SECTION 1: CRUISE SUMMARY

PRE-WOCE ISS01, RRS DISCOVERY CRUISE 164
SEASOAR AND CTD SECTIONS IN THE SOUTHWEST INDIAN AND SOUTHERN 
OCEANS FROM 22S TO 52S

Expedition Designation (EXPOCODE) 74DI164_1

Chief Scientist:		Raymond T. Pollard
				IOS, Deacon Laboratory, UK
				Now at
				School of Ocean and Earth Sciences,
				Southampton Oceanography Centre,
				University of Southampton,
				Empress Dock,
				Southampton, SO14 3ZH, UK
				E-mail: rtp@soc.soton.ac.uk

Ship: RRS Discovery owned and operated by the Natural Environment 
Research Council, UK.

Ports of Call: Port Louis, Mauritius to Port Louis, Mauritius.

Cruise Dates: 19 December 1986 to 21 January 1987.

CRUISE OVERVIEW

Cruise Track
	The  station locations are available in the accompanying
summary file.

Number of Stations
	A total of 61 CTD/Rosette stations were occupied employing a 
12 place 2 litre  Niskin Bottle Rosette with a Neil Brown CTD.

Sampling
	Water samples measurements were made for salinity and 
oxygen. Although CTD data were carefully reconciled with the 
sample values the latter are no longer available.

XBTs
	39 Deep Blue XBTs were deployed during the cruise. Their 
fate is unknown.


		Table 1 CRUISE PERSONNEL & TASKS

Pollard, Raymond T.	IOS	Principal Scientist
Diddams, Paul D.	IOS	XBTs
Goy, Keith M.		IOS	XBTs
Griffiths, Gwyn		IOS	ADCP
Grohman, Dave		IOS	Winches
Hooker, Nigel J.	IOS	CTD
King, Brian A.		IOS	Hydrography/Navigation
Moorey, John A.		IOS	Salinity/Oxygen
Read, Jane F.		IOS	Hydrography
Smithers, John		IOS	CTD/Rosette
Stirling, Moragh W.	IOS	Biology
Wild, Roy A.		IOS	Winches
Brook, Andrew J.	RVS	Computing
Lewis, Derek		RVS	Computing
May, Stephen J.		UCNW	Biology

IOS	Institute of Oceanographic Sciences, Wormley UK
RVS	Research Vessel Services, Barry UK
UCNW	University College of North Wales, Bangor UK


SECTION 2: SCIENTIFIC PROGRAMME

1.	To document the incidence and nature of frontal structures 
in the Southern Ocean. Can the Subantarctic front be identified on 
all transects?  How do upper ocean properties change across each 
front? Are the Polar, Subantarctic and Subtropical Fronts the only 
ones?
 
2.	To document downstream (zonal) variations in the structure 
of the Antarctic Circumpolar Current (ACC).  Is the ACC banded in 
mid-0cean as it is in the Drake Passage, with geostrophic currents 
concentrated in frontal zones?

3.	To estimate the transport of the ACC.

4.	To observe spatial variations in the T/S properties of mode 
waters and compare the observations with theories of mode water 
formation.

5.	To explore the potential of the SeaSoar and Acoustic Doppler 
Profiler to quantify the scales and meridional transport of 
mesoscale eddies in the Subantarctic Zone.


SECTION 3: UNDERWAY MEASUREMENTS

A) NAVIGATION
	Only GPS and transit satellites could provide absolute 
position fixes in the area of operation.  The track plot was 
therefore computed in the standard way perfected on Discovery, 
using the two-component EM log to interpolate between satellite 
fixes, assuming a constant current between each pair of transit 
fixes.  The EM log had been calibrated on Cruise 162 (Pollard, 
Swallow and Saunders), and these recent calibrations were used on 
Cruise 164, namely

	Misalignment angle

	of EM log = 1.7deg clockwise from ship's head

	vFA(tru) = 0.1955 + 0.93145 vFA(est) (knots)

	vSP(tru) = 0.015 + 0.962 vSP(est) (knots)

Transit satellite fixes were transferred from the Level C to the 
PDP11/34 via magnetic tape, eliminating all duplicate and suspect 
fixes.  These were further culled to eliminate all fixes with 
elevations less than 10deg or greater than 70deg, with more than 3 or 
occasionally 4 iterations, and closer together than about one 
hour.  About 400 fixes remained over the 33 day cruise, an average 
of about one fix every two hours.  The currents calculated from 
these fixes and the EM log DR can be seen in Fig. 3 of Pollard et 
al. (1987).

The GPS system provided position fixes for between 3 and 5 hours 
per day throughout the cruise, typically for periods of one to two 
hours duration.  Once per minute during these periods, fixes were 
typed on a dedicated printer and time, latitude, longitude and 
number of space vehicles were logged onto the ship's level A/B/C 
system.

The GPS user has some control over how the system chooses which 
space vehicles (SVs) are used to calculate position:

(1)	The minimum angle of elevation of a good SV.  This was kept 
at 10deg throughout the cruise.

(2)	The maximum acceptable value of PDOP (Position Dilution of 
Precision) for a group of SVs, which is a function of the relative 
positions of the SVs.  The default value for this is 7.0.  
However, a value of 15 was used for most of the cruise and the 
fixes obtained when PDOP was greater that 7.0 were not obviously 
worse that when PDOP was less than 7.0.

(3)	At certain times during the cruise, four or more SVs were 
visible.  If permitted, the system will calculate a horizontal and 
vertical position from four SVs.  However, the resulting fixes had 
a high value of PDOP (which may or may not matter).  When four SVs 
were available, the receiver was programmed to calculate a 2-D fix 
from the best combination of three SVs; the choice of SVs was made 
by the system.

It is recommended that when a group of SVs comes into view, the 
system is initialised with a position correct to within half a 
degree.  After initial experimentation, the receiver was left 
unattended throughout the day and periods of GPS fixes were 
initialised with the previous GPS fix, which was up to 13 hours 
earlier and 1.5 degrees distant.  This method of operation seems 
to have been quite adequate.

Accuracy  A preliminary analysis of the position fixes shows that 
even with values of PDOP up to 15, differences between GPS and 
SATNAV/DR were usually less than 500m, which is the limiting 
accuracy of the transit fixes themselves.

Note  After a mains power failure, the GPS receiver had to be 
reset to the required mode of operation.  Some fixes were lost 
before it was discovered that the receiver had not recovered to 
its state before the loss of power.

Navigation data resides at the British Oceanographic Data Centre 
(BODC).

(Brook, King, Pollard)


B) SEASOAR
	Seasoar - towed yoyo CTD measurements were made between the 
surface and a nominal 400m depth along 3000km of track. These data 
can be recovered from the British Oceanographic Data Centre (BODC) 
and information concerning them can be found in Pollard et al. 
(1987).

C) ADCP MEASUREMENTS
	These were early days for the RD Acoustic Doppler Current 
Measurements (ADCP) and considering this and given the lack of 
precision in the navigation, the data has large error estimates.
See Pollard et al. (1987) for further details. 
	  
D) DEPTH MEASUREMENTS
	The Precision Echo-Sounder was run throughout the cruise. 
Depths every 6 minutes were logged and corrected using Carter 
tables.  The data resides at the British Oceanographic Data Centre 
(BODC).

E) XBT OBSERVATIONS
	See the section on Cruise Overview.


SECTION 4: STATION MEASUREMENTS - CTDS

A CTD station list can be seen in Table 1 of Pollard et al. (1987) 
and in the Summary file.  All casts except 11399 and 11400 were 
made with the IOS Neil Brown Instrument Systems "New Deep CTD", 
and were made to full ocean depth.  A transponder attached to the 
CTD frame was used to make casts to within 20m of the bottom 
whenever a good bottom echo could be seen.

The earth connection was found to be faulty and the cable 
termination was remade prior to station 11418.  The oxygen sensor 
failed for station 11430.  For all other stations it was found to 
drift way off calibration, but reasonable oxygen profiles could be 
recovered using oxygen samples.  Only once did a station have to 
be delayed with the vessel hove to in rough weather, prior to 
11450, when winds up to 45kt were recorded.

On station 11452, it was found that the wire on the midships winch 
did not lay properly at about 4600m, and some time was lost on 
that and subsequent casts trying to achieve a perfect lay.

The conductivity cell failed after cast 11458 and had to be 
replaced.

All casts were logged on a Digidata tape deck interfaced to the 
NBIS deck unit and displayed in real time on a BBC micro computer 
system.  Indeed, for the first ten casts no other data recording 
was possible, because source code for the CTD Level A computers 
was not on board so the CTD data cycle definition could not be 
modified.  This was later patched, and data were averaged to one 
second values by a Level A micro computer, and transferred to both 
the Level C and PDP11/34 computers.  It proved useful to have all 
three routes, as two out of three paths failed on a number of 
occasions, when either the Digidata tape was inadvertently not 
started, or the logging program to the PDP timed out, or the Level 
C failed.  Consequently, no data were lost.
(Smithers, Hooker, Brook)

CTD CALIBRATION
	The 12 bottle multisampler was used to collect samples for 
calibration on all CTD casts.  On occasion, the multisampler does 
not trip the sample bottle, but this can be detected by the CTD 
recovering rather quickly from the firing signal.  Thermometer 
frames were placed on bottles one and four, to keep a check on the 
NBIS calibration. The CTD reads high by a few millidegrees 
compared to thermometers throughout the cruise, so was taken to be 
correct.

Salinities were drawn at 12 levels for the first six deep casts.  
The first-guess CTD calibration appeared to be stable, and about 
0.050psu too low at all depths, so salinities were only drawn at 
three levels thereafter, to keep the number of samples to be 
analysed to reasonable levels.  The Guildline salinometer did not 
function properly from the start of the cruise.  The fault was 
diagnosed to be a faulty cell, and the Autolab salinometer had to 
be used for the rest of the cruise.  Because it is not as 
repeatable as the Guildline (when working properly), triplicate 
samples were drawn at each level.  If the duplicate samples proved 
inconsistent, the triplicate was analysed to decide the matter.

The Beckman oxygen sensor on the NBIS CTD is known to be unstable 
and hard to calibrate.  For this reason, 9 to 12 samples were 
drawn on casts 11401-15.  However, it became apparent that 
reagents for oxygen titration would run out, as significantly more 
CTD casts were being made than originally planned.  Oxygen samples 
had therefore to be drawn from a restricted number of levels, 
namely the bottom, 2500m, oxygen minimum, oxygen maximum, 
thermocline and near surface, six in all.  These proved barely 
adequate to fit the exponential temperature and pressure 
coefficients, as the sensor calibration drifted wildly during the 
course of the cruise.  Further details will be given in the CTD 
data report (Pollard, Read and Smithers, 1987).

Two shallow casts were made at the start of the cruise to provide 
an approximate calibration for the conductivity ration of the 
shallow CTD to be used in the SeaSoar.  In the absence of daily 
deep CTD casts during the week long SeaSoar runs, high priority 
was given to surface salinity samples drawn half hourly from the 
non-toxic supply.  These were entered on the PSTAR computer system 
and differenced from 6m SeaSoar values after careful correction of 
the latter for obvious offsets in the T/S relation.  It was found 
that the SeaSoar was within 0.03psu of the samples, with long term 
drifts (over days) or order 0.01psu, which can be corrected later.  
The technique is thus a satisfactory way of maintaining the 
salinity calibration within 0.01psu.
							(Moorey, Read, Smithers)


REFERENCES

Pollard, R.T., et al. 1987 RRS Discovery Cruise 164, 19 December 
1986 - 21 January 1987. SeaSoar and CTD sections in the Southwest 
Indian and Southern Oceans from 22 S to 52 S. Institute of 
Oceanographic Sciences, Cruise Report, No. 191, 31pp.

Pollard, R.T., Read, J.F., and Smithers, J. 1987 CTD sections 
across the Southwest Indian Ocean and Antarctic Circumpolar 
Current in southern summer 1986/7. Institute of Oceanographic 
Sciences, Report, No. 243, 161pp.



