Evaluation of Drifter Salinities in the Subtropical North Atlantic

• Published in: Journal of atmospheric and oceanic technology Vol. 32; no. 1; pp. 185 - 192
• Main Authors: Hormann, Verena, Centurioni, Luca R, Reverdin, Gilles
• Format: Journal Article
• Language: English
• Published: Boston American Meteorological Society 01.01.2015
• Subjects: Accuracy; Air bubbles; Algorithms; Atmospherics; Bias; Data transmission; Datasets; Drift; Drifters; Drogues; Editing; Marine; Noise levels; Noise measurement; Oceanography; Precipitation; Procedures; Quality control; Salinity; Salinity data; Salinity effects; Salinity measurements; Sampling; Satellite tracking; Satellites; Sciences of the Universe; Sea surface; Sensors; Statistical methods; Strain gauges; Studies; Surface salinity; Time series; Upper ocean; AN, North Atlantic; Data quality control; North Atlantic Ocean; Surface observations; Buoy observations; Salinity; Subtropics
• ISSN: 0739-0572; 1520-0426
• DOI: 10.1175/JTECH-D-14-00179.1

Salinity measurements from drifters constitute an important in situ dataset for the calibration and validation of the sea surface salinity satellite missions. A total of 114 satellite-tracked salinity drifters were deployed within the framework of the first Salinity Processes in the Upper Ocean Regional Study (SPURS) experiment in the subtropical North Atlantic focusing on the period August 2012-April 2014. In this study, a subset of 83 drifters, which provided useful salinity measurements in the central SPURS region from a few weeks to more than one year, is evaluated and an ad hoc quality-control procedure based on previously published work and the new observations is described. It was found that the sampling algorithm of the drifters introduces a predominantly fresh bias in the noise level of the salinity data, probably caused by the presence of air bubbles within the measuring cell. Since such noise is difficult to eliminate using statistical methods, extensive editing was done manually instead. Such quality-control procedures cannot be routinely applied to the real-time data stream from the drifters. Therefore, a revision of the sampling algorithm of the drifter's salinity sensor is needed. Comparisons of the drifter's salinity measurements with independent datasets further indicate that the sensor can provide reliable observations for up to one year. Finally, little evidence was found that the quality of the drifter's salinity measurements depends on the presence of the drogue.