One of the main parameters which determines the state of aquatic ecosystems is water temperature. Determination of the temperature of the underlying surface by remote sensing methods is carried out using scanning in the infrared range of spectrum. Many researchers have faced the issue of uncertainty in the sea surface temperature (SST) determination from remote sensing data. As part of the study of climatic variability of the thermohydrodynamic regime of the Caspian Sea based on remote sensing data, studies were carried out to refine algorithms for calculating SST using data from Landsat-5 and Landsat-7 satellite products. A correct determination of the SST is crucial for monitoring thermal and ice regime of the Caspian Sea, evaporation rates, algal bloom events, ecosystem functioning, stocks status of Caspian sturgeon, ctenophore invasion, and even the rate of evaporation of oil spills at the sea surface, which have their own negative impact on the Caspian Sea ecosystem. To calibrate satellite data, in-situ measurements of SST at SVPB drifters deployed in the Caspian Sea in 2006-2008 were used. Synchronous sub-satellite measurements of SST were selected and compared with satellite data obtained from thermal channels of Landsat-5 and Landsat-7 (Level-1 and Level-2 products from Collection 2). The results of this study allow us to recommend using SST obtained from the Landsat Level-2 product with a correction of +1°C. The study also evaluates the possible error when using brightness temperature at the upper boundary of the atmosphere obtained from Landsat Level-1 product data. These data do not take into account the influence of the atmosphere and the state of underlying surface. In this case, the average error in determining SST was significantly higher; for the Caspian Sea it was 5.1°C.
