Evaluation of AIRS and CrIS SST measurements relative to three globally gridded SST products between 2013 and 2019.

Globally gridded sea surface temperatures (SST) are useful for providing boundary conditions for atmospheric models and for the long-term monitoring of satellite derived SSTs. The Real Time Global (RTG) SST from NOAA has been used for the past decade to evaluate the stability and accuracy of the AIRS calibrated radiances since 2002. In order to find an alternate and improved SST product, we evaluated SST products provided by the Canadian Meteorological Centre (CMC) and the UKMetOffice (OSTIA). We analyzed the mean, standard deviation and trend of the daily SST derived from clear sky AIRS and CrIS radiances relative to daily matchups with gridded SST products for the 2013 and 2019 period. During this time the standard deviation of the difference between AIRS and the RTG was 0.55K and increasing. In contrast, the standard deviation of the difference between the AIRS, the CMC and the OSTIA dropped steadily to below 0.4 K, a level previously only seen in SST products relative to the Argo buoys. Unexplained biases between the observed and the gridded SSTs at the 100 mK level are consistent with already existing estimates of the AIRS and CrIS absolute calibration accuracy. However, the AIRS and CrIS observations also independently confirm artifacts in all three SST products, increasing with distance from the equator, with the CMC artifacts being the smallest. We selected the CMC as the replacement for the RTG. In the 2013-2019 data, the CMC artifacts for the 30S-30N oceans create a 4 mK/yr trend artifact.