Land surface temperature retrieval from HJ-1B satellite thermal infrared data and error analysis using partial differential equation

In this study, we retrieved the land surface temper ature (LST) of Guangzhou on Jan 14, 2013 from HJ-1B satellite data. The retrieval was based on the characteristic s of HJ-1B thermal infrared band and revised QKB The transmittance error of 0.1 resulted in LST error of around 1 K. M eanwhile, the atmospheric water vapor error and the LST error exh ibited a linear relationship; The atmospheric water vapor error of 0.1 g/cm 2 resulted in LST error of around 0.2 K. The LST ret rieval error was proportional to both the nearsurface air temperature error and the average atmos pheric error; The near-surface air temperature erro r of 1 K led to the LST retrieval error of around 1 K. Overall, at a constant ratio relation between emissivity and atmospheric transmittance, the LST retrieval error are related to the average atmospheric temperature error as wel l as the nearsurface air temperature error. The retrieved land stemperature of Guangzhou was in strong spati al accordance with the MOD11_L2 LST product. The tempe rature difference curve exhibited a normal distribu tion, concentrating in the range of -0.9K to 0.9K. Six ob servation areas in Guangzhou were chosen to compare the LST obtained by the revised QK&B algorithm with the mea sured average land surface temperature. The differe nce between the LST obtained using the algorithm and th e measured temperature was around 0.31 K, whereas t he MOD11_L2 product had a difference of around 0.65 K with the measured surface temperature, both of them are less than 1 K. By deriving the partial differential equa tions of the revised QK&B algorithm, a more detaile d and precise analysis was performed on the LST retrieval from HJ -1B/IRS data. This study offers a reference for sim ilar LST retrieval algorithms based on thermal infrared band of environmental satellites, as well as a scientific basis for future accuracy improvement of LST retrieval.