Data was read from glider stored memory after recovery.
Quality control (QC) steps applied to CTD data (bad flag means values were replaced with NaN in the matlab files):
1. data when glider was in testing modes or at the surface were bad flagged.
2. obvious spikes in temperature and conductivity were bad flagged based on a difference threshold
3. salinity errors were significant due to the unpumped conductivity cell and were corrected following Garau et al., 2011. Matlab scripts available at http:// www.socib.es/;glider/doco/gliderToolbox/ctdTools/ thermalLagTools) were used with modified parameters as specified in the data file under config.thermal_lag_correction_parameters to create salinCorrected and densCorrected.
4. poor flushing of the conductivity cell and thus invalid salinity data were found to occur when the glider forward speed fell below 10 cm/s or the pitch was between +5 and +15 degrees (on an upcast). All CTD output values were bad flagged during these times.
Salinity (salin) was calculated from conductivity using PSS-78, as well as water density (dens) using the seawater routines.
Normally, density should not be archived with data submission to national repository. However since this parameter is integral to the main study, density is being reported with other measured parameters but with caveat that it is derived based on older conventions.
BCO-DMO Processing:
-extracted data from MatLab .mat files
-added conventional header with dataset name, PI name, version date, reference information
-renamed parameters to BCO-DMO standard
-added yrday_utc and ISO_DateTime_UTC to served view
-reduced number of significant digits