Generation of high resolution sea surface temperature using multi-satellite data for operational oceanography

摘要

In the present article, we introduce a high resolution sea surface temperature (SST) product generated daily by Korea Institute of Ocean Science and Technology (KIOST). The SST product is comprised of four sets of data including eight-hour and daily average SST data of 1 km resolution, and is based on the four infrared (IR) satellite SST data acquired by advanced very high resolution radiometer (AVHRR), Moderate Resolution Imaging Spectroradiometer (MODIS), Multifunctional Transport Satellites-2 (MTSAT-2) Imager and Meteorological Imager (MI), two microwave radiometer SSTs acquired by Advanced Microwave Scanning Radiometer 2 (AMSR2), and WindSAT within-situ temperature data. These input satellite andin-situ SST data are merged by using the optimal interpolation (OI) algorithm. The root-mean-square-errors (RMSEs) of satellite andin-situ data are used as a weighting value in the OI algorithm. As a pilot product, four SST data sets were generated daily from January to December 2013. In the comparison between the SSTs measured by moored buoys and the daily mean KIOST SSTs, the estimated RMSE was 0.71°C and the bias value was –0.08°C. The largest RMSE and bias were 0.86 and –0.26°C respectively, observed at a buoy site in the boundary region of warm and cold waters with increased physical variability in the Sea of Japan/East Sea. Other site near the coasts shows a lower RMSE value of 0.60°C than those at the open waters. To investigate the spatial distributions of SST, the Group for High Resolution Sea Surface Temperature (GHRSST) product was used in the comparison of temperature gradients, and it was shown that the KIOST SST product represents well the water mass structures around the Korean Peninsula. The KIOST SST product generated from both satellite and buoy data is expected to make substantial contribution to the Korea Operational Oceanographic System (KOOS) as an input parameter for data assimilation.