Physical oceanography data from a Seaglider mission west of Svalbard, October 2020 - February 2021.

• Main Author: Fer, Ilker
• Format: Dataset
• Language: unknown
• Published: University of Bergen 2021
• Subjects: Svalbard; Sea ice
• Online Access: https://dx.doi.org/10.21335/nmdc-1878084716; http://metadata.nmdc.no/metadata-api/landingpage/a0dbad151c9d9d7509f898f3833da152

The Seaglider is sg562 (WMO ID 6801612). SBE-CT sail (SN 191) was last calibrated on 3 November 2017. The optode AA4330 (SN 797) was last calibrated on 29 January 2018. The glider was deployed on 7 October 2020, went into recovery mode on 8 February 2021, and was recovered on 10 February 2021. The last good profile was on 7 February. The sampling rate was variable. In the first 129 dives, sampling rate was 10 s in the entire 1000 m. Later on, 10 s in the upper 600 m and 15 s below 600 m in dives 130 to 173, and 13 s in the upper 600 m and 20 s below 600 m for the rest of the mission (dives 174 to 625). The data are processed using the University of East Anglia Seaglider toolbox (http://www.byqueste.com/toolbox.html) downloaded on 25 March 2021 (most recent toolbox change prior to download was 15 September 2020). The processing is based on the methods described by Garau et al. (2011) for the thermal lag correction of the conductivity cell, and Frajka-Williams et al. (2011) for the flight model regression. TEOS-10 is used (IOC, SCOR, and IAPSO 2010). Hard limits on salinity and temperature were applied to exclude obvious outliers before the flight model regression. Variable pitch and roll dives are selected for the flight model regression and 30 m range from the surface and apogee are excluded. Correction coefficients for the thermal lag of the conductivity cell are obtained minimizing the area between the up and down temperature-salinity curves as described in Garau et al. (2011). Improved salinity values are then obtained accounting for the thermal lag. Unfortunately, salinity profiles obtained from this mission are particularly problematic, despite the best estimates for flow past sensor and thermal lag corrections. Substantial manual flagging was applied to salinity and density profiles to remove spikes and unstable overturns appearing in density profiles. After manual flagging, each profile is despiked by flagging values exceeding twice the rms value of (x – xs), where x is the profile data and xs is a 5-point median filter. In addition, data in the upper 5 m in dives and 1 m in climbs, and in the bottommost 5 m in dives and climbs are flagged. All flagged data, either detected automatically or manually, are replaced by NaNs. The density profile is sorted (retaining NaNs), and salinity is obtained from the gravitationally stable density profile. The user is warned about this nonstandard approach. Normally, processed Absolute Salinity and Conservative Temperature are accurate to 0.01 g kg-1 and 0.001C, respectively, and DAC is accurate to 0.01 m s-1 (p. 9 “Seaglider Quality Control Manual,” 2012). However, we expect unquantified uncertainties in salinity at finescale (1-5 m) because it is calculated from the sorted density profile. Furthermore, DAC values for the dives under sea ice (dives 422-425) and dives missing GPS fix (230-231, 235-236) are not accurate and obtained by linear interpolation of the last known predive and postdive positions.