In situ validation of sea surface temperatures from the GCOM‐W1 AMSR2 RSS calibrated brightness temperatures

Abstract

AbstractRemote Sensing Systems AMSR2 v7.2 data from 25 July 2012 to 9 October 2014 are collocated with in situ sea surface temperature (SST) data. The RSS SST algorithm uses AMSR2 brightness temperatures calibrated using a methodology developed at RSS rather than using the standard JAXA AMSR2 product, which includes the JAXA calibration. The new RFI exclusion methodology used for the AMSR2 v7.2 data is described. Buoy data are quality controlled using an internal quality indicator. Daytime collocations with wind speeds of less than 6 m s−1 are excluded to avoid diurnal contamination of the results. A mean bias (AMSR2 minus in situ) of −0.04 K and standard deviation 0.55 K with 109,350 collocations is found. The geographical distribution of biases is investigated, with a small increase in biases found at higher latitudes. At lower SST the uncertainty increases and the bias. The dependencies of the bias and uncertainties on other geophysical variables are shown to be negligible. The time series of the bias and uncertainty show little variability, but a small seasonal dependence is determined to be related to a seasonal shift in wind speeds. Overall, the AMSR2 SSTs are of comparable quality to the AMSR‐E SSTs and continue the climate microwave SST record that started in 1997.