WHOTS (16-17) Cruise Shipboard Observations
Contents
4. WHOTS (16-17) Cruise Shipboard Observations¶
The hydrographic profile observations made during the WHOTS cruises were obtained with a Sea-Bird CTD package with dual temperature, salinity, and oxygen sensors. This CTD was installed on a rosette-sampler with 5 L Niskin sampling bottles for calibration water samples. Furthermore, the ship Oscar Sette came equipped with a thermosalinograph system that provided a continuous depiction of the near-surface layer’s temperature and salinity. Horizontal currents over the depth range of 30-700 m were measured from the shipboard 75 kHz Ocean Surveyor (OS75) ADCP (narrowband) with a vertical resolution of 16m for the WHOTS-16 and WHOTS-17 cruises. Broadband mode for the OS75 provided additional current data over the range upper 200 m with a vertical resolution of 8m.
Data gaps occurred when the system was shut down temporarily during communications with the acoustic releases used for the moorings during both cruises. Periods of missing data between 300 and 450 m in the broadband ADCP were apparent due to the lack of scattering material in the water.
4.1. Conductivity, Temperature, and Depth (CTD) Profiling¶
Continuous measurements of temperature, conductivity, dissolved oxygen, and pressure were made with the UH Sea-Bird SBE-9/11Plus CTD underwater units #91361 and #850 during WHOTS-16 and WHOTS-17 cruises respectively. The CTD was equipped with an internal Digiquartz pressure sensor and pairs of external temperature, conductivity, and oxygen sensors.
Each temperature-conductivity sensor pair used a Sea-Bird TC duct, which circulated seawater through independent pump and plumbing installations. The CTD configuration also included two oxygen sensors, installed in the plumbing for each sensor set. In both cruises, the CTD was mounted in a vertical position in the lower part of a rosette sampler, with the sensors’ water intakes located at the bottom of the rosette.
The package was deployed on a conducting cable, which allowed for real-time data acquisition and display. The deployment procedure consisted of lowering the package to approximately 10 dbar and waiting until the CTD pumps started operating. The CTD was then raised until the sensors were close to the surface to begin the CTD cast. The time and position of each cast were obtained via a GPS connection to the CTD deck box. Four salinity samples were taken on each cast for calibration of the conductivity sensors.
4.1.1. Data Acquisition and Processing¶
CTD data were acquired at the instrument’s highest sampling rate of 24 samples per second. Digital data were stored on a laptop computer, and, for redundancy, the analog signal was recorded on a separate computer using a sound card and Audacity (TM) software. Backups of CTD data were made onto USB storage cards.
The raw CTD data were quality controlled and screened for spikes described in the WHOTS Data Report 1 [Santiago-Mandujano et al., 2007]. Data alignment, averaging, correction, and reporting were done as described in [Tupas et al., 1993]. Spikes in the data occur when the CTD samples the disturbed water of its wake. Therefore, the downcast samples were rejected when the CTD was moving upward or when its acceleration exceeded 0.5 m s-2 in magnitude. The data were subsequently averaged into 2-dbar pressure bins after calibrating the CTD conductivity with the bottle salinities.
The data were additionally screened by comparing the T-C sensor pairs. These differences permitted the identification of problems with the sensors. The data from only one T-C pair, whichever was deemed most reliable, is reported here. Only data from the downcast are reported, as wake effects from the rosette commonly contaminate upcast data.
Temperature is reported on the ITS-90 scale. Salinity and all derived units were calculated using the UNESCO (1981) routines; salinity is reported in the Practical Salinity(SA) scale (PSS-78). Oxygen is reported in umol kg-1.
4.1.2. CTD Sensor Calibration and Corrections¶
4.1.2.1. Pressure¶
The pressure calibration strategy for CTD pressure transducers #75434 and #1430 used during WHOTS-16 and WHOTS-17 cruises respectively employed a high-quality quartz pressure transducer as a transfer standard. Periodic recalibrations of this lab standard were performed with a primary pressure standard. The only corrections applied to the CTD pressures were a constant offset determined when the CTD first enters the water on each cast. Also, a span correction determined from bench tests on the sensor against the transfer standard was applied. These procedures and corrections are thoroughly documented in the HOT-2019 data report [Fukieki et al., 2021].
4.1.2.2. Temperature/Conductivity¶
Sea-Bird SBE-3-Plus temperature and SBE 4C conductivity transducers were used during WHOTS-16 and -17 cruises. These sensors’ history and performance have been monitored during HOT cruises, and calibrations and drift corrections applied during WHOTS cruises are thoroughly documented in the HOT-2019 data report [Fukieki et al., 2021].
4.1.2.3. Dissolved Oxygen¶
Sea-Bird SBE-43 oxygen sensors were used during the WHOTS-16 and -17 cruises. The WHOTS-16 oxygen data were calibrated using calibration coefficients obtained during the HOT-315 cruise conducted on 3-7 September 2019, before the WHOTS-16 cruise, which used the same oxygen sensors. The CTD empirical calibration was performed using oxygen water samples and the procedure from [Owens and Millard, 1985]. See [Tupas et al., 1996] for details on these calibrations procedures. The oxygen data from WHOTS-17 were calibrated using calibration coefficients obtained during the HOT-327 cruise conducted on 15-19 February 2021, before the WHOTS-17 cruise, which used the same oxygen sensors.
4.2. Water Sampling and Analysis¶
4.2.1. Salinity¶
Salinity samples were collected by a rosette sampler during CTD casts at selected depths during WHOTS-16 and -17, and then sub-sampled in 250 ml glass bottles. The top of each bottle and thimble were thoroughly dried before being tightly capped to prevent water from being trapped between the cap or thimble and the bottle’s mouth. It has been observed that residual water trapped in this way increases its salinity due to evaporation, and it can leak into the sample when the bottle is opened for measuring. Samples from each cruise were measured after the cruise in the UH laboratory using a Guildline Autosal 8400B SN 70168 for WHOTS-16 and SN 73647 for WHOTS-17. International Association for Physical Sciences of the Ocean (IAPSO) standard seawater samples were measured to standardize the Autosal, and samples from a large batch of “secondary standard” (substandard) seawater were measured after every 24-48 samples to detect drift in the Autosal. Standard deviations of the secondary standard measurements were less than ± 0.001 for WHOTS-16 and -17 cruises Table 4.1.
The substandard water was collected by a rosette sampler from 1020 m at station ALOHA during HOT cruises and drained into a 50-liter Nalgene plastic carboy. In the laboratory, the water was then thoroughly mixed in a glass carboy for 20 minutes by manually shaking, rolling, and tilting the carboy vigorously, after which a 2-inch protective layer of white oil was added on top to deter evaporation. The substandard water was allowed to stand for approximately three days before it was used and was stored in the same temperature-controlled room as the Autosal, protecting it from the light with black plastic bags to inhibit biological growth. Substandard seawater batch #67 was prepared on August 18, 2019, and it was used for WHOTS-16. The batch #71 was prepared on August 27, 2021, and it was used for WHOTS-17.
Samples from the WHOTS-16 were measured on October 28, 2019 and samples from WHOTS-17 were measured on September 13, 2021. Table 4.1 shows the precision measurements of the secondary sub-standards.
Cruise |
Mean Salinity +/- SD |
# Samples |
Substandard Batch |
IAPSO Batch |
|---|---|---|---|---|
WHOTS-16 |
34.4985 ± 0.0005 |
4 |
67 |
P163 |
WHOTS-17 |
34.5011 ± 0.0004 |
28 |
71 |
P164 |
4.3. Thermosalinograph Data Acquisition and Processing¶
4.3.1. WHOTS-16 Cruise¶
Near-surface temperature and salinity data during the WHOTS-16 cruise were acquired from the thermosalinograph (TSG) system installed on the NOAA Ship Oscar Sette. The sensors were sampling water from the continuous seawater system running through the ship. They comprised one thermosalinograph model SBE-21 (SN 3168) and a micro-thermosalinograph model SBE-45 (SN 0290), both with (internal) temperature and conductivity sensors located in the ship’s chemistry lab, about 70 m from the hull intake; and an SBE-38 (SN 266) external temperature sensor located at the entrance of the water intake. All instruments recorded data every second. The water intake is located at the ship’s bow, forward from the starboard side bow thruster at a depth of 3 m. The system has a flow meter in the chemistry lab, showing a flow rate of about 1.1 liters/minute during the cruise. Only the SBE-45 has a debubbler. Salinity water samples were taken every 8 hours from the exhaust in the Chemistry lab using 0.25-liter glass bottles, to be measured in the UH lab to correct any drift in the thermosalinograph conductivities.
4.3.1.1. Temperature Calibration¶
External temperature data from the SBE-38 sensor (last calibrated at Sea-Bird on December 29, 2019) were used to measure the seawater temperature. These data were compared to the data collected during CTD casts.
4.3.1.2. Nominal Conductivity Calibration¶
Data from the SBE-45 conductivity and temperature sensors were used to calculate the intake seawater salinity. These sensors were last calibrated at Sea-Bird on February 19th, 2019. All conductivity data from the thermosalinograph were nominally calibrated with coefficients from this calibration. However, all the final salinity data reported here were calibrated against bottle data, as explained below.
4.3.1.3. Data Processing¶
Daily files containing navigation data recorded every second were concatenated with the thermosalinograph data. The thermosalinograph data were then screened for gross errors, with upper and lower bounds of 18°C and 35°C for temperature and 3 and 6 Siemens/m for conductivity. There were 92 points outside the valid temperature range and no points outside the valid conductivity range.
A 5-point running median filter was used to detect one- or two-point temperature and conductivity glitches in the thermosalinograph data. Glitches in temperature and conductivity detected by the 5-point median filter were immediately replaced by the median. Threshold values of 0.3°C for temperature and 0.1 Siemens/m for conductivity were used for the median filter. After running the filter, there were 13 internal temperature, 100 external temperature, and 913 conductivity points replaced with the median.
A 3-point triangular running mean filter was used to smooth the temperature and conductivity data after passing the glitch detection.
The thermosalinograph aboard the Ship Oscar Sette was set to record data every second. The system had to be secured on the last day of the cruise due to the bad weather because it kept shutting down due to air going into the plumbing, causing the pumps to stop working.
Data were visually scanned to flag spikes likely caused by contamination due to the introduction of bubbles to the flow-through system during transits or rough conditions. Of 69,311,136 data points, 141,365 conductivity data points were flagged as bad.
4.3.1.4. Bottle salinity and CTD Salinity Comparisons¶
The thermosalinograph salinity was calibrated by comparing it to bottle salinity samples drawn from a water intake next to the thermosalinograph every 8 hours throughout the cruise. Of the twenty-one thermosalinograph bottles sampled, bottle #18, #19, #20, and #21 were identified as a conductivity outlier and were discarded from the analysis. Samples were analyzed as described in Water Sampling and Analysis. The comparison was made in conductivity to eliminate the effects of temperature. The conductivity of each bottle sample was computed using the salinity of the bottle, thermosalinograph temperature, and a pressure of 10 dbar, which includes the pressure of the flow-through system’s pump.
Salinity samples were drawn from the flow-through system, located less than 0.5 m from the SBE-45. Consequently, there should be virtually no delay between when the water passes through the thermosalinograph and sampled. A 90-second average centered on the sample draw time was chosen for processing purposes.
In order to make the comparison in conductivity units, the CTD conductivity was calculated using the 4 dbar downcast CTD salinity, the internal thermosalinograph temperature, and a pump pressure of 10 dbar. There were eleven CTD casts conducted during WHOTS-16 while the thermosalinograph was running. Casts 4, 5, 6, 7, and 9 were removed from the analysis as temperature and conductivity outliers.
A cubic spline was fit to the time series of the differences between the bottle and TSG conductivity, and a correction was obtained for the TSG conductivities. Salinity was calculated using these corrected conductivities, the thermosalinograph temperatures, and 10 dbar pressure. After applying corrections, the mean difference between the bottle and thermosalinograph salinities was -0.00004 psu with a standard deviation of 0.001886 psu. The mean CTD - thermosalinograph difference was -0.0028 psu with a standard deviation of 0.001118 psu.
4.3.1.5. CTD Temperature Comparisons¶
There were 11 CTD casts conducted during WHOTS-16, one of which was a test cast offshore Honolulu (Station 20) and five at Station 52 (WHOTS-16), and five at Station 50 (WHOTS-15), respectively. The 4 dbar downcast CTD temperature data from those casts were used to compare with the thermosalinograph data at the time of the casts. This comparison gives an estimate of the quality of the thermosalinograph measurements. Of the 11 casts, five were identified as temperature outliers after comparing it against the thermosalinograph data and removed from the analysis. The mean difference between the CTD and the internal temperature sensor was 0.058°C, with a standard deviation of ± 0.056°C.
4.3.2. WHOTS-17 Cruise¶
Near-surface temperature and salinity data during the WHOTS-17 cruise were acquired from the thermosalinograph (TSG) system installed on the NOAA Ship Oscar Sette. The sensors were sampling water from the continuous seawater system running through the ship, and comprised one thermosalinograph model SBE-21 (SN 3168) and a micro-thermosalinograph model SBE-45 (SN 0290), both with (internal) temperature and conductivity sensors located in the ship’s chemistry lab, about 70 m from the hull intake; and an SBE-38 (SN 266) external temperature sensor located at the entrance of the water intake. All instruments recorded data every second. The water intake is located at the bow of the ship, forward from the starboard side bow thruster at a depth of 3 m. The system has a flow meter in the chemistry lab, showing a flow rate of about 1.1 liter/minute during the cruise. Only the SBE-45 has a debubbler. Salinity water samples were taken every 8 hours from the exhaust in the Chemistry lab using 0.25 litter glass bottles, to be measured in the UH lab to correct for any drift in the thermosalinograph conductivities.
4.3.2.1. Temperature Calibration¶
External temperature data from the SBE-38 sensor (last calibrated at Sea-Bird on November 26, 2020) were used to measure the seawater temperature. These data were compared to the data collected during CTD casts.
4.3.2.2. Nominal Conductivity Calibration¶
Data from the SBE-45 conductivity and temperature sensors were used to calculate the intake seawater salinity. These sensors were last calibrated at Sea-Bird on November 17, 2020. All conductivity data from the thermosalinograph were nominally calibrated with coefficients from this calibration. However, all the final salinity data reported here were calibrated against bottle data, as explained below.
4.3.2.3. Data Processing¶
Daily files containing navigation data recorded every second were concatenated with the thermosalinograph data. The thermosalinograph data were then screened for gross errors, with upper and lower bounds of 18°C and 35°C for temperature and 3 and 6 Siemens m -1 for conductivity. There were 488 points outside the valid temperature range and no points outside the valid conductivity range.
A 5-point running median filter was used to detect one- or two-point temperature and conductivity glitches in the thermosalinograph data. Glitches in temperature and conductivity detected by the 5-point median filter were immediately replaced by the median. Threshold values of 0.3°C for temperature and 0.1 Siemens m-1 for conductivity were used for the median filter. After running the filter, there were 283 internal temperature, 1998 external temperature, and 341 conductivity points replaced by the median. A 3-point triangular running mean filter was used to smooth the temperature and conductivity data after passing the glitch detection.
The thermosalinograph aboard the Ship Oscar Sette was set to record data every second. Both thermosalinographs exhibited a number of conductivity and temperature glitches due to air going into the plumbing. In addition, the system had a drainage problem according to the ship’s technician. The data between August 26 at 13:30 and 27 at 07:00 are particularly bad because it was during transit back to Oahu to disembark a crew member with medical problems, and the flow through the system was stopped during that time.
Data were visually scanned to flag spikes likely caused by contamination due to the introduction of bubbles to the flow-through system during transits or rough conditions. Of 649,826 data points, 133,851 conductivity data points were flagged as bad.
4.3.2.4. Bottle salinity and CTD Salinity Comparisons¶
The thermosalinograph salinity was calibrated by comparing it to bottle salinity samples drawn from a water intake next to the thermosalinograph every 8 hours throughout the cruise. Of the sixteen thermosalinograph bottles sampled, bottle #1 was identified as a conductivity outlier and were discarded from the analysis. Samples were analyzed as described in Water Sampling and Analysis. The comparison was made in conductivity to eliminate the effects of temperature. The conductivity of each bottle sample was computed using the salinity of the bottle, thermosalinograph temperature, and a pressure of 10 dbar, which includes the pressure of the flow-through system’s pump.
Salinity samples were drawn from the flow-through system, located less than 0.5 m from the SBE-45. Consequently, there should be virtually no delay between when the water passes through the thermosalinograph and sampled. A 90-second average centered on the sample draw time was chosen for processing purposes.
In order to make the comparison in conductivity units, the CTD conductivity was calculated using the 4 dbar downcast CTD salinity, the internal thermosalinograph temperature, and a pump pressure of 10 dbar. There were ten CTD casts conducted during WHOTS-17 while the thermosalinograph was running. Casts 1, and 10 were removed from the analysis as temperature and conductivity outliers.
A cubic spline was fit to the time series of the differences between the bottle and TSG conductivity, and a correction was obtained for the TSG conductivities. Salinity was calculated using these corrected conductivities, the thermosalinograph temperatures, and ten dbar pressure. After applying corrections, the mean difference between the bottle and thermosalinograph salinities was less than 1 mpsu with a standard deviation of 0.000662 psu. The mean CTD - thermosalinograph difference was -0.00018 psu with a standard deviation of 0.00124 psu.
4.3.2.5. CTD Temperature Comparisons¶
There were ten CTD casts conducted during WHOTS-17, one of which was a test cast offshore Honolulu (Station 20), one at Aloha Station (Station 2), five at Station 50 (WHOTS-17), and two at Station 52 (WHOTS-16). The 4 dbar downcast CTD temperature data from those casts were used to compare with the thermosalinograph data at the time of the casts. This comparison gives an estimate of the quality of the thermosalinograph measurements. Of the ten casts, two were identified as temperature outliers after comparing it against the thermosalinograph data and removed from the analysis. The mean difference between the CTD and the internal temperature sensor was -0.247°C, with a standard deviation of ± 0.067°C.
4.4. Shipboard ADCP¶
4.4.1. WHOTS-16 Deployment Cruise¶
Currents were measured for the cruise duration over the depth range of 30-700 m with a 75 kHz RDI Ocean Surveyor (OS75) ADCP working in narrowband mode with a vertical resolution of 16 m and broadband mode with a vertical resolution of 8 m. The system yielded good data [Santiago-Mandujano et al., 2021] during operations near the WHOTS-15 and WHOTS-16 moorings. The broadband system only recorded good data in the upper 200 m. The times of the datasets from the OS75 kHz are shown in Table 4.2.
WHOTS-16 |
OS75nb |
OS75bb |
|---|---|---|
File starting time |
10/04/19 19:38:54 |
10/04/19 19:38:54 |
File ending time |
10/12/19 20:14:30 |
10/12/19 20:14:30 |
4.4.2. WHOTS-17 Deployment Cruise¶
Currents were measured for the duration of the cruise over the depth range of 30-700 m with a 75 kHz RDI Ocean Surveyor (OS75) ADCP working in narrowband mode with a vertical resolution of 16 m, and in broadband mode with vertical resolution of 8 m. The system yielded good data (see [Santiago-Mandujano et al., 2022]) during operations near the WHOTS-16 and WHOTS-17 moorings. The broadband system only recorded good data in the upper 200 m. The times of the datasets from the OS75 kHz are shown in Table 4.3.
WHOTS-17 |
OS75nb |
OS75bb |
|---|---|---|
File starting time |
08/20/21 01:22:22 |
08/20/21 01:22:22 |
File ending time |
09/01/21 19:44:41 |
09/01/21 19:44:41 |