Abstract:
This thesis is a contribution to estimate surface temperature (Ts) from satellite thermal infrared data. For estimate Ts by satellite we must first correct the atmospheric effects. In this framework, we propose three algorithms for estimating Ts. The first algorithm is mono-channel: it can estimate Ts from Meteosat-7 data. The second algorithm is splitwindow who depends on the integrated water vapor content (W); it can estimate Ts from the Meteosat Second Generation data (MSG-1). The third algorithm is split-window global for estimating Ts in the case where the value of W is unknown. The results were validated by theoretical simulation, by comparison with other algorithms, and also by comparison with the in-situ measurements.
