DIMES James Cook Cruise JC041 Underway Meteorology and Surface Hydrography Document

Cruise details

The JC041 cruise is the second cruise of the DIMES experiment, and the first conducted in a British vessel. Its primary goals were: (1) deploy a cluster of six U.K. moorings in Drake Passage to investigate how the energetic mesoscale eddies of the ACC are dissipated, and how they interact with the regional internal wave field; and (2) deploy four Florida State University (FSU) sound sources in the same area to track a large set of isopycnal RAFOS floats deployed elsewhere by American collaborators.

All work planned for the JC041 cruise was carried out successfully. The work commenced with deployment of the six U.K. moorings in a cross-like pattern centered near 56.012 S, 57.825 W, in northeastern Drake Passage. A CTD / LADCP station was occupied at the location of each mooring, and a CTD / LADCP tow-yo transect was conducted in the lee of a topographic feature at the centre of the U.K. mooring area. Subsequently, the three FSU sound source moorings were deployed at different locations further south in Drake Passage. At the most easterly of those locations and in another site in northern Drake Passage upper-ocean CTD / LADCP tow-yos were conducted to survey the submesoscale structure of interleaving features in the vicinity of the Polar Front. Finally, three profiling floats were deployed at various locations near the western edge of Drake Passage. The data provided to BODC cover the period from 04/12/2009 to 20/12/2009. The data do not cover this whole time period, with two large gaps in the data related to the cruise being within Argentine waters.

Content of data series

Instrumentation

Navigation and bathymetry (HIPLAT)

Navigation instruments used on board the ship and their main roles.

Meteorology (HIMET)

Meteorological sensors used on board the ship.

Surface hydrography (HIUWAY)

Surface hydrography sensors used on board the ship.

Data Processing Procedures

Navigation and bathymetry (HPPLAT)

Originator's Data Processing

Underway data (including navigation and bathymetry) were logged daily by the onboard logging system (TECHSAS). The data storage method used was NetCDF and pseudo-NMEA (ASCII). The NetCDF data files were manually parsed through an application in order to convert them to RVS format for data processing. Data were transferred from the TECHSAS system to the user Unix area and reformatted into Mstar format NetCDF using NOC-generated Mstar processing scripts on a daily basis. Transfer and formatting was achieved via Matlab. Further details of the originator's data processing is detailed in the cruise report. The following files were submitted to BODC:

Navigation data streams

Bathymetry data streams

BODC Data Processing

POSMVPOS processed navigation channel (averaged to 1 minute bins) in Mstar format was the ships primary GPS system and was provided as the navigation source by the originator. However, the navigation seemed to contain regular spikes which were reflected in the calculated ship's northward and eastward velocities which contained regular (every 10 mins or so) and large (changes of 100 cm^-s) spikes. Therefore the ship's bestnav in ASCII format was used to provide the main source whilst the POSMVPOS processed navigation was used to add supplementary positions in order to fill several gaps where latitude and longitude were not available from the bestnav.

Gap 1 Start: 2009/12/04 16:48:30 End: 2009/12/04 21:53:29 Duration: 306 minutes
Gap 2 Start: 2009/12/06 21:24:30 End: 2009/12/07 12:28:30 Duration: 905 minutes
Gap 3 Start: 2009/12/07 14:14:30 End: 2009/12/07 14:33:30 Duration: 20 minutes
Gap 4 Start: 2009/12/07 14:36:29 End: 2009/12/07 14:39:30 Duration: 4 minutes
Gap 5 Start: 2009/12/16 20:27:30 End: 2009/12/16 20:34:29 Duration: 8 minutes
Gap 6 Start: 2009/12/16 20:36:29 End: 2009/12/16 20:36:29 Duration: 1 minutes
Gap 7 Start: 2009/12/16 20:38:30 End: 2009/12/16 20:38:30 Duration: 1 minutes
Gap 8 Start: 2009/12/18 22:30:29 End: 2009/12/18 22:45:30 Duration: 16 minutes

The Ship's velocity was also calculated and any periods where unrealistic speeds were found, the navigation channels were flagged as suspect.

The ship's heading was taken from GYROPMV in ASCII format.

The ASCII formatted files of 1 second cycles were reformatted to the internal QXF format via the nearest neighbour method.The Mstar formatted channels which were 1 minute averaged bins were also reformatted to the internal QXF format (1 min data cycle intervals) via the nearest neighbour method.

The ship's eastward (APEWGP01) and northward (APNSGP01) velocities were calculated from the composite latitudinal (ALATGP01) and longitudinal (ALONGP01) channels. The distance run was also calculated based upon the latitude (ALATGP01) and longitude (ALONGP01) to create a new channel DSRNCV01. The data were also supplemented with a GEBCO (The General Bathymetric Chart of the Oceans) bathymetry channel (MBANGBCE). Depths were extracted from GEBCO based on the latitude (ALATGP01) and longitude (ALONGP01) of each datacycle.

bestnav

POSPMV

GYROPMV

EA 600 echo sounder

Meteorology (HPMET)

Originator's Data Processing

Data from the meteorological suite of sensors were logged daily by the onboard logging system (TECHSAS). The data storage method used was NetCDF and pseudo-NMEA (ASCII). The NetCDF data files were manually parsed through an application in order to convert them to RVS format for data processing. Data were transferred from the TECHSAS system to the user Unix area and reformatted into Mstar format NetCDF using NOC-generated Mstar processing scripts on a daily basis. Transfer and formatting was achieved via Matlab. Further details of the originator's data processing is detailed in the cruise report. The following files were submitted to BODC:

Meteorological data streams

BODC Data Processing

All Mstar formatted channels, which were 1 minute averaged bins, were reformatted to the internal QXF format (1 min data cycle intervals) via the nearest neighbour method. The ASCII formatted channels of 1 second cycles were also reformatted to the internal QXF format via the nearest neighbour method.

Relative wind speed (ERWSSS01) and direction (ERWDSS01) were corrected for the ship's velocity, to absolute wind speed (EWSBSS01) and absolute direction (EWDASS01) using relative wind speed in m s^-1, relative wind direction, the ship's eastward and northward velocities and the ship's heading. An anemometer orientation correction of 180 degrees was applied.

It was identified that some of the channels in the surfmet ASCII file did not contain the correctly corresponding data. This could have occurred when the data streams were parsed or by the wiring being changed over. The pressure channel actually contained the SPAR data, the SPAR channel contained the STIR data and the humidity channel contained the pressure data. The following tables show how the variables within the files were mapped to appropriate BODC parameter codes:

met Mstar file

surfmet ASCII file

Calibration

The voltages from the TIR and PAR channels were calibrated from volts to W m^-2 using the manufacturer's calibration sheets. PTIR (port-side) (CVLTRPO1) was calibrated using the Kipp and Zoden CM6B 973134 calibration sheet, using the equation: W m^-2= output volts * (1/10.96 x10⁶). STIR (starboard-side) (CVLTRSO1) was calibrated using the Kipp and Zoden CM6B 973135 calibration sheet, using the equation: W m^-2= output volts * (1/11.84 x10⁶). It was unclear where the PAR Skye SKE 510 sensors where located on the ship (starboard or port-side). Therefore, in order to determine which sensor calibration was the most appropriate to use for each channel, the channels were calibrated using both sensor calibrations. Which manufacturers calibration was chosen for each PAR channel was based upon which resulting calibrated values reflected their corresponding TIR channel the most, i.e. as PAR is about 45% of TIR, where TIR is the largest PAR should be expected to be the largest, and vice versa. This resulted in PPAR (port-side) (DVLTRPSD) being calibrated using the SKYE 28561 calibration: W m^-2= output volts * (1/10.17 x10⁶), and SPAR (starboard-side) (DVLTRSSD) being calibrated using the sensor SKYE 28562 calibration: W m^-2= output volts * (1/10.86 x10⁶). In order to eliminate shading effect, the two PAR sensor values were merged by taking the greater value out of the port and starboard sensors and the channel DWIRRXMX was created.

All reformatted and calibrated data were visualised using the in-house Edserplo software. Suspect and missing data were marked by adding an appropriate quality control flag.

Sea surface hydrography (HPUWAY)

Originator's Data Processing

Data from the sea surface hydrography suite of sensors were logged daily by the onboard logging system (TECHSAS). The data storage method used was NetCDF and pseudo-NMEA (ASCII). The NetCDF data files were manually parsed through an application in order to convert them to RVS format for data processing. Data were transferred from the TECHSAS system to the user Unix area and reformatted into Mstar format NetCDF using NOC-generated Mstar processing scripts on a daily basis. Transfer and formatting was achieved via Matlab with some of the data remaining in matlab format. Further details of the originator's data processing is detailed in the cruise report. The following files were submitted to BODC:

Sea surface hydrography data streams

BODC's Data Processing

Thermosalinograph measurements (TSG) were made by Falmouth Scientific Instruments OTM and OCM as well as a SBE45 Micro TSG. Only the SBE45 Micro TSG data were supplied from the originator as it was proven previously to have better reliability than the FSI sensors. All matlab formatted channels, which were 1 minute averaged bins, were reformatted to the internal QXF format (1 min data cycle intervals) via the nearest neighbour method. The ASCII formatted channels of 1 second cycles were also reformatted to the internal QXF format via the nearest neighbour method. The year, month, day, hour, minute were combined in matlab into one 'time' channel which the 1 min data cycle intervals were based on.

The salinity data and calibration as contained in the cruise report are different from that provided to BODC. The cruise report details that salinity was calculated from the conductivity and the remote temperature using the subroutine sw_salt (from the free CSIRO sea_water package) and then calibrated against 21 data points from salinity bottle samples, taken from the thermosalinograph (TSG) every 4-6 hours. The salinity data provided by BODC appears to have been calculated using conductivity and the housing temperature with no calibration applied. Therefore salinity was re-derived by BODC using conductivity housing temperature using the routine sw_salt based upon UNSECO (1983).

The following tables show how the variables within the files were mapped to appropriate BODC parameter codes:

TSG matlab file

surfmet ASCII file

Calibrations

The TSG SBE45 contains its own calibration coefficients and has outputted calibrated temperature and conductivity. The BODC derived salinity (PSALSU01) was calibrated against the bottle salinity data collected from the TSG. A total of 21 bottle salinity samples were provided by the originator, but only 18 bottle samples were used in the calibration as 4 samples did not have a corresponding underway date and time. The offset between the bottle salinity and the underway salinity was examined to identify any trends, outliers or steps in the data, with time or bottle salinity. No outliers were identified and regression analysis revealed there to be no trend in the relationship between the salinity offset and bottle salinity (p = 0.235, r₂(adjusted) = 3.0%). A mean offset calibration equation was therefore used; calibrated underway salinity (PSALSG01) = raw underway salinity + 0.027977778.

The manufacturer's calibration was applied to the raw fluorometer output to calibrate the raw volts to concentration of chlorophyll-a per (milligrams per cubic metre) using the Wetlabs WET-Star fluorometer calibration sheet. The following equation was used: mg m^-3= 14.7 * (output volts - 0.064).

The transmissometer was calibrated using the manufacturers calibration from volts to, light transmission (%) and beam attenuation (per metre). The equation below, taken from Sea-Bird electronics was calculated using values taken from the calibration sheet only. No field measurements were available to be used for the calibration. M=(Tw/(W0-Y0))*((A0-Y0)/(A1-Y1)); B=-M*Y1; where: A0 = Vair = manufacturer voltage output in air (4.806) A1 = most recent voltage output in air (not known, 4.806 used) Y0 = manufacturer dark or zero voltage (0.062) Y1 = most recent dark or zero voltage (not known, 0.062 used) W0 = manufacturer voltage output in pure water (4.698) Tw = % transmission in pure water relative to water (100%) Light transmittance (%) = M * output volts + B Beam attenuation (per metre) = -1/path length * ln (light transmittance).

All reformatted and calibrated data were visualised using the in-house Edserplo software. Suspect and missing data were marked by adding an appropriate quality control flag.

References

Unesco 1983. Algorithms for computation of fundamental properties of seawater, 1983. _Unesco Tech. Pap. in Mar. Sci._, No. 44, 53 pp.

Quality control

Large sections of the cruise track are within Argentine waters and therefore data are unavailable for these time periods, 06/12/2009 4:29 to 07/12/2009 16:19 and from 17/12/2009 07:20 to 20/12/2009 09:58.

Navigation

The heading channel does not begin until 04/12/2009 21:54:30 and ends on 17/12/2009 07:19:30. There is a large gap in the heading data between 06/12/2009 04:29:30 and 07/12/2009 16:19:30.

The only data available to fill the gaps in the bestnav was the POSMVPOS data. These data contained regular spikes which were reflected in the calculated ship's northward and eastward velocities which contained regular (every 10 mins or so) and large (changes of 100 cm-s) spikes. These spikes in the velocity channels are therefore present when the POSMVPOS data have been used to fill the navigation gaps (as above).

Meteorology

Caution should be taken when interpreting the TIR results as there is a large consistent offset between the PTIR and STIR values, with the PTIR values being consistently larger than the STIR values. It appears that the PTIR values may be higher than expected for this time of year and latitude, however, it was concluded that the values are not high enough that the data are completely unreliable and therefore have not been flagged. It is possible that the sensor calibrations are the wrong way round but as there is no way of confirming this, the calibrations have remained the same as the information given in the cruise report. Caution should also be taken when interpreting the PAR results as it is also possible that the sensor calibrations are the wrong way round as this was not stated in the cruise report but the most likely calibrations were chosen for each sensor. The PAR and TIR data are also short in length and only span the time period between 04/12/2009 21:54:30 and 09/12/2009 18:13:30.

Caution should be taken when interpreting the absolute wind channels at times when the bestnav gaps have been filled with the POSMVPOS data, as the calculated ship's northward and eastward velocities which contain regular (every 10 mins or so) and large (changes of 100 cm-s) spikes which may be reflected in these channels. The times when absolute wind have been calculated and the navigation have been filled are below;

Start: 2009/12/16 20:27:30 End: 2009/12/16 20:34:29 Duration: 8 minutes
Start: 2009/12/16 20:36:29 End: 2009/12/16 20:36:29 Duration: 1 minutes
Start: 2009/12/16 20:38:30 End: 2009/12/16 20:38:30 Duration: 1 minutes

Sea surface hydrography

The temperature, conductivity and salinity values were found to have regular spikes of 1 degree and 0.1 PSU every 12 hours or so, which is likely to be system maintenance. These have been flagged as suspect where necessary.

Problem report

Navigation and bathymetry

Caution must be taken when interpreting the bathymetry data, as the data are very patchy and look unreliable when compared to the GEBCO data.