General practices & notes on CalCOFI's Seabird 911/911+ CTD-Rosette

• Data are logged at 24hz so all data may be rederived with different coefficients if necessary
  • Raw cast data files are distributed independently of the processed data so reprocessing or other methods of processing may be applied by end users
  • Processed 1m binavg ascii data are also distributed - refer to our data pages or CTD data pages for files
• CalCOFI has always used Seabird 911 or 911+ CTD systems. The primary sensor array includes T, C, O2 with pump; the secondary array was T, C and pump only until 2009 when we started running a second O2. During a 911+ system upgrade in July 2009 (0907), we acquired an additional O2 sensor so we had enough SBE43s to start running dual sets and have adequate spares.
• CTD data are processed according to Seabird's recommendations for 911+ casts to 500m. After Data Conversion, the Window Filter module is applied to help smooth the rough ISUS nitrate sensor -see the SBE Data Processing Manual and/or the CTD Data Processing Methods web page for additional info.
• Loop edit is not usually applied to our data. In order to maintain the sensor data collected when the CTD stops for a bottle closure, loopedit is not applied since it eliminates data collected when the CTD is not moving.
• The CTD "fish" is mounted horizontally in its cage with the T intake and pump output at roughly equal height to minimize any pressure differential. The transmissometer is mounted on its side adjacent to the "fish" with an unobscured waterflow through the optical path. The fluorometer and ISUS optics are mounted optics down, at the same height as the fish with unobscured views downward. The Remote PAR is mounted on the upper rosette ring, high on the frame to prevent reflections or shading of the sensor. The altimeter is mounted vertically low on the frame with an unobstructed view down. Refer to CTD-rosette photos for additional info.
• The ISUS nitrate sensor is powered by an external 12v battery pack. The lead-acid battery is charged between casts on the rosette and vented prior to deployment.
• On 3500m casts, the ISUS nitrate sensor and battery, the remote PAR, and the SBE18 pH sensor are removed.
• The CTD is powered on ~15mins before station; ISUS battery plugged in ~20mins before station, during bottle prep.
• The CTD-rosette is deployed, sent to 10m for 2mins. If pump status is "on" and sensor pairs agree, the CTD-rosette returns to surface, data archiving (In Seasave: Real-Time Data/Start Archiving) is started, and after ~45-60secs, the CTD-rosette is lowered at 30m/mins to 100m then 60m/min, weather-permitting, to 515m. 515m terminal depth is a historical terminal depth from earlier bottle casts protocol when the terminal target depth was 500+m. In order to insure the deepest bottle was below 500m, an extra 15m was added to the cast card. This allowed for some wire angle and still be below 500m. It's practice is to preserve the continuity of the time-series.
• We try to keep the CTD sensor configuration the same for each cruise: V0/1=transmissometer/fluorometer; V2/3: altimeter/rPAR; v4/5: O21/O22; V6/7: ISUS/pH. But occasionally we test new sensors such as the RINKO O2 optode on CalCOFI 1210NH & 1611SR, which we ran as the secondary oxygen sensor. Note that altimeter and secondary SBE43 O2 sensors are removed to accommodate the RINKO which requires two voltage channels.
• Transmissometer M & B coefficients are recalculated prior to the first cast by deck testing the in-air transmission dark & light, freshly RBS-rinsed lenses. Voltage values are key-entered into a tabulation/calculation spreadsheet and M & B calculated (since 2002).
• After each cast, the dual plumbed sensor arrays are flushed with DI water for ~10secs. The carousel trigger array is freshwater rinsed. ISUS battery charging cable is attached to the ISUS battery on the rosette and charged between casts.

CalCOFI CTD data, particularly the thermodynamic properties, are computed by Seasoft based on EOS-80. Temperatures, typically from the primary temperature sensor, are merged with bottle sample data into station files which produce the hydrographic database and other data products, Hydrographic Reports, figures, IEHs. CTD sensor salinities, and oxygen values may also replace bottle measurements on mistrip, interpolated standard levels (in place of calculated interpolations) or missing samples.

Currently (May 2014) no TEOS-10 calculation for absolute salinities are calculated.

Information from Seasoft v7.23.1 (May 2014)

Algorithms used for calculation of derived parameters in Data Conversion, Derive, Sea Plot, SeaCalc III [EOS-80 (Practical Salinity) tab], and Seasave are identical, except as noted in Derived Parameter Formulas (EOS-80; Practical Salinity), and are based on EOS-80 equations.

Derived Parameter Formulas (EOS-80)

For formulas for the calculation of conductivity, temperature, and pressure, see the calibration sheets for your instrument.
 
Formulas for the computation of salinity, density, potential temperature, specific volume anomaly, and sound velocity were obtained from "Algorithms for computation of fundamental properties of seawater", by N.P. Fofonoff and R.C Millard Jr.; Unesco technical papers in marine science #44, 1983.

• Temperature used for calculating derived variables is IPTS-68, except as noted. Following the recommendation of JPOTS, T68 is assumed to be 1.00024 * T90 (-2 to 35 °C).
• Salinity is PSS-78 (Practical Salinity) (see Application Note 14: 1978 Practical Salinity Scale on Sea-Bird's website). By definition, PSS-78 is valid only in the range of 2 to 42 psu. Sea-Bird uses the PSS-78 algorithm in SBE software, without regard to those limitations on the valid range. Unesco technical papers in marine science 62 "Salinity and density of seawater: Tables for high salinities (42 to 50)" provides a method for calculating salinity in the higher range (http://unesdoc.unesco.org/images/0009/000964/096451mb.pdf). Salinity measurements in the CalCOFI sampling area never been outside the 2 to 42 psu range - typically between 32 - 36 psu.

Read more ...

Please note, the processing methods described use SIO-CalCOFI-developed software (DECODR & BtlVsCTD) to merge and correct Seabird 911+ CTD data with bottle samples. The Seasoft portion of our data processing protocol follow their recommended settings for our 911+ v2 CTD & deck unit.
"Sta.csvs" are bottle data collected at each station cast combined with 1m binavg CTD data from matching bottle depths (~20). Preliminary comparisons & plots of CTD sensor data versus bottle data are generated after each cruise. These comparisons are used to point-check the bottle data and QC the CTD sensor data. Once final bottle data are generated, eliminating fliers & mistrips, a final comparison of 1m binavg CTD to bottle data is performed. This generates final CTD data csv files for each cast that contain 1m binavg Seasoft-processed CTD sensor data, bottle-corrected Seasoft-processed CTD sensor data, and final bottle data.
The CTD temperatures are Seasoft-processed with no additional corrections applied. CTD salinities are corrected using bottle samples from depths >350m, where the profile is near vertical - offset are typically very small (<0.001). CTD SBE43 oxygen data are significantly improved using bottle oxygen samples for calibration; a linear regression is applied to bottle-correct SBE43 O2 profiles. Fluorometer and ISUS-NO3 sensor data require bottle data to generate regressions and convert their voltages to quantitative measurements. Transmissometer & PAR data are collected, processed using Seasoft and not QC'd further. 
Seabird's CTD data processing software suite, Seasoft, is available for free at seabird.com                       JRW 06/2019

CTD Data Processing Quick Guide (rev Jun 2019)

see sections below for more information & settings - these instructions reference SIO-CalCOFI inhouse software which merges CTD & bottle data then applies regression corrections to sensor data based on sensor vs calibration sample plots. CTD data compared to bottle data are 4-second averages prior-to-bottle-closure.

1. Data Conversion... (SBE):  Match CON or XMLCON files accordingly.
    1a. Calculate transmissometer M & B values from deck tests if not done during the cruise, edit/update tranmissometer .xmlcon coefficients to derive beam attenuation coefficients and % light transmission.
2. Window Filter... (SBE):  To smooth profiles & reduce spikes (median filter: 9 for all except cosine 500 for ISUS Voltage, usually setup on V6).
3. Filter... (SBE):  Filter B = 0.15 on Pressure only.
4. Align CTD... (SBE):  Oxygen sensors only (4 seconds).
5. Cell Thermal Mass... (SBE):  Default settings.
6. Derive... (SBE):  Match CON or XMLCON files accordingly.
7. ASCII Out... (SBE):  Select desired variables (see figure below for selection example), check output HDR for errors. DatCNV overwrites the existing .hdr files with new .hdr from embedded information. If the hdr information was keyed-in incorrectly during the cast and the .hdr file was corrected (cast or line.station number, for example) post-cast. The correction is lost when DatCNV generates a new hdr from the .hex file. Hexediting the .hex file, done carefully with a hex editor, will eliminate this problem.
    7a. Sta.csvs updated with .btl data (for 4sec depth value) then 1m .asc data followed by bottle data (salt, O2, chl, nutrients)
8. Step 1... (BTLvsCTD): Individual station BL files and recent preliminary sta.csvs (ieh option still works), unaveraged CTD asc files are required.
9. CSV to .XLS... (Excel):  Open the YYMM_CCTD.CSV generated by BtlVsCTD Step 1 in Excel then save as an XLS file. 
10. Salt... (Excel):  Calculate BTL-CTD offsets (>350m), remove outliers, derive means & statistics.
11. Oxygen... (Excel):  Calculate regressions for both oxygen sensors Sbeox0/1ML/L vs.  BTL_O2.
    11a. Oxygen (Excel):  Calculate regressions for both oxygen sensors Sbeox0/1uM/KG vs.  BTL_O2.
12. Nitrate... (Excel):  Calculate regression for ISUS vs. BTL_NO3.
13. Chlorophyll... (Excel):  Calculate regression for FLSP vs. BTL_CHL.
14. Split... (SBE):  Apply the split to the derived CNV files.  
15. Bin Average... (SBE):  Bin by Depth in 1 meter increments.
16. ASCII Out... (SBE):  Select correct variables.

Generating CTD.csvs for point-checking and plots

17. Step 2... (BTLvsCTD):  ASC- HDR with current sta.csvs (or IEH) and corrected Events file in folder.
18. Plot Profiles... (Matlab):  Plot up cast CSV data with bottles from BTLvsCTD Step 2. 
    18a. Zip & post preliminary CTD+Bottle csvs with metadata files online.
    18b. Copy preliminary CTD+Bottle csvs to dataserver for GTool point-checking.

For Final CTD data, repeat steps 8 - 18 using final data sta.csvs
19. Step 1... (BTLvsCTD): Regenerate comparisons using final sta.csvs (or IEH).
20. Coefficients... (Excel):  Recalculate final regressions and coefficients.
21. Step 2... (BTLvsCTD):  Regenerate final CTD.CSV files.
22. Plot Profiles... (Matlab): Create final up and down cast profiles with bottles.
23. Misc... (BTLvsCTD):  Rename ASC/HDR files, column voltages (optional), and create ZIP files.
24. Database... (Access):  Import files into annual database (usually done by JLW).
25. Data Upload... (WinScp):  Upload files to sio-calcofi.ucsd.edu in correct folders.

Recent Changes

2013 - large data files warehoused on sio-calcofi.ucsd.edu, includes CTD data & underway data
201108 - new sta.csv format replaces 00/20/ieh
5-5-11 Removed Check on Tau Correction during Step One CNV creation. 
5-6-11 No longer plotting downcast without bottles

Read more ...

CalCOFI's primary hydrographic instrument is a Seabird 911+ CTD equipped with dual temperature, conductivity and oxygen sensors mounted on a 24-10L bottle rosette. Additional CTD sensors mounted on the rosette frame include a fluorometer, transmissometer, nitrate sensor, PAR, pH and altimeter.
The CTD-Rosette is lowered into the ocean measuring a suite of seawater properties throughout the water column. Occupying the same stations (specific GPS locations) four times a year - Winter, Spring, Summer, Fall - we measure physical & biological properties: temperature, salinity, oxygen, fluorescence (chlorophyll), nutrients, and productivity from surface to 500m. Additional measurements from seawater samples collected using the rosette are combined with CTD sensor data, filling out the dataset. These seasonal measurements are published in Cruise Data Reports & added to our time-series database, both available online.
The CTD-Rosette is electronically tethered to the ship using a winch with conductive wire. This allows a computer on the ship to control the CTD and monitor the temperature, conductivity, oxygen sensor arrays plus single fluorometer, transmissometer, altimeter & nitrate sensors. As the CTD-Rosette is lowered to 500m, the temperature, salinity, oxygen, chlorophyll, nitrate, and other measurements are displayed real-time & stored on the ship's computer. Depending on the downcast profiles, mainly the depth of highest chlorophyll & mixed layer, bottles are closed at standard depths as the CTD-Rosette is brought back to surface. The seawater samples collected will be analysed & used to calibrate the various sensors or provide measurements that cannot be measured electronically.

• Temperature (T) - Seabird Electronic SBE 3plus temperature sensors: are dependable, requiring no additional calibration other than annual service and calibration performed by Seabird. Dual T sensors are plumbed horizontally, separately to SBE 5T submersible pumps. Measurements are reported in degC.
• Conductivity (C) - Seabird SBE 4C: paired/plumbed with SBE 3plus temperature sensor, measure seawater pumped by the SBE 5T pump. Standard SBE Data Processing offsets are applied to this measurement prior to deriving salinities as PSU. When using the SBE11 v2 Deck Unit, a SBE Data Processing Alignctd offset is not applied to secondary conductivity sensor since it is applied by the deck unit.
  • Salinities (in PSU) - are derived from T & C after standard SBE Data Processing modules are applied. Although salinities directly from the CTD are considered excellent, when bottle salts from greather than 350m are available, bottle-correction offsets are applied. These data are reported alongside the uncorrected salinities in our CTD data products.
• Oxygen - Seabird SBE 43 Dissolved Oxygen Sensor: dual SBE 43 O2 sensors are plumbed inline with the paired T & C sensors, between the T-C pair and pump. Since the response time of the SBE 43 oxygen sensor is slower that T & C, SBE Data Processing's AlignCTD module is used to apply a 4-second to sensor data before oxygen measurements are derived.
  In addition to standard Seabird SBE Data Processing recommended data processing, when seawater oxygen samples are analysed during the cruise. They are used to calibrate sensor oxygen values: cruise-corrected CTD oxygen & station-corrected CTD oxygen data are derived and included in the CTD data products.
• Wetlabs ECO AFL/FL FLuorometer: measures fluorescence-chlorophyll-a; used during the downcast to identify the depth of the chlorophyll-a maximum, which determines the bottle sample depths. Although an estimated chlorophyll-a can be derived using the fluorometer factory calibration. CalCOFI uses cold-extracted chlorophyll-a sample data versus fluorometer voltages to derive estimated chlorophyll-a measurements.
• WetLabs C-Star 25cm 440nm Transmissometer measures % light transmission & beam attenuation coefficients. Before the first cast, CalCOFI measures the dark & light voltages on deck, calculates the M & B coefficients and enters these values under the transmissometer calibration in the CTD .xmlcon file. This is the only calibration done on the transmissometer during a cruise. A Triton or RBS (mild soap, squirt bottle) rinse of the transmissometer optics is done routinely before a cast to clean the optical surfaces.
• Satlantic MBARI-ISUS v3 Nitrate Sensor: recently upgraded to firmware version 3, the ISUS can now be recalibrated by CalCOFI. MBARI-ISUS Version 2 deployed on cruises prior to 1708SR required calibration by Satlantic (prior to Seabird Electronics merger). Estimated nitrate may be displayed real-time using user-polynomial coefficients (from the previous cruise) entered into Seasave .xmlcon. Estimated nitrate data are derived post-cruise from the ISUS voltages using seawater nitrate samples plotted versus average ISUS voltages. Cruise-corrected and station-corrected values are calculated and reported in the CTD data products. The ISUS has been deployed on most cruises since 2004, on casts 1000m or less.
• Seabird SBE 18 pH Sensor: has been deployed on all cast (1000m or less) since 2009. This sensor is serviced annually and is checked before each cruise using 3 buffer solutions (pHs 4, 8, 10). The sensor electrode is stored in buffer solution between casts to prevent drying out.
• Biospherical Single-Channel Photoradiometers (QSP-2300 PAR & QSR 2200 Surface PAR): a remote PAR is deployed on casts up to 1000m; surface PAR is interfaced with the SBE 11 Deck Unit on most cruises. Factory calibration and coefficients are entered into Seasoft .xmlcon file for both sensor. No additional calibration is performed; standard SBE Data Processing is performed but nothing specific to either PAR other than WFilter.
• Benthos Altimeter (PSA-916): used on all cruises and all casts unless a oxygen optode is deployed (CC1210NH, CC1611SR) then only on stations less than 500m deep. Factory calibration coefficients are entered into Seasave .xmlcon but no calibration is performed. Height off the bottom typically is displayed ~50m from bottom but depends on sea state, bottom composition or wire angle.

A Sea-Bird Electronics 911plus V2 CTD collects vertical profile data at every CalCOFI station. In addition to being a dual TCO (temperature, conductivity, and oxygen) system, the CTD also interfaces with a transmissometer, fluorometer, PAR/SPAR meters, altimeter, nitrate, and pH sensor. Connected to a shipboard data-acquisition computer through an electronically-conductive winch wire, sensor data are collected and displayed real-time using Seasave V7 on a Windows PC. The CTD is normally lowered to terminal depth of 515 m, bottom-depth permitting, but is routinely deployed within meters from the seafloor at nearshore SCCOOS and basin stations. To ensure high resolution sampling in areas with significant hydrological and biological gradients a speed of ~30 m/min is used for the first 100 m then ~60 m/min to depth without stopping. During retrieval, the CTD is paused for at least 20 seconds at target bottle depths to adequately flush each 10 liter sample bottle prior to closure. Seawater samples are analyzed onboard (e.g., salinity, oxygen, nutrients, chl) and are used to correct measured CTD values.

Read more ...

Setting up and using the ISUS with Seabird CTD

1). Connect the ISUS Analog Out port to a 6-pin CTD port – we use voltage channel 6 on a universal Y-cable, allowing pH to occupy channel 7.  Map that port in Seasave as user-polynomial; under the user-polynomial, enter: A0 = - 7.1168, A1 = 27.155. 
These are linear regression coefficients from an earlier CalCOFI (1501NH).

2). Connect the battery cable to the ISUS fifteen minutes prior to deployment to "warm-up". This improves data quality & stability.

3). SIO-CalCOFI uses a 10m charging cable strung from an AC adapter in the wetlab to the battery mounted on the CTD-rosette. Between casts, the battery is left mounted on the rosette and charged. This method has improved the reliability of the ISUS to have adequate power for the full cast, provided the battery has been charged between deployments. Occasionally, the CTD operator forgets to unplug the ISUS and it logs deck data until discovered - usually next station, draining the battery & filling the ISUS internal memory.
The original method was swapping batteries every three 500m cast then charging inside the ship. The Wetlabs 12v lead-acid batteries usually lasted ~3.5 hours before requiring removal and charging. Charging takes 12-15hrs so three primary batteries were rotated. 
Regardless of charging method, be sure to unstopper the battery's vent plug to allow hydrogen emissions.  Re-stopper, greasing lightly if necessary, before mounting the fresh battery on rosette. When charging on the rosette, protect from seawater intrusion by keeping the vent-plug closed but vent the battery before deployment.

4). Remove the ISUS on casts deeper than 1000m. The 512kb memory allows storage of ~75 casts without the need to download data and purge the memory. Data are downloaded by connecting the ISUS to a Windows laptop’s serial port. If the ISUS memory is full, the sensor will still work but the sensor data profile will "stair-step" significantly.

• For v2 :
  • Use terminal.exe to handshake with the ISUS, settings are 38400,n,8,1.  You will see the ISUS boot messages and after several seconds it will start counting down for data acquisition.
  • Press S to stop the countdown then M (or H) to get the system menu, follow the prompts. 
  • Use the file commands to delete all the data or download all data.
    • To download data
      • initiate the download in terminal.exe by sending the command to the ISUS, selecting 115200 baud
      • disconnect terminal.exe then switch to HyperTerminal (Windows default com program – I usually have it already setup in the background at 115200,n,8,1)
      • connect to the ISUS then  Transfer/Receive File, select ymodem-g.
        It will download three files then requires you to Transfer/Receive File YModem-G again, repeat 10x to download 30 files. 
    • Delete all data to free up the memory.
    • Collect more data, repeat step 4 as necessary.
• For v3:
  • connect the ISUS USB connector to laptop's USB port
  • run Satlantic ISUScom program on the laptop then connect the battery to the ISUS if needed to power up the sensor
  • download all the data files then delete them. This should free up the internal memory for the next set of casts.