Abstract: Nuclear power industry developed rapidly in China in recent years and the environmental problems that followed have attracted much attention. In this study, three kinds of data source, MODIS thermal data, HJ-1B infrared scanner data and Landsat-8 thermal infrared data, which were imaged at simultaneous phase and different spatial resolution, were used to calculate sea surface temperature in the waters near the Hongyanhe nuclear power plant. Based on the corrected gulf average temperature method, the thermal plume distribution of nuclear power plants with three kinds of data source was obtained. Analysis shows that: Landsat-8 data achieves the highest accuracy and more details in thermal plume distribution. For restriction of the "mixed pixel" effect caused by low spatial resolution, HJ-1B and MODIS data have a bad performance. According to the differences in hydrogeological conditions between different monitoring areas, it’s wise to use satellite data with different time phase and spatial resolution comprehensively. It can be foreseen that unmanned aerial vehicles with infrared payload is expected to be a powerful complement to satellite remote sensing monitoring methods.