Retrieval of land surface temperature by cross-calibrated SVISSR thermal infrared data onboard China geostationary satellite

Abstract

A practical algorithm is developed to retrieve spatial-temporal land surface temperature (LST) using the Stretched Visible and Infrared Spin Scan Radiometer (SVISSR) time series data from China Feng-Yun 2C (FY-2C) geostationary satellite. A cross-calibration method and a general split-window algorithm for FY-2C/SVISSR data are developed. An automatic procedure is developed to implement the proposed methods for LST retrieval from SVISSR. Results from cross-calibration show that a good linear relationship between the TOA brightness temperatures from FY-2C/SVISSR and that from MODIS was found with correlation coefficients <i>R</i>² as 0.95 notwithstanding the differences of spectral response function between the two sensors. The results show that the SVISSR derived LST can be evaluated with aggregated AATSR derived LST and In-situ data. Results indicate that, the SVISSR and aggregated AATSR give comparable results (within 4K) both in Arou and YK, on the condition that AATSR LST product overestimates by about 3K than the ground measurement.