Split Window Algorithm Calibration and Validation for TASI Sensor

In this work, we present the calibration and validation method we have applied in order to retrieve the split window (SW) coefficients for land surface temperature (LST) estimations from Thermal Airborne Spectrographic imager (TASI). For calibration and validation two different datasets has been used, both extracted from SeeBor V5.0 training dataset. The coefficients have been retrieved by a multiple regression analysis and MODTRAN simulations. For the radiative transfer experiment, we considered seven different viewing angles in a range between 0° and 60° with a step of 10°. Simulations have been performed considering all TASI channel combinations and the sensor spectral response functions. Preliminary results are presented for best band combinations suitable for SW algorithm application; these are channel 19 (10.034 gm) with 28 (11.024 gm), and channel 29 (11.134 gm) with 31 (11.354 gm). Finally, validation of the LST retrievals presents a RMSE lower than 0.6 K for both band combinations.