BESST - Inter-sensor Bias Estimation in Sea Surface Temperature

Context and objectives

Sea surface temperature (SST) is a fundamental parameter in oceanography, meteorology and climate research. It defines the exchange of heat and moisture between the sea surface and the atmosphere. Satellite SST measurements offer high spatial and temporal resolution measurements not available using conventional in situ SST measurements, but biases are present when comparing different SST sensors. BESST uses DINEOF (Data Interpolating Empirical Orthogonal Functions) to derive improved corrections of SST fields over the European Seas, reducing bias both in space and time. This methodology will benefit the wider community since it has been implemented at Meteo-France/CMS, a major producer of SST data in Europe and the world. This implementation is now undergoing a testing phase at CMS before becoming operational. Improved SST measurements will be used both in climatological studies as well as in weather forecasts leading to more accurate analysis.

Project outcome

Results show that using RECDIFF approach is preferred way in DINEOF based SST bias correction method as it constrains reconstruction of biases and gives improved results when compared to results without adjustment or when using optimal interpolation method in adjustment process. Although the SEVIRI SST validation results without any correction were already very good (-0.03 ± 0.52 K against AATSR and -0.05 ± 0.44 K against in situ), results were improved for RECDIFF approach both for bias (<0.02 K) and for standard deviation (0.50 K for AATSR and 0.43 K for in situ). A higher spatial and temporal resolution of the bias fields resulted in more realistic corrections.