Drift Characteristics of a Moored Conductivity–Temperature–Depth Sensor and Correction of Salinity Data

Abstract

The temperature and conductivity drift (time change of the characteristics) of moored SBE37IM conductivity and temperature (CT) sensors was investigated by pre- and postdeployment calibration of the Triangle TransOcean Buoy Network (TRITON). This buoy network comprises the western portion of the basinwide (Tropical Atmosphere Ocean) TAO/TRITON buoy array, which monitors phenomena such as El Niño and contributes to forecasting climate change. Over the time of deployment the drift of the temperature sensors was very small, within 3 mK of the postdeployment calibration data. The drift of the conductivity sensors was more significant. After 1 yr of mooring, conductivity drift observed in the shallowest layer (1.5?100 m) was positive and 0.010 S m?1 [equivalent to 0.065 (PSS-78) at 30°C and 6 S m?1; here, 1 S is 1 Ω?1] at 6 S m ?1 on average. Drift observed in the thermocline layer (125?200 m) was also positive and 0.0053 S m?1 [0.034 (PSS-78)] at 6 S m?1 on average. Conversely, the drift of conductivity in the deepest layer (250?750 m) was 0.00002 S m?1 with a standard deviation of 0.001 S m?1 [0.0065 (PSS-78)]. Assuming a linear trend of conductivity drift with time, the authors attempted to correct the conductivity data using the postdeployment calibration data. The corrected data for about 80% of the sensors exhibited smaller differences than the uncorrected data when compared with the in situ conductivity?temperature?depth (CTD) data. However, the corrected salinity data became worse than the uncorrected data for about 20% of the sensors. The reasons for these errors are also discussed in this paper.