HB1102 ReadMe Class of Data: Surface ocean carbon dioxide concentrations Dataset Identifier: HB1102-Legs 1-5 Spring Bottom Trawl Survey Expocode: 33HH20110301 Statement of how to cite dataset: Bigelow website: http://www.aoml.noaa.gov/ocd/gcc/rvbigelow_introduction.php For more information on dataset citation please see Master Readme File. Measurement platform identifier: NOAA Ship Henry B. Bigelow, R225 Time period the dataset refers to: 1 March, 2011 - 12 May, 2011 Geographic area the dataset refers to: Latitude, 34.4 to 44.5 degree north Longitude, 65.7 to 76.1 degree west Standard and Reference Gases: The four CO2 standard cylinders come from Scott-Marrin, Inc., and are calibrated with primary standards that are directly traceable to the WMO scale. The zero gas is ultra-high purity air. Any value outside the range of the standards should be considered approximate (+/-5 ppm). While individual data points above 520 or below 200 may less accurate, the general trends would be indicative of the seawater chemistry. The standards used on the cruise are: STANDARD TANK # [CO2]ppm VENDOR STD1 JA02645 204.74 Scott-Marrin STD2 JA02273 317.49 Scott-Marrin STD3 JA02285 372.23 Scott-Marrin STD4 JA02647 518.24 Scott-Marrin STD5 JA03093 0.00 Scott-Marrin Data quality issues: This cruise was the first deployment of the analytical system on this ship. Several of the conditions detailed below impacted operations and data quality but were resolved for later cruises. During HB1102, the real-time Sea Surface Temperature (SST) data that was appended to the pCO2 data records was for the hull mounted sensor that is close to the water line (Furuno2000-HI). After the HB1102 cruises, it was recognized that the SBE38 sensor in the sea water piping close to the main water pump provided a better estimate of the in situ SST. The ship’s underway data files for the SBE45 thermosalinograph (TSG) measurements contained SBE38 values, most of the time. When available these SBE38 values were used for the in situ SST. If there was a Furuno2000-HI value but no SBE38 value, the offset between these two sensors was determined for data during the same leg and then the offset was applied to the Furuno2000-HI value to estimate the SBE38 value. For 66 analyses between 15:00 and 17:38 on 5 March, 2011, the SST was estimated by subtracting 0.22 degrees from the Furuno2000-HI value. For over 6000 analyses on leg 1 that had values for both the SBE-38 and the Furuno2000-HI temperature sensors, the average difference was 0.22 (+/- 0.09) deg C. For 406 analyses between 18:00 on 26 March and 13:00 on 31 March, the SST was estimated by subtracting 0.20 degrees from the Furuno2000-HI value. For over 5000 analyses during legs 2 and 3 that had values for both the SBE-38 and the Furuno2000-HI temperature sensors, the average difference was 0.20 (+/- 0.09) deg C. There remained CO2 data records that did not have a value for SST. Typically, there were CO2 data records with SST preceding and following them. A value for SST was interpolated from bracketing data for 183 CO2 data records over the 5 legs of HB1102, mostly during leg 1. The interpolation is noted in the QC comments column and QC flag is changed from 2 (= good) to 3 (= questionable). Some of the CO2 data records did not have a salinity value from the ship’s SBE45 TSG appended to it in real-time. The salinity from the ship’s data files was assigned to 1647 of these CO2 data records over the 5 legs of HB1102. There remained 138 CO2 data records without a salinity value. A salinity value was interpolated from bracketing data records, and these records had the same annotation in QC and QC comment columns as the records with interpolated SST. There were gaps in the atmospheric pressure values that were appended in real-time to the CO2 data records. An attempt was made to fill these gaps using the raw data files for the SAMOS and Young barometers provided by the ship. For the approximately 2000 occurrences when the pCO2 data record and the ship's data files contained a barometer pressure value within a minute of each other, the average difference was 0.001 (+/- 0.184) mbar with no difference exceeding 1.4 mbar. The atmospheric pressure appended to the CO2 data records in real time compare well with both the SAMOS and Young barometer readings. On 24 March, 2011, at ~13:00 (UTC; ~83.55 YearDay) during leg 2 of HB1102, the barometric pressure reading increased ~4.6 mbar between CO2 analyses separated by 2.5 minutes. If one assumes that the jump in barometer is due to changing the barometer from in situ to sea surface calibration, the change in height is ~38 meters. In hopes of having the data before and after this jump comparable, the data before the jump (i.e. leg 1 and first half a day of leg 2) was processed with the height of the atmospheric barometer equal to 35 meters. The data after the jump (i.e. the last 5 days of leg 2 and legs 3-5) was processed with a barometer height of 0 meters. The data reduction software takes into account the height of the barometer and makes the appropriate adjustments to sea level pressure needed for data reduction. For leg 1 and part of leg 2 of HB1102, there were 1366 data records without atmospheric pressure values. Half of these gaps were filled with data from the ship’s barometer files. The other half of the missing barometer values were estimated by subtracting 4.23 mbar from the LICOR pressure. For 6970 of the surrounding analyses that had both LICOR and ship’s barometer values, the difference was 4.23 (+/- 0.37) mbar. For part of leg 2 and leg 3 of HB1102, there was no barometric pressure appended to 465 CO2 data records. For the surrounding 5850 records that did have barometer values, the average difference between the LICOR and barometric pressures was 0.06 (+/- 0.38) mbar; on average the LICOR pressure was smaller. The missing barometric pressure values were estimated by adding 0.06 mbar to the LICOR pressure. For legs 4 and leg 5 of HB1102, there were 339 CO2 data records whose atmospheric pressure values were obtained from the ship’s data files for the SAMOS and Young barometers. There were 616 CO2 data records whose atmospheric pressure values were estimated by adding 0.11 to the LICOR pressure value. For the surrounding 9163 records with LICOR and atmospheric pressure values, the average difference was 0.11 (+/- 0.33) mbar. When the system was started back up on 21 March for leg 2 (YearDay 80.8), the condenser temperature reading (-15 C) was much lower than the normal 5 deg C. The gas flow rates for the EQU and ATM analyses decreased to essentially zero during the first 18 hrs of operation - possibly due to ice formation. Those conditions persisted from YearDay 81.5 until YearDay 84.9, when the system was shutdown. Most of these sample analyses were deleted because of low sample gas flow. When the analytical system was started up on 26 March (YearDay 85.7) the gas flows were restored. Optimal pCO2 measurements with this type of analytical system can be made when the sea water flow rate is steady and between 1.5 and 2.5 L/min. The sea water flow was rarely in this range during HB1102. During legs 1 and 2, the water flow was often between 1.0 and 1.7 L/min, but there were only a few intervals when the flow rate remained above 1.0 L/min. The difference between the SST and the equilibrator temperature values was quite variable and averaged between 0.6 and 0.8 deg C. For legs 3 through 5, the sea water flow rate appeared to have two patterns, both with less variability than legs 1 and 2. The average water flow was mostly between 0.2 and 0.4 L/min but was occasionally between 0.6 and 0.9 L/min. The difference between the SST and the equilibrator temperature values was more variable than previous legs and averaged between 0.7 and 1.1 deg C. No relationship between the water flow rate and the temperature difference could be recognized. Under optimal conditions, the temperature difference should be less than 0.2 C. The differential pressure transducer on the equilibrator displayed greater variability than optimal. Most of the differential pressure (del-P) values were between 1 and -1 mbar; the headspace of the equilibrator was within 1 mbar of the ambient barometric pressure. For essentially all 5 legs, the del-P values were scattered between 1.5 and -9.5 mbar. For the last 4 days of leg 4, the del-P values were scattered between 1.1 and -1.7 mbar. No relationship between the sea water flow rate and the del-P variability or value could be recognized. Patterns in the calculated fugacity of CO2 in water can be related to patterns in the salinity and/or SST values, as expected. The low and variable sea water flow rates, however, should caution interpretations of the CO2 fugacity data without close examination. The sea water flow rate was included in the data records to allow a closer scrutiny of its impact. Timestamp for dataset reduction: 11 August, 2011 Timestamp for most recent update of dataset: 12 August ,2011 Matlab version used for reduction: 1.10 Contact person for this dataset: Kevin Sullivan (305)361-4382 Kevin.Sullivan@noaa.gov