JR17007 Originator CTD Data Processing
Sampling Strategy
A total of 20 CTD casts were performed during JR17007.
Data Processing
For each CTD cast the following raw data files were generated:
- JR17007_XXX.bl (a record of bottle firing locations)
- JR17007_XXX.hdr (header file)
- JR17007_XXX.hex (raw data file)
- JR17007_XXX.con (configuration file)
where XXX is the cast number of the CTD data series.
The CTD processing was started using Seabird Data Processing version 7.26.7.114 where the following modules were run:
- Data conversion - converted raw data from engineering units to binary .cnv files using any calibrations in the instrument configuration file and created .ros files.
- Wild edit - flagged any major spikes however this was not applied to conductivity and temperature as it resulted in bad data values of oxygen concentration after dynamic corrections.
- Filter - smoothed the high frequency pressure and depth data using a low-pass filter (values of 0.15 - recommended by SeaBird).
- AlignCTD - shifted conductivities and oxygen relative to pressure to compensate for sensor time lag. A 5s correction was applied to all casts.
- CellTM - ran a recursive filter to remove conductivity cell thermal mass effects from measured conductivity.
- Soak - identified the soak, up and downcast for each profile.
- Section - ran to remove the soak period for each cast.
- LoopEdit - procedure run with a minimum CTD velocity of 0 m s -1 and exclude scans marked bad ticked.
- Derive - derived computation salinity, oxygen saturation and concentration in ml l -1 .
- BinAverage - averaged all variables to 0.2 db and 1 db.
- Strip - removed the primary and secondary salinity channels and oxygen concentration (in ml l -1 ) variables obtained at the Data Conversion stage.
Calibrations
Temperature
A total of 308 independent temperature readings where used to determine a calibration equation for temperature data. Two hypothesis were tested, a regression and offset, after outliers were identified and removed from the final dataset, a total of 292 readings for the primary sensor and 290 for the secondary remained. The two hypothesis were tested and it was found that none improved the data quality significantly, therefore it was decided that no calibration would be applied to these data.
Salinity
A total of 31 discrete salinity samples were analysed throughout the cruise, covering a wide range of salinity values. For each sample the bottle was rinsed 3 times with the Niskin seawater, filled, plastic insert fitted, bottle neck wiped and lid put on. Once a crate of 24 bottles was full, it was placed in the Autosal laboratory to acclimatise to temperature for at least one day prior to analysis. At the start and end of each crate a standard seawater (SSW) sample was analysed, in order to monitor the drift of the instrument but no clear drift pattern was visible.
There did not appear to be any temporal drift in the sensors, or a drift relative to pressure and once all the outliers were identified and removed from the dataset, a constant offset was used to correct the data for both sensors.
| # Samples | Salinity offset | |
|---|---|---|
| Sensor 1 | 23 | 0.007043 |
| Sensor 2 | 22 | 0.002773 |
Oxygen
The oxygen calibration equation was determined by Tim Brand from SAMS and applied to the data, 18 samples, from a total of 21 contributed to this procedure (R 2 = 0.9258)
O 2 _cal = (O 2 _raw * 0.8177) + 41.654
