Evaluation and fusion of SST data from MTSAT and TMI in East China Sea, Yellow Sea and Bohai Sea in 2008

Abstract

Two typical satellite sea surface temperature (SST) datasets, from the Multi-functional Transport Satellite (MTSAT) and Tropical Rainfall Measuring Mission Microwave Imager (TMI), were evaluated for the East China Sea, Yellow Sea, and Bohai Sea throughout 2008. Most monthly-mean availabilities of MTSAT are higher than those of TMI, whereas the seasonal variation of the latter is less than that of the former. The analysis on the one-year data shows that the annual mean availability of MTSAT (61%) is greater than that of TMI (56%). This is mainly because MTSAT is a geostationary satellite, which achieves longer observation than the sun-synchronous TMI. The daily availability of TMI (28%–75%) is more constant than that of MTSAT (9%–93%). The signal of infrared sensors on MTSAT is easily disturbed on cloudy days. In contrast, the TMI microwave sensor can obtain information through clouds. Based on in-situ SSTs, the SST accuracy of TMI is superior to that of MTSAT. In 2008, the root mean square (RMS) error of TMI and MTSAT were 0.77 K and 0.84 K, respectively. The annual mean biases were 0.14 K (TMI) and −0.31 K (MTSAT). To attain a high availability of SSTs, we propose a fusion method to merge both SSTs. The annual mean availability of fusion SSTs increases 17% compared to MTSAT. In addition, the availabilities of the fusion SSTs become more constant. The annual mean RMS and bias of fusion SSTs (0.78 K and −0.06 K, respectively) are better than those of MTSAT (0.84 K and −0.31 K).