Development of View Angle Dependent Split-Window Algorithm for Retrieving Land Surface Temperature From Modis Thermal Infrared Data

The study focuses on development of a method for retrieving Land Surface Temperature (LST) using MODIS (MODerate Resolution Imaging Spectroradiometer) thermal infrared (TIR) band-31 (10.78-11.28µm) and band-32 (11.7712.27µm) data through view-angle dependent split-window algorithm for atmospheric correction for the retrieval. For simulating the at-sensor radiance, we have used atmospheric Radiative Transfer Model (RTM) MODTRAN 5.3 (MODerate resolution atmospheric TRANsmittance). A large numbers of theoretical simulations were carried out for different land covers along with many tropical atmospheric profiles pertaining to Indian landmass for varying view zenith angles in two TIR bands. The Split Window (SW) coefficients were generated from this modeling exercise and were used in retrieval process of LST. The retrieved LST images generated by our method were compared with the standard LST products (MOD11C1 LST) from the MODIS. The results indicated that the proposed method showed a good match with the reference MODIS products having a high correlation (with R
²
range 0.71-0.99) and Root Mean Square Error (RMSE) below 3.25K with minimum bias of 0.023 K.