The standard practices for CTD deployment, seawater sampling, and general protocols are linked to the 70+ year time series.

• • Originally, 20 Nansen/Niskin bottles were suspended on the non-conductive hydro wire at target depths based on secchi depth (1% light level). A cast card was composed then presented to the winch shack with specific wireout measurements. These measurements suspended the 5L Niskin or 2L Nansen bottles at specific target depths in the water column. Typically, bottom-depth permitting, 20 bottles were hung with reversing thermometers & messengers. After an incubation period of ~10mins for the reversing thermometers to equilibrate. A messenger was dropped to trip & close the surface bottle, starting the chain reaction as the next bottle closed and released its messenger. Hopefully, if all bottle messengers were rigged correctly, all 20 bottles tripped. The person dropping the messenger would hold the wire and feel for 20 vibrations up the wire, indicating all the bottles had closed. The bottles were returned to surface and removed from the wire and hung on racks in the wet lab. Seawater samples were drawn from each bottle in similar fashion as current rosette seawater samples. After ~45min, the reversing thermometers were read and tabulated. A separate 6-bottle primary productivity cast was done around noon each day. 

• • When the CTD was introduced in the early 1990s, in order to maintain continuity with the time series. Protocols were established to insure the new technology would maintain data measurements comparable to the bottle method. Electronic sensors in the ocean behave differently at different pressures and temperatures (hysteresis) affecting their responsiveness. This FAQ was composed to explain certain practices done on CalCOFI CTD casts.
    
    • • Deck test – usually once a day (usually by the CTD primary operator), a deck test is performed to log voltages on-deck/in-air & saved to disk. The fluorometer and transmissometer light paths are rinsed with DI, voltages recorded for ~1min. 
        On the fluorometer, a standard is placed in the light path to max out the voltage. This is to test the fluorometer is responding as it should and establish the 0 & 100% voltage ranges.
        The transmissometer 100% voltage is the in-air voltage after rinsing the lenses. Then the light path is blocked to measure the 0% light voltage. These are used to calculate the M & B coefficients entered into the transmissometer’s seasoft.con cfg. This improves the %light transmission and attenuation measurements of the transmissometer. If not done daily, this should be done before the 1st cast so the M & B can be entered into seasoft.con. There is an Excel spreadsheet that automates the calculation – just enter the two voltages for the appropriate transmissometer #SN and it will calculate the M & B. Check/Change the light path distance when using a new transmissometer.
    • • Lower the CTD/Rosette to 10m and wait 2-3mins: as mentioned, temperature can have a significant affect on certain sensors. Particularly, the conductivity sensors and the SBE43 oxygen sensors. Without going into too much detail about membrane permeability and electrolyte temperature, soaking these sensor for a few (to several) minutes improves their downcast surface measurement. Plus, since the plumbing will have some air and DI water in it, it allows the pumps to turn on and purge the DI and air from the tubing. The Seabird pumps are activated by contact to seawater, so if you ever need to troubleshoot the system on-deck. Place the pumps in a bucket of seawater to check the flow rate.
    • • Lowering at 30m/min for the first 100m then 60m/min from 100m to 515m. The Seabird CTD originally did not have pumps or tube-plumbed sensors so ship roll caused significant data variability. Seabird added the pumps plumbed to the Temperature, Conductivity, and Oxygen sensors to stabilize the flow of seawater past the sensors, regardless of ship roll or descent rate. The descent rates have more to do with minimizing the conductive wire going slack then taunt when the ship rolls. So 30m/min, when the CTD is shallower, reduces the stress of ship roll on the wire. It also improves the resolution of sensor data through the thermo/halocline. The slower descent rate is most important in rougher seas, but it’s less confusing to have the winch operators to do it all the time. After 100m, the weight of the wire and rosette are enough that the ship roll is not as critical. Although, when the seas are really rough, both the 30m/min and 60m/min rate can be reduced to insure the wire tension does not jump above what’s reasonable. We often did 20m/min then 30m/min to 515m during rough weather. Since the pumps are moving water past the sensors, it doesn’t affect the data significantly.
    • • Why 515m? When hanging bottles, to insure a large wire angle did not result in the terminal bottle being shallower than 500m, the terminal bottle was placed at 515m. The messenger was not dropped until the wire angle was less than 10deg. Since we usually display CTD pressure as dbs on the console during the cast, we’d send the CTD to 515m db. Unless you have the CTD remote depth readout box in the winch shack, sending the CTD to 515m of wireout usually resulted in a terminal depth below 500m. If there is a severe CTD wire angle, then the CTD operator should have the winch operator continue until the CTD terminal depth is at least 505m. Having a downcast without stoppages to 505+m results in the optimal data quality.
    • • 20 CTD Primary Bottle Depths: bottom depth permitting, the CTD cast goes to 515m without stopping. After 1min or more at terminal depth to let the wire angle settle, the 515m bottle is closed. BTW – these protocols carry-over to deeper casts as well. The time-series is primarily 20 bottle depths per station. Standard levels may be interpolated data if the bottle did not close at a standard level, such as 100m. If the ship is rolling substantially, the CTD operator should try to close the bottle at the desired depth by timing the triggering. 20 bottles were hung on the wire originally, and how many seawater samples were drawn for analyses.
        * The six primary productivity bottle depths are integrated into a 24-bottle noon-time prodo cast. This combines the two separate wire casts that were performed at the noon-time station.
        * On non-prodo 500m stations, you have 4 extra bottles for LTER or ancillary project sampling. Tripping only 12 bottles or fewer is a way of reducing the sample load when personnel are limited. But the more oxygen samples, chlorophyll samples, and salts deeper than 350m the better for bottle-correcting the sensor data.
        * Like the hydrowire casts secchi depth, the CTD bottle depths are based on the CTD’s fluorometer chlorophyll max and mixed layer. Type I, II, or III should set the 10m bottle spacing around the thermocline/halocline.
    • • 6 Prodo bottle depths: this is an easy one to explain – there are 6 incubation tubes of various light levels. So based on the secchi depth and calculated 1% light level, the prodo bottle depths are supposed to hit the 6 light levels.
    • • Why only use the salt samples deeper than 350m to bottle-correct the CTD salinity data? If you look at the CTD salinity or termperature+conductivity profiles, after 350m the lines are fairly vertical. This makes the ship roll have less of an affect on the seawater sampled by the bottles. Mixed layer bottles could also be used to correct CTD salinity but the depth of the mixed layer often changes station to station. The 350m or deeper profile slopes are pretty stable.

CalCOFI CTD data collection is always evolving so many of these protocols may be fluid particularly based on personnel availability. Collecting high-quality CTD sensor data is the primary mission. Bottle sample measurements can improve sensor data, so when possible, are highly recommended. Sensor technology is improving with oxygen optodes, nitrate sensors, and pH sensors requiring fewer calibration seawater samples. But electronics in the ocean will always have challenges and collecting seawater can insure the data quality and believeability.

JRW 08/03/2024