Abstract: To verify the feasibility and applicable conditions of extending the Improved Radiative Transfer Equation Algorithm the differences of LandSAT8 SST parameters retrieved by Unimproved Algorithm and Improved Algorithm are compared, in this paper respectively. MODIS SST validation shows that the Improved Algorithm has significantly improved the atmospheric transmittance deviation and SST deviation except for the 2019-08-21 image while the SST deviation of the other three images is around 0.5 ℃. Buoy SST validation shows that the Improved Algorithm also has significant improvement effect on SST deviation in the nearshore sea, its bias and rmse are reduced to 0.14 ℃ and 0.18 ℃, respectively. Further research find that the SST retrieval accuracy of Improved Algorithm has a significant positive correlation R²bias(SST)=0.920 7, R²rmse(SST)=0.934 0 with the uniformity of atmospheric transmittance distribution in the study area. Using the Improved Algorithm to monitor the thermal discharge of the power plants, the surface temperature rise plume of Songyu Power Plant is not obvious. However, the surface temperature rise of Houshi Power Plant is the most significant in that it has the largest temperature rise amplitude and diffusion range, as well as the presence of a stable high temperature (>3 ℃) water body near its outlet. The surface temperature rise plume of Jinjiang Power Plant is a fan-shape distribution with a short flow axis and mainly concentrated near the outlet.