Practical split-window algorithm for retrieving land surface temperature over agricultural areas from ASTER data

A practical split-window algorithm which involves two parameters (transmittance and emissivity) utilized to retrieve land-surface temperature over agricultural areas from the Advanced Spaceborne Thermal Emission and Reflection Radiometer data is presented. First, by calculating the relationship between thermal radiation intensity and temperature, the Planck function is simplified using exponential function which is applied to deduce the split-window algorithm. Second, how to obtain transmittance from water vapor content and the method for estimating emissivity using normalized difference vegetation index are discussed in detail. Sensitivity analysis demonstrates that the algorithm is not sensitive to these two parameters. Finally, a standard atmospheric simulation method has been used to validate the proposed algorithm, and comparison between the algorithm and the prior study has been carried out. The results indicate that the average accuracy is 0.32 K for the case without error in both transmittance and emissivity, which is better than the prior algorithm. The accuracy is also 0.32 K when the transmittance is computed from the water content by piecewise cubic polynomial fit. The accuracy is about 0.30 K similar to 0.33 K corresponding to different Pv (Pv is the proportion of vegetation) values, which indicates that this algorithm is suitable for different land surface types over agricultural areas. (C) 2014 Society of Photo-Optical Instrumentation Engineers (SPIE)