An enhanced semi-empirical method to estimate land surface temperature from AMSR2 observation

Land surface temperature (LST) is an important parameter in many research fields. Compared to the thermal infrared (TIR) remote sensing, passive microwave (MW) remote sensing can better overcome the atmospheric influences and has advantages in LST estimation. However, there are still many problems in estimating LST by MW: traditional empirical methods mainly rely on the statistic relation; therefore, their accuracies are generally limited; physical methods are not suitable for wide applications because they need to be constructed based on complicated surface cases. Based on the optimal time-window fitting the MW radiation transfer (RT) equation, this paper facilitates the semi-empirical method to estimate LST over the Chinese landmass from the Advanced Microwave Scanning Radiometer-2 (AMSR2) data. The results show that the method has higher accuracy than the traditional semi-empirical method. The study is beneficial for estimating LSTs in cloudy conditions and merging the TIR and MW LST in difference spatial scales.