Land surface temperature retrieval from FY-3E/MERSI using an optimized water vapor scaling method

This paper presents a new temperature and emissivity separation (TES) algorithm for retrieving land surface temperature (LST) from the MEdium Resolution Spectral Imager-LL (MERSI) onboard Fengyun-3E (FY-3E), the first satellite of the FY-3 series placed in a dawn-dusk, sun-synchronous orbit. An optimized Water Vapor Scaling (WVS) method was proposed for atmospheric correction, scaling reanalysis water vapor profiles on a pixel-by-pixel basis for both graybody and nongraybody pixels without the need for external land surface emissivity (LSE) data as input. In situ validation showed the retrieved FY-3E/MERSI LST achieved a bias of −0.61 K and an RMSE of 2.24 K at the SURFace RADiation budget (SURFRAD) network sites. Cross-comparison with the LST product of Geostationary Operational Environmental Satellite-16 (GOES-16) showed a high degree of spatial distribution consistency, with an average bias and RMSE of −0.21 K and 1.64 K, respectively. Simulation experiments showed that for nongraybodies, the optimized WVS method more effectively improved the accuracy of the TES-derived LST compared to the traditional interpolation-based WVS method. This study holds the potential to offer valuable insights into improving atmospheric correction for remotely sensed TIR data.