in specific depth windows, in PS3 and SEG-Y formats. The recorded 
data of both parametric and NBS signals are stored in a hybrid raw data 
format (ASD format) in either a ring buffer or into user selected folders on 
hard disc. Either full soundings of both signals or defined depth windows can 
be acquired and stored. Optionally, the parametric data can be converted 
online into standard SEG-Y format for further processing using industrial 
post-processing software. In addition, the parametric data can be stored in 
the formerly used PS3 format. This function, however, requires careful watch 
keeping because only data kept in a selected depth window of commonly 200 m 
length is stored.

On this cruise, the data was stored as ASDF and PS3 formats for all time 
periods the system was running and data was stored. In addition, the replay 
option of PARASTORE-3 gives the opportunity to replay the soundings of 
parametric data and/or NBS data and record them again as PS3 and/or SEG-Y 
data.

Because all the data can be produced and stored simultaneously, a 
well-organized data management is necessary. In order to avoid confusion with 
the different formats and the status of the data, the data management was 
carried out by one designated person.

The data storage was organized in such a way that the parametric raw data 
(.asd files) and the subsequent data set of PS3 data (.ps3) were packed in 
four hour packages of the UNIX based TAR format (.tar files) and saved to the 
ships intern server space as well as the visualised graphics (.gif). For 
backup reasons, all changes to the server the resulting data collection is 
saved to are simultaneously copied to tape. Finally, LTO-tapes and a set of 
30 DVDs with the complete data set of ASD, PS3 and GIF files were sent to AWI 
directly after the cruise. In addition to this, meta-files were produced to 
ease the input of these files into the data base PANGAEA.



7.2   Bathymetry

Precise knowledge of the seafloor morphology is a basic requirement for 
geological work. Bathymetric data combined with parasound profiles allow a 
three-dimensional reconstruction of history of the upper sediment layers. 
Therefore, bathymetric data were recorded along with parasound data in the 
same areas.

During the ARK-XXII/2 expedition, depth measurements were carried out using 
ATLAS Hydrosweep DS-2, a deep-sea multibeam echo sounding system installed 
permanently on Polarstern. The hull-mounted Hydrosweep DS-2 system operates 
at a frequency of 15.5 kHz transmitting sound pulses perpendicular to the 
ship's long axis. In addition to depth measurements, echo amplitudes can be 
recorded and converted into multibeam sidescan and angular backscatter data.

Hydrosweep was operated with a swath width of 90°. Accuracy of the depth 
measurements is mainly dependent on precise sound velocity profiles of the 
water column. Therefore, CTD measurements carried out by the oceanography 
group were used for calibration (21 CTD stations in total). Only before the 
first stations (i.e. Barents Sea), the Hydrosweep system was calibrated by 
comparison of slant and vertical beams of sweep profiles perpendicular to the 
ship's long axis with profiles from the ship's longitudinal direction.

The recorded depth and positioning data were stored in 8 hour intervals in a 
sensorindependent raw data format (SURF). Post-processing such as cleaning of 
erroneous position and depth data was not carried out onboard Polarstern.


7.3   Geological sampling

Geological sampling was performed in two areas on the northern Kara Sea 
continental margin, along a transect from the Kara Sea across the Gakkel 
Ridge and the Lomonosov Ridge (at 88°40'N, 140°E) to the Alpha Ridge, and on 
a shorter transect from the Alpha Ridge transect back to the Lomonosov Ridge 
at 86°30'N, 152°E (Fig. 7.1). Coring sites were selected based on Parasound 
results, with emphasis on morphological highs in the central and eastern 
Arctic Ocean (Lomonosov and Alpha Ridges). In the two working areas on the 
Kara Sea continental margin, one continuous N-S coring transect was performed 
in each area, covering water depths from the shelf (ca. 300 m) to the 
continental foot (ca. 4,000 m).


Fig. 7.1: Bathymetric map of the working area. Coring positions are indicated 
          by black dots and the station number (add PS70/ for full number). 
          Referenced sites PS2185 and 96/12 are shown as white dots.


Surface-near sediments were sampled at 24 geological sampling sites with a 
large volume grab (box corer, 50 cm x 50 cm x 50 cm). Except for the very 
first station PS701261, where the technical problems prevented successful 
coring, the box corer always provided excellent results with recoveries of 
0.33 - 0.50 cm of undisturbed surface-near sediments. After opening of the 
cores they were photographed. A detailed description of the surface and 
geological profile was followed by sampling for various purposes, including 
the recovery of 3 archive tubes and 1 - 3 archive boxes which preserved the 
full sequence of deposits in the corer box. Additional samples were taken 
downcore for X-ray analyses and dry bulk density determinations. All samples 
were stored at 4°C.

In addition to the box corer, a multicorer with 12 plastic tubes of 8 cm 
diameter was used at 7 selected stations to retrieve sediments with an 
absolutely undisturbed water-sediment interface. The very first and the very 
last multicorer casts (stations /239 and /385) were performed without other 
geological sampling tools, mainly for microbiological sampling purposes. 
Sampling was done as 1 cm-slices. Samples were frozen (-18°C) or stored at 
4°C. At station /352 a microcorer (technically similar to a multicorer) with 4 
plastic tubes was used, hanging under the CTD/rosette. It recovered 21 cm of 
sediment which were handled as described above.

To obtain long sediment cores at selected stations, a gravity corer (SL) of 5 
or 10 m length and 12 cm diameter with a penetration weight of 1.5 t or a 
square-barrel kasten lot corer (30 x 30 x 1,200 cm) with a penetration weight 
of ca. 3 t were used. At station /294 both systems were run. In total, 12 
gravity cores (length 0.50 - 5.93 m) and 4 kastenlot cores (length 2.93 - 
7.70 m) were recovered (Fig. 7.2). A characteristic lithological boundary in 
the two cores from station /294 (at 282 and 322 cm core depth in the gravity 
and kastenlot core, respectively) allows to determine a sediment compression 
of ca. 15% in the gravity core at this site, if compared to the kastenlot 
core.


Fig. 7.2: Core recovery of gravity and kastenlot cores.


All cores were opened on board, photographed, visually described, and sampled 
for X-ray analyses, dry bulk density determinations, and other purposes. Core 
descriptions are given in the annex. From the kastenlot cores, several 1 m 
long archive boxes were taken in parallel from along the entire core and 
frozen or stored at 4°C. In addition to the visual description, the sediment 
colour was determined at 1 cm intervals using a Minolta CM-2002 
spectrophotometer (see next chapter).

All sediment cores were be stored cool on board and transported to the AWI 
core depository in Bremerhaven from where they will be made available for 
investigations.



7.4   Physical properties

Multi-Sensor Core Logging

Non-destructive whole-core physical properties provide initial core 
characterization with a very high vertical resolution and are commonly used 
for lateral correlation of cores. Physical property measurements were carried 
out onboard using a Multi-Sensor-Core-logger (MSCL, GEOTEK Ltd., UK). P-wave 
travel times, magnetic susceptibility, g-ray-absorption and non-contact 
electrical resistivity were measured simultaneously including control 
measurements of core diameter and sediment temperature. From these data, the 
physical properties wet bulk density, magnetic susceptibility, fractional 
porosity, P-wave velocity, and formation factor can be calculated using the 
MSCL software package. The technical description of the system is given in 
Table 7.1:


Tab. 7.1: Technical specifications of the GEOTEK MSCLI4


          P-wave velocity and core diameter

                     Plate transducer diameter: 4 cm
                     Transmitter pulse frequency: 250 kHz
                     Pulse repetition rate: 1 kHz
                     Received pulse resolution: 50 ns
                     Gate: 5000
                     Delay: 0 µs


          Density

                     Gamma ray source: Cs-137 (1983)
                     Activity: 356 MBq
                     Energy: 0.662 MeV
                     Collimator diameter: 5.0 mm for gravity and box cores,
                     2.5 for Kastenlot cores
                     Gamma detector: Gammasearch2, Model SD302D, Ser.
                     Nr. 3043, John Count Scientific Ltd.,
                     10 s counting time


          Fractional porosity

                     Mineral grain density = 2.65, water density = 1.026
                     

          Magnetic susceptibility 

                     Loop sensor: BARTINGTON MS-2C, Ser. Nr. 208 
                     Loop sensor diameter: 14 cm 
                     Point sensor: BARTINGTON MS-2F, Ser. Nr. 139 
                     Alternating field frequency: 565 Hz, counting time 10 s, 
                     precision 0.1 * 10-5 (SI) 
                     Magnetic field intensity: ca. 80 Nm RMS 
                     Krel: 1.56 (SL, 12 cm core-ø), 0.60 (KAL, cross section 
                     50.27 cm2, = 8 cm core-ø) 
                     Loop sensor correction coefficient: 6.391 (SL) and 
                     16.689 (KAL) for 10-6 (SI), respectively 
                     Point sensor: BARTINGTON MS-2F, Ser. No. 139 
                     counting time 10 s



Gravity and box cores were measured in coring liners with 12.5 cm diameter, 
whereas Kastenlot cores were measured in sub-cores retrieved from the 
original core using length-wise open transparent plastic boxes of 1,000 mm 
length and 75 times 75 mm cross section. In order to allow for 
magnetic-susceptibility sensor correction according to the Bartington 
correction requirements, the rectangular cross section of the box was 
equalized to a size-equivalent circular section, of which a fictive core 
diameter was calculated as input parameter for loop-sensor correction 
coefficient (Tab. xx). For both core sizes (gravity/box and Kastenlot cores), 
the density calibration was carried out using a set of defined mixtures of 
aluminium and distilled water in a gravity liner and Kastenlot sub-sampling 
box, respectively (Best and Gunn, 1999).

In addition, measurements of magnetic susceptibility using a Bartington point 
sensor were performed directly on the sediment surface. Gravity cores were 
measured shortly after splitting. For box core point sensor magnetic 
susceptibility measurements, a second sub-core was taken in a transparent 
plastic box with open surface. MSCL and point sensor data of the box cores 
thus origin of different subcores. For Kastenlot cores, the measurements were 
carried out on the same subcore as MSCL measurements. For practical reasons, 
the point sensor was hooked up on the GEOTEK Color-line-Scan-logger. The step 
intervals of the measurements were 1.0 cm.


Spectrophotometer measurements

Quantitative estimates of spectral reflectance and sediment colour were 
carried out using a handheld Minolta CM-2002 spectrophotometer (Minolta 
Camera Co., Osaka, Japan). A precise documentation of the camera settings is 
given in Balsam et al. (1998). The split core surfaces of the archive halves 
were covered by a standard film to protect the spectrophotometer. Calibration 
and operation of the spectrophotometer were done according to the Minolta 
CM-2002 users' manual (Minolta Camera Co., 1991). Measurement spacing was 
generally set to 1 cm.


Preliminary results

Parasound surveys and data interpretation

The first PARASOUND profile was run continuously from the northern Barents 
Sea north into the Nansen Basin and then south into the first marine 
geological working are on the NE Kara Sea continental margin (St. Anna 
Trough). The system had to be turned off for transit to the second working 
are and was switched on again at ca. 83°E on the NW Kara sea continental 
margin (Voronin Trough). From there on the system worked continuously again 
on the way to the Gakkal, Lomonosov, and Alpha Ridges and back to the Gakkel 
Ridge. The Parasound survey ended at ca. 79°N in the Laptev Sea. In the 
following, a short description of the Parasound profiles and the 
characteristics of the deposits will be given, as imaged from the PARASOUND 
data.


Barents Sea

The first, more southern part of the northward profile in the Barents Sea is 
characterized by a smooth topography with a distinct hard reflector at the 
surface. Only in a few locations, a penetration deeper than 10 m could be 
recognized. The northern part, close to the continental margin, displays a 
rough glacial topography with steep troughs and walls. It is interpreted to 
reflect the erosional activity of the Barents Sea ice sheet which had covered 
the area during the last glacial maximum (cf. Svendsen et al., 2004). Data 
recovery on the steep slope was poor in many depth intervals. Otherwise there 
is evidence for levee structures and a strong activity of slumps and slides.


The Nansen, Amundsen, and Makarov Basins

Sediments in the basins are generally flat lying. Penetration was 20 - 40 m. 
During earlier expeditions (ARK-IV/3, ARK-VIII/3) the long sediment cores 
obtained from the basins revealed the presence of thick turbidite sequences, 
intercalated with thinner layers from more pelagic sedimentation.


Northern Kara Sea continental margin

Sediments on the lower slope NW of the St. Anna Trough and N of the Voronin 
Trough up to ca. 2,500 m depth usually show 10 - 20 m of good penetration. 
The images are characterized either by a single or double layer signature. 
Below the uppermost, ca. 3 - 4 m layer without a clear internal structure, an 
acoustically more transparent layer can be observed. Its thickness is 
sometimes only 1 - 2, but can increase in lense-shaped bodies to more than 20 
m. These bodies are interpreted as slides. Below the transparent layer, a 
second more reflective layer of a few meters in thickness can often be 
observed. The succession is tentatively interpreted as intercalated slides 
and layers of pelagic sedimentation.

Northwest of the St. Anna Trough the slope between 2,500 m and 1,200 m is 
extremely steep and does not allow a characterization from the records. North 
of the Voronin Trough this depth interval also shows evidence of slides, but 
especially between 1,800 and 2,800 m the reflective layers are thicker (5 - 
15 m) and show a distinct internal layering (Fig. 7.3).

In both areas, profiles up to ca. 650 m showed an undulated morphology and 
often evidence for large- and small-scale slide structures (Fig. 7.4). 
Penetration was variable, but often the reflective top layer of 4 - 8 
captured most of the energy and it is difficult to analyze the deeper 
structure.

Above 650 m, a very rough small-scale topography is usually dominating. It 
partly consists of channels incising 5 - 15 m into the otherwise more 
undulated surface, but in other areas the entire profile is hummocky and no 
horizontal layering is visible. The first type is interpreted as the result 
of erosional effects during times of lower sea level, whereas the second type 
was probably formed by an overriding ice sheet, followed by severe iceberg 
ploughing.


Fig. 7.4: PARASOUND-Profile of sediment layers at the upper slope of the St. 
          Anna Trough. Note the undulated surface of sedimentary layers.

Fig. 7.3: PARASOUND-Profile of sediment layers at the lower slope of the 
          Voronin Trough. Note the layering and the transparent lenses in 
          certain parts of the profile.


Gakkel Ridge

PARASOUND records from the Gakkel Ridge revealed an extremely rough 
topography. Only in a few pockets between the volcanoes of this active 
mid-ocean spreading ridge small subbasins could be found where well-layered 
sediments have accumulated. In general, the data recovery in the Gakkel Ridge 
are was very poor.


Lomonosov Ridge

Both transects across the Lomonosov Ridge gave similar results as during 
earlier crossings (ARK-VIII/3, ARK-XIV/la). While the steep slopes did only 
rarely allow an analysis of the underground, the upper slope and the ridge 
crest showed the well-layered structure of the uppermost part of the 
sedimentary sequence. A special feature within the ridge is the interior 
basin at ca. 88°30'-89°N and 140 - 180°E. The cruise track crossed the basin 
in its southern part and a profile from the ridge crest at ca. 1,300 m into 
the basin at ca. 2,700 m water depth was recorded. The flanks of the basin 
were very steep and data quality was poor. However, in the basin, the records 
reveal horizontally layered sediments. These observations fully confirm the 
results of Björk et al. (in press).


Fig. 7.5: PAR4SOUND-Profile of sediment layers from the crest of the 
          Lomonosov Ridge into the interior basin of the ridge. Note the 
          distinct horizontal layering in the basin.


Alpha Ridge

Independent of the water depth which reached <1,500 m in the shallowest 
regions, the topography of the Alpha Ridge was generally rough and the 
surfaces were undulated (Figs. 7.6 and 7.7). Well-layered sediments of a 
wider regional or local extent (i.e., a few kilometre) were never observed. 
Almost all PARASOUND records from the Alpha Ridge show abundant side echoes 
which prohibited a detailed analysis of the subsurface layers. From these 
shipboard data it seems difficult to analyze the internal structure of the 
upper sedimentary layers.


Fig. 7.6: PARASOUND-Profile of sediment layers on the lower Alpha Ridge. Note 
          the irregular surface caused by side echoes.

Fig. 7.7: PARASOUND-Profile of sediment layers on the Alpha Ridge. Note the 
          irregular surface caused by side echoes.


Sediment characteristics and paleoenvironmental interpretations

Coring operations during ARK-XXII/2 recovered only soft sediments and there 
is no obvious evidence for pre-Quaternary deposits. Based on the sediment 
descriptions and the different lithologies found in the cores, a subdivision 
of 4 sedimentary provinces seems possible (Kara Sea continental margin, 
Gakkel Ridge, Lomonosov Ridge, and Alpha Ridge). The box cores from the 
central Nansen and Amundsen Basins (stations /299, /301, and /309) held 
sequences of mostly sandy silty clays, as observed earlier during other 
expeditions (e.g., ARK-VIII/3). Typically, these sequences are underlain by a 
sandy layer which has been determined to be ca. 50 - 60 ky old 
(Nørgaard-Pedersen et al., 1998, 2003).


Kara Sea continental margin

12 box cores, 6 gravity cores, and 2 kastenlot cores were obtained from the 
Kara sea continental margins at water depths between 4,000 and 300 m. The 
uppermost few decimetres of sediment usually consist of dark yellowish brown 
silty clays and sandy silty clays, underlain at sediment depths of 30 - 40 cm 
by more olive greyish sandy silty clays. The sediment surfaces often contain 
calcareous micro- and macroorganisms (incl. bivalve fragments and benthic 
foraminifers) and the sediments are homogenized or strongly mottled from 
bioturbation, but sporadic sieving tests of subsurface sediments rarely 
revealed microfossils in the coarse fraction.

Grain sizes in the gravity and kastenlot cores typically range from silty 
clays to sandy silts. The deposits have mainly greyish olive colours of 
variable lightness, but dark grey and olive brown layers also occur. 
Especially the dark gray (colour codes N4 and N3) and dark olive gray (5Y411 
and 5Y311) sediments are characteristic features. In all 6 cores from the 
Kara Sea margin which were longer than 3 meters, the dark greyish deposits 
were found in the bottom part of the cores. The maximum recovered thickness 
of this ubiquitous layer was 290 cm in kastenlot core PS70/294-5, and in none 
of the cores this layer was fully penetrated. It often contained 
iceberg-rafted dropstones in variable amounts and in cores from the 
continental margin off the Voronin Trough also coal particles of up to 5 cm. 
One or two similar dark grey layers also occur higher up in the cores, but 
their thickness is in most cores restricted to 10 - 50 cm. In kastenlot core 
PS70/294-5 the dark grey layer at ca. 150 - 250 cm consists of silty clay 
with sandy laminae, indicating a redistribution of sand particles by 
turbidity or contour currents. The dropstone content clearly marks the dark 
grey layers as of glacial origin. Since the cores stem from the margin of the 
two major glacial outlets of the Kara Sea, it is tempting to correlate the 
dark grey layers with the glaciation episodes of northern Eurasia as revealed 
by terrestrial and marine investigations in the last decade (e.g., Svendsen 
et al., 2004; Spielhagen et al., 2004). According to this correlation the 
dark greyish layers in the long cores were deposited at ca. 50 - 60 ka, 80 - 
90 ka, and 130 - 180 ka. In most cases the thick dark grey layer at the 
bottom of the cores would thus stem from the Saalian glaciation which 
probably had the most extensive ice sheet on the Siberian shelves and in 
northern Eurasia ever in the Quaternary. Following this interpretation, the 
more olive and olive brownish to brownish layers in the sediment cores which 
are often homogeneous or strongly mottled from bioturbation represent the 
climatically warmer and intermediate periods in the Late Quaternary. Detailed 
investigations will be performed in the home laboratories to establish a 
high-resolution age model for the cores and reconstruct the 
paleoenvironmental history of the Kara Sea continental margin, the adjacent 
Arctic Ocean, and the hinterland.


Gakkel Ridge

At site PS70/306 (ca. 4,000 m water depth) a box core and a 3.79 m long 
gravity core were retrieved. With very few exceptions the sediments consisted 
of sandy silty clay and sandy silt. The sequence of sediments is similar to 
that from the Kara Sea continental slope. Intercalated among dark brownish 
and olive deposits it contains three dark grey (N4) layers which probably can 
be correlated to those in cores from the margin, implying an iceberg 
transport of shelf-derived sediments to the Gakkel Ridge area during times of 
continental glaciations. The lowermost dark grey layer is underlain by olive 
brownish deposits. If the above described correlation holds true, the bottom 
sediments in PS70/306-2 may be ca. 200 ky old.


Lomonosov Ridge

Three box cores, one gravity core, and two kastenlot cores were retrieved 
from the eastern sector of the Lomonosov Ridge between 86°30' and 88°40'N. 
Sites PS70/316 and /319 are only 50 km apart. While site /316 is situated 
slightly below the ridge crest (1,300 m), site /319 is in a small interior 
basin within the Lomonosov Ridge (2,700 m deep), which is connected both with 
the Makarov and Amundsen Basins, allowing a water mass exchange above the 
1,900 m depth level (cf. Björk et al., in press). The sediments consist 
mostly of sandy silty clays which are strongly mottled, indicating a high 
degree of bioturbation. Lithological correlation seems difficult based on 
sediment colours and grain sizes, possibly because of winnowing effects on 
the ridge top. Preliminary investigations of the microfossil content in 
PS70/319-3 revealed uninterrupted occurrences of planktic foraminifers within 
the uppermost 75 cm. Since the deposits in nearby core PS2185 from 87°32'N 
144°E have a distinct foraminifer-barren interval below 45 cm (50 ka, 
Spielhagen et al., 2004), it can be concluded that the time resolution in the 
interior basin sediments is by a factor of 1.5 or more higher than on the top 
of the ridge, possibly due to trapping effects of sediments winnowed from the 
crest.

At site PS70/358 the longest core obtained during cruise ARK-XXII/2 was 
recovered from near the crest of the Lomonosov Ridge at 86°32'E. It holds a 
sequence of mostly sandy silty clays which are well-known from other sites on 
the ridge (e.g., 96/12, PS2185). The lithological similarity in this area 
suggests that the age models of nearby cores can be tentatively transferred 
to PS70/358-4. This method would give an age of ca. 200 ka at ca 250 cm and a 
bottom age of the core near 1 My. The strongly mottled or homogeneous colours 
of the sediment are evidence for a high activity of burrowing organisms. 
According to Jakobsson et al. (2000) the various colour changes from brown to 
olive brown reflect a cyclic variability of manganese oxide contents which 
are thought to be related to changes in intermediate and bottom water 
ventilation caused by oceanographic and climate changes in the Arctic.


Alpha Ridge

Four box cores and four gravity cores (up to 5.82 m) were retrieved from the 
morphologically very heterogeneous area of the Alpha Ridge. The water depths 
of the coring sites lie between 1,500 and 3,300 m. The lithology of the cores 
was rather monotonous; they consisted almost entirely of sandy silty clays 
with very little changes in grain size. Sediment colours ranged from dark 
brown to olive brown. Almost all layers were strongly mottled and sometimes 
no basic colour could be determined ("sploshy colours"). According to the 
interpretation of Jakobsson et al. (2000) for similar sediments on the 
Lomonosov Ridge, the colour variability may be interpreted as evidence of 
changes in manganese oxide contents, which in turn reflect climate 
variability.

At site PS70/338 both the box core and the gravity core contained a layer of 
carbonate-cemented worm tubes a few centimetre below the sediment surface. 
The layer was hard and ca. 3 - 4 cm thick. Washed pieces were whitish grey in 
colour and up to 10 cm in size. Two further hard layers at 40 and 75 cm in 
the gravity core were of similar thickness, but consisted of strongly 
burrowed, stiff sandy silty clay of the similar colours as the soft sediments 
above and beneath. The origin of these hardground layers remains yet elusive



References

Balsam W. L., Deaton B. C., & Damuth J. E., 1998. The effects of water 
    content on diffuse reflectance spectrophotometry studies of deep-sea 
    sediment cores. Marine Geology 149, 177-189.

Best, A. I. and D. E. Gunn, 1999. Calibration of marine sediment core loggers 
    for quantitative acoustic impedance studies. Marine Geology 160, 137-146.

Björk, G., Jakobsson, M., Rudels, B., Swift, J.H., Anderson, L., Darby,. 
    D.A., Backman, J., Coakley, B., Winsor, P., Polyak, L., and Edwards, M. 
    (in press). Bathymetry and deep-water exchange across the central 
    Lomonosov Ridge at 88°-89°N. Deep-Sea Res.

Jakobsson, M., Løvlie, R., Al-Hanbali, H., Arnold, E., Backman, J., Mörth, 
    M., 2000. Manganese and colour cycles in the Arctic Ocean sediments 
    constrain Pleistocene chronology. Geology 28, 23-26.

Nørgaard-Pedersen, N., Spielhagen, R. F., Thiede, J., Kassens, R., 1998. 
    Central Arctic surface ocean environment during the past 80,000 years. 
    Paleoceanography 13, 193-204.

Nørgaard-Pedersen, N., Spielhagen, R. F., Erlenkeuser, H., Grootes, P. M., 
    Heinemeier, J., Knies, J., 2003. The Arctic Ocean during the Last Glacial 
    Maximum: Atlantic and Polar domains of surface water mass distribution 
    and ice cover. Paleoceanography.

Spielhagen, R.F., Baumann, K.-H., Erlenkeuser, H., Nowaczyk, N.R., Nørgaard
    Pedersen, N., Vogt, C., Weiel, D., 2004. Arctic Ocean deep-sea record of 
    Northern Eurasian ice sheet history. Quat. Sci. Rev., 23 (11-13): 
    1455-1483.

Svendsen, J.I., Alexanderson, H., Astakhov, V.1., Demidov, I., Dowdeswell, 
    J.A., Funder, S., Gataullin, V., Henriksen, M., Hjort, C., 
    Houmark-Nielsen, M., Hubberten, H.W., IngOlfsson, 0., Jakobsson, M., Kjr, 
    K.H., Larsen, E., Lokrantz, H., Lunkka, J.P., Lysâ, A., Mangerud, J., 
    Matioushkov, A., Murray, A., Möller, P., Niessen, F., Nikolskaya, 0., 
    Polyak, L., Saarnisto, M., Siegert, C., Siegert, M.J., Spielhagen, R.F. 
    and Stein, R., 2004. Late Quaternary ice sheet history of northern 
    Eurasia. Quat. Sci. Rev., 23(11-13): 1229-1271.





APPENDIX
A.1 PARTICIPATING INSTITUTIONS
A.2 CRUISE PARTICIPANTS
A.3 SHIP'S CREW
A.4 STATION LIST (see pdf)
A.5 ANNEX CORING POSITIONS (see pdf)
A.6 SEDIMENT CORE DESCRIPTIONS (see pdf)





A.1  PARTICIPATING INSTITUTIONS

Adresse / Address

AWI           Alfred-Wegener-Institut
              für Polar- und Meeresforschung
              Am Handelshafen 12
              27570 Bremerhaven / Germany

DWD           Deutscher Wetterdienst Hamburg
              Abteilung Seeschifffahrt
              Bernhard-Nocht Str. 76
              20359 Hamburg / Germany

FAZ           Frankfurter Allgemeine Zeitung GmbH
              Wissenschaftsredaktion
              Hellerhofstraße 2-4
              60 327 Frankfurt / Germany

FIELAX        FIELAX
              Gesellschaft für wissenschaftliche
              Datenverarbeitung mbH
              Schifferstraße 10-14
              27568 Bremerhaven
              Germany

FIMR          Finnish Institut of Marine Research (FIMR),
              Finnland
              Erik Palménin aukio 1,
              P.O. Box 2,
              00561 Helsinki / Finnland

GEOKHI        Vernadsky Institute of Geochemistry and
              Analytical Chemistry,
              Russian Academy of Sciences,
              19, Kosygin Str, 117975 Moscow / Russia

Heli Service  Heli Service International GmbH
              Im Geisbaum 2
              63329 Egelsbach
              Germany

IFM-GEOMAR    Leibniz-Institut für Meereswissenschaften
              IFM-GEOMAR
              Düsternbrooker Weg 20
              24105 Kiel / Germany

IFM HH        Universität Hamburg, Fakultät 6, Fachbereich
              Geowissenschaften, Institut für Meereskunde
              Bundesstraße 53,
              20146 Hamburg / Germany

IPÖ           Institut für Polarökologie
              Wischhofstr. 1-3, Geb. 12,
              24148 Kiel / Germany

JAMSTEC       Institute of Observational Research for Global
              Change, Japan Agency for Marine-Earth
              Science and Technology (JAMSTEC)
              2-15, Natsushima-cho, Yokosuka
              Kanagawa, 237-0061 / Japan

Laeisz        Reederei F. Laeisz (Bremerhaven) GmbH
              Brückenstrasse 25
              27568 Bremerhaven / Germany

LOB           Laboratoire d'Oceanographie et de
              Biogeochemie (LOB) UMR 6535
              Centre d'Oceanologie de Marseille Campus
              de Luminy, Case 901 13288
              Marseille Cedex 09 / France

NIOZ          Royal Netherlands Institute for Sea Research
              PO box 59,
              1751 AB Den Burg, Texel / The Netherlands

OPTIMARE      OPTIMARE Sensorsysteme AG
              Am Luneort 15a
              27572 Bremerhaven / Germany

SIO           P.P. Shirshov Institute of Oceanology
              Russian Academy of Science, Russia
              36 Nachimovsky prospect,
              Moscow, 117851 / Russia

Texas A&M     Texas A&M University at Galveston
              5007 Avenue U, Galveston,
              Texas 77551 / USA

UAB           Institut de Ciència i Tecnologia Ambientals
              Universitat Autànoma de Barcelona
              08193 Bellaterra / Spain

UCD           School of Physics, University College of
              Dublin, Belfield,
              Dublin 4 / Ireland

UG            Molecular Systems Group,
              Department of Zoology an Animal Biology,
              Universität von Genf
              30, Quai Ernest Ansermet,
              CH-1211 Genf 4 / Switzerland

VNIIO         VNIIO All-Russia Research Institute for
              Geology and Mineral Resources of the World
              Ocean VNII Okeangeologiya
              1, Angliysky ave.,
              190121 St.Petersburg / Russia

XU            XU Research Center for Environmental
              Science
              Xiamen University
              Xiamen 361005 / China





A.2  CRUISE PARTICIPANTS

Name               Vorname/         Institut/     Beruf/ 
                   First Name       Institute     Profession
-----------------  ---------------  ------------  -----------------------
Bakker             Karel            NIOZ          Geochemist
Barz               Kristina         AWI           Biologist
Basilico           Adrian           AWI           Biologist
Boom               Lorendz          NIOZ          Tech nician/Geotraces
Breier             Florian                        Journalist
Büchner            Jürgen           Heli Service  Pilot
Cai                Pinghe           XU            Geochemist
Camara             Patricia         UAB           Geochemist
Damm               Ellen            AWI           Geochemist
Daniel             Kristin          AWI           Student/Geology
Gebauer            Manfred          DWD           Meteorologist
Gebhardt           Catalina         AWI           Geologist
Gremlowski         Lars             AWI           Geochemist
Heckmann           Hans H.          Heli Service  Pilot
Heckmann           Markus           Heli Service  Technician
Hendricks          Stefan           AWI           Geophysicist
Kern               Stefan           IfMH          Geophysicist
Kiko               Rainer           IPÖ           Biologist
Kikuchi            Takashi          JAM STEC      Oceanographer
Klunder            Maarten          NIOZ          Geochemist
Kosobokova         Ksenia           SIO           Biologist
Kramer             Maike            IPÖ           Biologist
Kuckero            Lilith           AWI           Student/Geotraces
Laan               Patrick          NIOZ          Geochemist
Lechtenfeld        Oliver           AWI           Geochemist
Lecroq             Beatrice         UG            Biologist
Leinweber          Volker           AWI           Student/Sea ice phys
Lensch             Norbert          AWI           Technician/Geology
Lepore             Kate             UCD           Geochemist
Mechler            Sebastian        Optimare      Engineer/Oceanogr
Mertineit          Sabine           AWI           Geochemist
Middag             Rob              NIOZ          Geochemist
Nehmiz             Wiebke           FIELAX        Geologist
Nowazyk            Antoine          COM           Biologist
Ober               Sven             NIOZ          Technician/Geotraces
Pisarev            Sergey           SIO           Oceanographer
Rabe               Benjamin         AWI           Oceanographer
Rabenstein         Lasse            AWI           Geophysicist
Rudels             Bert             FIMR          Oceanographer
Rusakov            Valery           GEOKHI        Geologist
Rutgers v d Loeff  Michiel          AWI           Geochemist
Schauer            Ursula           AWI           Chief Scientist
Schmidt            Anna             IFM-GEOMAR    Geologist
Schneider          Alice            IPÖ           Biologist
Semenov            Pjotr            VNIIO         Geologist
Siebert            Stefan           IPÖ           Biologist
Sonnabend          Hartmut          DWD           Meteorol. Technician
Spielhagen         Robert           IFM-GEOMAR    Geologist
Spreen             Gunnar           lfM HH        Geophysicist
Stimac             Michael          Heli Service  Technician
Thuroczy           Charles-Edouard  NIOZ          Geochemist
Vöge               Ingrid           AWI           Technician/Geochemistry
Walker             Sally            Texas A&M     Student/Biochemistry
Winderlich         Andreas          lfM HH        Geophysicist
Wisotzki           Andreas          AWI           Tech nician/Oceanogr





A.3  SHIP'S CREW

No.  Name                   Rank
---  ---------------------  ----------
01.  Schwarze, Stefan       Master
02.  Spielke, Steffen       1.Offc.
03.  Farysch, Bernd         Ch. Eng.
04.  Fallei, Holger         2.Offc./L.
05.  Wunderlich, Thomas     2.Offc.
06.  Kaufmann, Tino         2.Offc.
07.  Schuhardt              Doctor
08.  Hecht, Andreas         R.Offc.
09.  Minzlaff, Hans-Ulrich  2.Eng.
10.  Schäfer, Marc          3.Eng.
11.  Sümnicht, Stefan       3.Eng.
12.  Scholz, Manfred        Elec Eng.
13.  Nasis, Ilias           Electron.
14.  Verhoeven, Roger       Electron.
15.  Muhle, Helmut          Elec.Tech
16.  Himmel, Frank          Electron.
17.  Loidl, Reiner          Boatsw.
18.  Reise, Lutz            Carpenter
19.  Rhau, Lars-Peter       A. B.
20.  Stutz, Heinz-Werner    A. B.
21.  Winkler, Michael       A. B.
22.  Reichert, Jörg         A. B.
23.  Hagemann, Manfred      A. B.
24.  Schmidt, Uwe           A. B.
25.  Bäcker, Andreas        A. B.
26.  Wende, Uwe             A. B.
27.  Preußner, Järg         A.B.
28.  Ipsen, Michael         Storek.
29.  Voy, Bernd             Mot-man
30.  Elsner, Klaus          Mot-man
31.  Hartmann, Ernst-Uwe    Mot-man
32.  Pinske, Lutz           Mot-man
33.  Müller-Homburg, R.-D.  Cook
34.  Silinski, Frank        Cooksmate
35.  Martens, Michael       Cooksmate
36.  Jürgens, Monika        Stwdess
37.  Wöckener, Martina      1.Stwdss/Kr
38.  Czyborra, Bärbel       2.Stwdess
39.  Silinski, Carmen       2.Stwdess.
40.  Gaude, Hans-Jürgen     2.Steward
41.  Huang, Wu-Mei          2.Steward
42.  Möller, Wolfgang       2.Stwdard
43.  Yu,Kwok Yuen           Laundrym.











Die "Berichte zur Polar- und Meeresforschung"

(ISSN 1866-3192) werden beginnend mit dem Heft Nr. 377 (2000) in Fortsetzung 
der früheren "Berichte zur Polarforschung (Heft 1-376, von 1982 bis 2000; 
ISSN 0176 - 5027) herausgegeben. Ein Verzeichnis aller Hefte beider Reihen 
befindet sich im Internet in der Ablage des electronic Information Center des 
AWI (ePIC) unter der Adresse http://epic.awi.de. Man wähle auf der rechten 
Seite des Fensters "Reports on Polar- and Marine Research". Dann kommt eine 
Liste der Publikationen und ihrer online-Verfügbarkeit in alphabetischer 
Reihenfolge (nach Autoren) innerhalb der absteigenden chronologischen 
Reihenfolge der Jahrgänge.

To generate a list of all 'Reports'past issues, use the following URL: 
http://epic.awi.de and select the right frame: Browse. Click on "Reports on 
Polar and Marine Research". A chronological list in declining order, author 
names alphabetical, will be produced. If available, pdf files will be shown 
for open access download.


Verzeichnis der zuletzt erschienenen Hefte:

Heft-Nr. 567/2007 - "Effects of UV Radiation on Antarctic Benthic Algae - 
With Emphasis on Early Successional Stages and Communities", by Katharina 
Zacher.

Heft-Nr. 568/2007 - "The Expedition ANTARKTIS-XXIII/2 of the Research Vessel 
'Polarstern' in 2005/2006", edited by Volker Strass.

Heft-Nr. 569/2008 - "The Expedition ANTARKTIS-XXIII/8 of the Research Vessel 
'Polarstern' in 2006/20067", edited by Julian Gutt.

Heft-Nr. 570/2008 - "The Expedition ARKTIS-XXI/1 a and b of the Research 
Vessel 'Polarstern' in 2005", edited by Gereon Budéus, Eberhard Fahrbach and 
Peter Lemke.

Heft-Nr. 570/2008 - "The Expedition ARKTIS-XXI/1 a and b of the Research 
Vessel 'Polarstern' in 2005", edited by Gereon Budéus, Eberhard Fahrbach and 
Peter Lemke.

Heft-Nr. 571/2008 - "The Antarctic ecosystem of Potter Cove, King-George 
Island (Isla 25 de Mayo). Synopsis of research performed 1999-2006 at the 
Dallmann Laboratory and Jubany Station", edited by Christian Wiencke, Gustavo 
A. Ferreyra, Doris Abele and Sergio Marenssi.

Heft-Nr. 57212008 - "Climatic and hydrographic variability in the late 
Holocene Skagerrak as deduced from benthic foraminiferal proxies", by Sylvia 
Brückner.

Heft-Nr. 573/2008 - "Reactions on surfaces of frozen water: Importance of 
surface reactions for the distribution of reactive compounds in the 
atmosphere", by Hans-Werner Jacobi.

Heft-Nr. 574/2008 - "The South Atlantic Expedition ANT-XXIII/5 of the 
Research Vessel 'Polarstern' in 2006", edited by Wilfried Jokat.

Heft-Nr. 575/2008 - "The Expedition ANTARKTIS-XXIII/10 of the Research Vessel 
'Polarstern' in 2007", edited by Andreas Macke.

Heft-Nr. 576/2008 - "The 6th Annual Arctic Coastal Dynamics (ACD) Workshop, 
October 22-26, 2006, Groningen, Netherlands", edited by Pier Paul Overduin 
and Nicole Couture.

Heft-Nr. 577/2008 - "Korrelation von Gravimetrie und Bathymetrie zur 
geologischen Interpretation der Eltanin-Impaktstruktur im Südpazifik", von 
Ralf Krocker.

Heft-Nr. 578/2008 - "Benthic organic carbon fluxes in the Southern Ocean: 
regional differences and links to surface primary production and carbon 
export", by Oliver Sachs.

Heft-Nr. 579/2008 - "The Expedition ARKTIS-XXII/2 of the Research Vessel 
'Polarstern' in 2007", edited by Ursula Schauer.







                        ARK-XXII/2 Carbon Measurements

                               Steven van Heuven
                          Dept. of Ocean Ecosystems
                 Faculty of Mathematics and Natural Sciences
                           University of Groningen
                              The Netherlands

                                  Sven Ober
                 Royal Netherlands Institute for Sea Research
                               The Netherlands

                             October 31st, 2008



Abstract

This document outlines the procedures followed in collection and analysis for 
DIC and TAlk of seawater samples during Polarstern cruise ARKXXII/2 and the 
subsequent processing of the results. Some of the various considerations 
during data processing are discussed, as well as the basic workings of the 
MATLAB-routines that were programmed specifically to facilitate rapid but 
thorough and well-documented processing of the measurements. Scripts are 
included verbatim, as are several figures and an extensive table containing 
raw and processed results. Briefly, sampling was performed according to 
D0E1994, but analysis follows D0E1994 less stringent. An assessment of data 
quality is presented at the end of this document. Although the measurement 
system was not absolutely calibrated, resulting data is very well correctable 
through the use of CRM, after which data is deemed to be of reasonably high 
quality.


1  Metadata

Name of cruise: ARKXXII/2 Research vessel: F/S Polarstern Time: 28 July 
(Tromsoe) to 10 October 2007 (Bremerhaven) Working area: Eurasian Arctic 
Parameters: Total alkalinity (TAlk) and dissolved inorganic carbon (DIC) 
Analyst: Sven Ober (Roya1NIOZ, The Netherlands) Analyzer: VINDTA #14 (owned 
by NIOZ) Data processing: Steven van Heuven (University of Groningen, The 
Netherlands) # of samples analyzed: 453 # of stations sampled: 40 # of CRM 
analyses: 140 (70 bottles two subsequent runs from each bottle).



2  Analytical methods

2.1  Sampling procedure

Samples were collected in 250m1 borosilicate bottles following D0E1994. 
Samples were poisoned with HgC1, closed with ground glass stoppers, using 
grease and rubber bands and stored dark and cold until analysis. All samples 
were analyzed on board, within 8 weeks after sampling. A correction for the 
diluting effect of adding HgCl is performed, but the headspace equilibration 
correction (∆D) is not, because the atmospheric concentration of CO2 in the 
rosette tapping room was not known. Since headspace volume was consistently 
very low (~1% of sample volume), no influence on DIG of more than .5 µmol/kg 
is expected.


2.2  Analyzer description

Samples were analyzed on a VINDTA 3C (Versatile Instrument for Determination 
of Titration Alkalinity) type 3C, developed and built by Dr. Ludger Mintrop, 
MARIANDA, Kiel, Germany. This device concurrently performs a potentiometric 
alkalinity titration and a coulometric DIC titration (see DOE1994 and 
references therein). Calculations performed by the VINDTA at time of 
measurement were considered to be preliminary, because of a lack of accurate 
salinity and nutrient values which are required for these calculations. 
Storage of all raw titration results allowed for post-cruise recalculation of 
all results.


2.3 Analysis procedure

Samples were allowed to reach analysis temperature by being placed in a 
waterbath of 25.0°C for ~1 hour. Sample bottle stoppers were removed 
immediately prior to insertion of the sampling line to the machine. Samples 
were drawn towards the machine by a peristaltic pump. Although a slight 
underpressure is generated in the sampling line in this way, no excessive 
bubble formation (dissolved gas being 'pulled out of solution') was observed. 
Sample was first used to rinse and fill the DIC pipette, and after that to 
rinse and fill the TAlk pipette. Care was taken to avoid sample carry-over 
(by separating subsequent samples by a small volume of air). Measurements 
were not performed around the clock but mainly during daytime. In order to 
set the measurement accuracy Certified Reference Material was analyzed at 
least twice during a measurements day: after the coulometric cell was 
successfully started (i.e., after a suitable number of dummy runs were 
performed and coulometer blank level was stable), and shortly before the end 
of the day. Occasionally CRM was analyzed halfway the day as well.


2.4 DIC-specific remarks

DIC determinations were performed using the standard coulometric method using 
a UIC model 5011 coulometer. No current-to-frequency calibration has been 
performed either pre- or post-cruise and an inaccuracy may be present in the 
coulometric data, up to 0.5% (i.e., 10 µmol/kg) in either direction. Also no 
highly accurate determination of pipette volume has been performed but 
pipette is known to have been 19.5±0.1 ml (accuracy thus ~.5% i.e., ~10 
µmol/kg). These errors are expected to have been very constant during the 
course of the cruise, and both have a linear effect on the measured values 
and are therefore exactly correctable by the use of CRM. No gas-loop 
calibration was used.


2.5 TAlk-specific remarks

TAlk determinations were performed by an acid titration that combines aspects 
from both the commonly used 'closed cell' method and the 'open cell' method, 
following the VINDTA's standard settings. Batches of acid were carefully 
prepared to be 0.1M, but no determinations of the exact acid strengths were 
performed (error assumed to be <0.25% i.e., ~6 µmol/kg). Also, no highly 
accurate determination of pipette volume has been performed, but pipette is 
known to have been 97.0±0.5 ml (accuracy ~0.5% i.e. ~12µmol/kg). Burette 
(on a Metrohm Titrino model 719) volume was determined post cruise to be 
exactly 5.00ml to within 0.1% (i.e., ~0.2µmol/kg). New electrodes from 
Metrohm (reference) and Orion (measurement) were used but no formal 
assessment of their quality (E°, Nernst response) was performed(1). 
Significant  and unquantifiable inaccuracies may therefore be present in the 
final dataset.












____________________________
(1) Please note that even when correcting values to CRM-results, a 
    non-Nernstian electrode will yield inaccurate values on samples with TAlk 
    values different from those of the CRM!


3  Post-cruise data processing

A suite of flexible MATLAB scripts was developed to assist in thorough though 
rapid processing of CO2-system data gathered during oceanographic 
expeditions. The scripts as they are detailed here are a customized version 
(specific to the particular format of the data gathered on this cruise) of 
the more general scripts. The following subsection will explain the basic 
workings of each routine

3.1  vindta222_loadvindtafile

This scripts loads the dataset output by the VINDTA CO2-analyzer. This 
contains such variables as runname, runtime, the sample's station cast and 
depth values, CRM batch numbers, etc. The DIC and TAlk values calculated by 
the VINDTA at the end of analysis runs are loaded as well although these will 
later on be overwritten with values recalculated from the raw coulometer and 
titration results and the final salinity and nutrients data.

3.2  vindta222_loadbottlefile

This scripts simply loads a cruise's bottlefile which in this case formatted 
as an ODV spreadsheet. For each oceanographic sample that was collected 
during the cruise station-, cast- and bottle numbers and values of salinity, 
phosphate and silicate are stored in memory.

3.3  vindta222_loadlogfile

This script loads the coulometric logfile generated by the VINDTA's software, 
associated with DIG related actions performed by the VINDTA. It contains all 
coulometer readings that were acquired during the cruise and some associated 
values (runname, time, etc.). This information was used to calculate the 
coulometer's blank level (and standarddeviation of that) for each run, by 
averaging the coulometer increments that were recorded after the carbon peak 
had been titrated away. To be sure none of the titration peak was 
accidentally taken into this calculation averaging was performed for minutes 
6 through 14 (which was always the end of each DIC run roughly coinciding 
with the end of the TAlk determination of the same sample). For each run, the 
calculated values for blank and blank-std are stored.

3.4  vindta222_findrawtitfiles

This scripts browses through the VINDTA's output directories, storing names 
and dates of these files. This information will be used to match the raw 
titration data to the data that was collected earlier in the process.

3.5  vindta22_2matchbottledata

This scripts matches to each of the CO2-system analyses performed on the 
VINDTA, the associated chemical and physical parameters (that were loaded in 
the previous step) by comparing station- castand bottlenumber between the two 
datasets. After this step, accurate salinity, phosphate, silicate, are thus 
associated with the VINDTA runs.

3.6  vindta222_matchblanks

Somewhat analogous to the previous step, this scripts matches the information 
about per-run coulometer blank level (collected in the first step), to the 
VINDTA runs.

3.7  vindta222_matchrawtitdata

Again analogous to earlier steps, this scripts matches the raw potentiometric 
titration results to the VINDTA runs, this time on the basis of the timestamp 
that is known for each run, assuming that the raw results were be written 
approximately 20 minutes after the start of an analysis. If the software 
cannot make a definitive match, user input confirmation is required.

3.8  vindta222_setbaseline

Plotting the calculated per-run baseline (see 4.3) against time, clearly 
reveals both trends and noise in the coulometer's per-run blank value. Some 
of this noise is of course associated with the stochasticity of the 
per-minute increments. It was determined preferable to set the per-run level 
manually, smoothing out that variability, but remaining close the actually 
observed vales. Note that this is different from the standard method of 
setting a blank level prior to doing the actual runs, in which case the set 
blank may not correspond closely to the actual values observed during the 
day. This scripts plots observed blank levels (and blankstd) against time and 
allows the user to graphically define a line that overrules the observed 
values. This level is allowed to gradually in- or decrease with time if the 
per-run blank level strongly suggests this to reflect the actual coulometer 
behavior (as is for instance visible at the third measurement series. This 
step does require some understanding of (or rather experience with) the 
behavior of coulometric systems. DIC-results are flagged as juestionable%hen 
they are obtained in a run that had a baseline that was more than one 
standard deviation away from the user-set baseline or that was higher than a 
certain threshold (set to l60cpm) or that had a standard deviation higher 
that a certain threshold (set to 80cpm). Please refer to figure 1 for an 
overview of the observed blanks, and the user-set blank that were used for 
calculations.

3.9  vindta222_recalculatedic

Since the original calculations of DIC, performed by the VINDTA upon 
completing each analysis, are invalid since the user manually set the 
baseline levels (see 4.8) a recalculation is required. Pipette volume, 
electronic calibration of the coulometer and the gas loop calibration are set 
in this script all potentially as a (stepwise) function of time. The script 
subsequently performs the DIC calculation using pipette volume, sample 
density, rawcounts, blank level, run duration and coulometer- and gas loop 
calibration.

3.10  vindta222_recalculatetalk

For each run raw titration data is read from disk. Now that for each run all 
information required to precisely (re-)calculate TAlk is available (salinity, 
pipette volume, sample temperature, PO4, Si, electrode readings acid volumes, 
acid salinity, acid concentration, etc.), a recalculation of TAlk is 
performed using the VINDTACALCALK routine, which is not covered here. The 
output of these routines is stored.

3.11  vindta222_flagbadcrms

Although care is taken to waste as little CRM as possible, these runs do 
occasionally fail. If the failure is obvious (caused by analyzer breakdown 
for instance) a rerun may be performed, but not-so-obvious runs may end up as 
valid in the dataset. This scripts graphically display results from all CRM 
runs (as deviations from the certified values) vs. time. Slow drifts through 
time are easily observed this way (for instance caused by the slow weakening 
of titration acid will slowly alter the TAlk's deviation). Obvious outliers 
are assumed not to represent the actual long term-deviation of the analytical 
setup but rather result from problems with sample handling. The script allows 
the user to graphically indicate these runs, upon which they will be flagged 
as badänd not subsequently used.

3.12  vindta222_correctcrmsdic

This script is probably the one that is of most influence to the final DIC 
results. The user is presented a figure displaying the CRM deviations (that 
is, the ratio between the CRM measurement and its certified value) versus 
time, three subsequent measurement periods (i.e., three days) at a time. This 
gives a quick overview of the changing response of the setup through time.

Because CRM's were only run early and late in the day (occasionally halfway, 
too), the changing response cannot be traced exactly, but will have to be 
interpolated between the 2 (or 3) known values. Various formal ways of doing 
this exist, from simply averaging the two (or three) value and assuming no 
change during the day, through linear interpolations, to complicated curve 
fits taken the average behavior of the system into account. This routine is 
potentially flexible in that it allow the user to define any curve he can 
come up with. Of course, the user must try to follow his expert's judgement 
when define this curve. To avoid goal-seeking by the user, no feedback on the 
proposed correction of the final DIC values is given.

After a curve is satisfactorily defined for a particular day, the figure 
shifts sideways and input for the next period is asked. See figure (2) for 
the final result for this particular dataset. User-defined correction curves 
are subsequently used to calculate the final DIC values.

3.13  vindta222_correct crmstalk

This script is identical to the previous one, except that it works on the 
TAlk data.

3.14  vindta222_flagbadsamples

This script presents the user with two figures displaying the final values of 
DIC (or TAlk) vs. depth profiles of all samples that have not already been 
flagged as bad: Obvious outliers can then be selected by the user and be 
flagged as 6adç which removes them from the figure, which makes for a highly 
convenient way of flagging data

3.15  vindta222_updatebottlefile

This script appends the final DIC and TAlk values, their quality flags and a 
lot of associated data (correction factors, TAlk calculation details, etc.) 
to the already existing cruise dataset ('bottlefile'). This file is easily 
imported into programs such as Microsoft's Excel or ODV for further analysis 
or graphing of the data.




4  Analytical quality assessment

Because not all components of the system were accurately calibrated a 
significant systematic error was expected to present in the results. On top 
of that medium-term drift in the VINDTA's response were expected, for 
instance due to changing coulometric cell chemistry or variations in room 
temperature over the course of a day. To determine (and correct for) these 
accuracy errors, CRM was analyzed. During the course of the cruise, two 
subsequent samples were drawn and analyzed from each of a total of 70 bottles 
of CRM. Running two analyses from one bottle also allows for an assessment of 
the the precision of measurements by means of calculating the standard 
deviation of all differences between these duplicates(2).

4.1 CRM DIC results: accuracy

See figure 2. Results of analyses of CRM were consistently (only) around 
0.25% (i.e., ~5µmol/kg) higher than the certified value. Apparently, errors 
in pipette volume calibration and coulometer current-to-counts calibration 
were smaller than feared or cancelled each other out to some extent. There 
were, however, significant trends during days (up to ~5µmol/kg) and a less 
pronounced day-to-day variation. Both short- and medium term variation are 
considered to have been quite captured through the use of CRM and corrected 
for to a final accuracy of ~2µmol/kg.

4.2 CRM DIC results: precision

See figure 3. Precision for DIC defined as the standard deviation of 
differences between duplicate analyses of CRM(3), is ~1.5µmol/kg.

NOTE: All DIC data from runs performed on the following days should be 
mistrusted due to high coulometer noise (see figure 1): 10-14 17-22 (first 
half of the day) 26, 28, 30 and 31 of August. All other days the coulometer 
displayed regular behavior and data may be considered accurate and precise.

4.3 CRM TAlk results: accuracy

See figure 2. Results for CRM can be seen to vary quite smoothly through time 
(except during measurement days 26/08 and 29/08, which display unexplainable 
high drifts and off-trend values, respectively). Within-day drift of up to 
~4µmol/kg are also clearly visible. Also shown in the figure are the 
correction factors applied to the oceanographic samples. Instrument drift is 
considered to have been corrected by use of CRM to within 0.1%-0.2%, i.e., to 
within ~2-4µmol/kg.

4.4 CRM TAlk results: precision

See figure 4. Precision for TAlk defined as the standard deviation of 
differences between duplicate analyses of CRM(4), is ~1.5µmol/kg.












_______________________________
(2) Admittedly, for DIC, duplicate analyses from a single bottle is not an 
    optimal solution due to potential gas exchange between the CRM and the 
    opened headspace in the 20 minutes that it takes to run the first sample.
    However, this effect seems to be of limited size as evident from the 
    histogram, which suggests an average loss of CO2 to the headspace of 
    ~1µmol/kg during those 20 minutes. It needs to be pointed out that any 
    potential variation in this rate of outgassing will lead to an 
    overestimation of the DIC measurement precision.
(3) For DIC, 19 obviously bad measurements were excluded. 56 usable pairs of 
    duplicates remain.
(4) For TAlk, 2 obviously bad measurements were excluded. 68 usable pairs of 
    duplicates remain.


Figure 1: Timeseries of per-run- and user-set-coulometer blank levels.

Figure 2: Timeseries of TAlk and DIC CRM deviations and user-set correction 
          factors for samples.

Figure 3: Histogram of differences between duplicate DIC analyses of the same 
          CRM.

Figure 4: Histogram of differences between duplicate TAlk analyses of the 
          same CRM.






CCHDO Data Processing Notes

Date        Person       Data Type   Action          Summary
----------  -----------  ----------  --------------  -----------------------------------
2012-07-23  Bob Key      BTL/CrsRpt  Submitted       Do not post until further notice 
            I do not have CTD data for this cruise. 
            Most of the cruise report is English 

2012-07-26  Bob Key      CrsRpt      Submitted       Carbon Data Report to go online 
            text on carbon measurements 

2012-09-11  Bob Key      BTL         Submitted       expanded parameters 
            As discussed via phone, this is a GEOTRACES file and can be 
            offered in "as recieved". The file I sent previously includes 
            only normal WOCE/CLIVAR variables.

            This merged file was provided to me by S.van Heuven on 8/22/12 

2012-09-17  CCHDO Staff  PDF         Website Update  Available under 'Files as received' ˚
            The following files are now available online under 'Files as 
            received', unprocessed by the CCHDO.
            ARKXXII2_CARBON_DATA_REPORT.pdf 

2012-09-17  CCHDO Staff  BTL/CrsRpt  Website Update  Available under 'Files as received' 
            The following files are now available online under 'Files as 
            received', unprocessed by the CCHDO.
            Sch2008ae.pdf
            06AQ20070728.exc.csv

2012-10-09  Bob Key      BTL         Submitted       to go online 
            As discussed via phone, this is a GEOTRACES file and can be 
            offered in "as recieved". The file I sent previously includes 
            only normal WOCE/CLIVAR variables.

            This merged file was provided to me by S.van Heuven on 8/22/12 

2012-10-10  CCHDO Staff  BTL         Website Update  Available under 'Files as received' 
            The following files are now available online under 'Files as 
            received', unprocessed by the CCHDO.
            ark222MAP.pdf
            Merged_Data_2012-08-20-12-12-42.csv 

2012-11-16  Jerry Kappa  CrsRpt      Processed       Fianl PDF ready to go online
            I've placed 1 new version of the cruise report:  
              06AQ20070728do.pdf
            into the directory 
              co2clivar/arctic/arkXXII2/arkXXII2_06AQ20070728/
            It includes the original AWI report, the carbon data report 
            submitted on 2012-07-26, summary pages and CCHDO data processing 
            notes as well as a linked Table of Contents and links to figures, 
            tables and appendices.

2014-04-24  Jerry Kappa  CrsRpt      Website Update  Fianl txt version online
            I've placed 1 new version of the cruise report:  
              06AQ20070728do.txt
            into the directory 
              co2clivar/arctic/arkXXII2/arkXXII2_06AQ20070728/
            It includes the original AWI report, the carbon data report 
            submitted on 2012-07-26, summary pages and CCHDO data processing 
            notes.
