Validation of satellite sea surface temperature analyses in the Beaufort Sea using UpTemp buoys

Many different blended sea surface temperature (SST) analyses are currently available and exhibit significant differences in the high latitude regions. It is challenging for users to determine which of these products is most accurate and best suited for their applications. Nine different SST analyses and two single sensor satellite products are compared with independent observations from Upper Temperature of the polar Oceans (UpTempO) buoys deployed in the Beaufort Sea in 2012 and 2013 during the Marginal Ice Zone Processes Experiment (MIZOPEX). The relative skill of the different SST products is evaluated using a combination of Taylor diagrams and two different verification scores that weight different statistical measures. Skill thresholds based on satellite accuracy requirements are chosen to map products with similar performance into three discrete skill categories: excellent, good, and poor. Results are presented for three subsets of the buoys corresponding to different regimes: coastal waters, northerly waters, and extreme weather. The presence of strong thermal gradients and cloudiness posed problems for the SST products, while in more homogeneous regions the performance was improved and more similar among products. The impact of variations in the ice mask between the SST products was mostly inconsequential. While the relative performance of the analyses varied with regime, overall, the best performing analyses for this region and period included the NOM Optimal Interpolation SST (OISST), the Canadian Meteorological Centre (CMC) SST, and the Group for High Resolution SST (GHRSST) Multi-Product Ensemble (GMPE). (C) 2016 Elsevier Inc. All rights reserved.