The reconstruction of MODIS land surface temperature products using NSSR

Land surface temperature (LST) is a key parameter in climatological and environmental studies [1]. The Moderate Resolution Imaging Spectroradiometer (MODIS), onboard the NASA Terra and Aqua Earth Observing System satellites, can provide global temperature and narrowband emissivity data on a daily basis. However, when the surface is obscured by clouds, the variable cannot be measured directly by using satellite thermal infrared channels, which leads to many invalid value pixels in the MODIS LST products. Methods for calculating LST of the MODIS cloudy pixels are important, yet few studies have been done. The objective of this paper is to estimate the LST values of the cloudy-pixels using the neighboring-pixel approach (NP) and MODIS NSSR (net surface shortwave radiation) product. In this study, the Heihe River Basin was selected as a case study area. The estimation was validated using ground-measured data of Huazhaizi (HZZ) desert station which is covered by homogeneous desert steppe. The validation shows that the reconstruction values of MODIS LSTs can agree well with the ground-measured data, and the biggest absolute error is 2.6K.