PhD defense : Développement méthodologique pour estimer la température de surface terrestre à partir des données infrarouge thermique hyper-spectrales

PhD defense: Xinke ZHONG

Team: TRIO

Title: Methodological development for retrieving LST from hyperspectral thermal infrared data

Abstract: We first developed a flexible multi-channel method for retrieving land surface temperature (LST) from hyperspectral thermal infrared data for high emissivity surfaces. In comparison with Metop-a/AVHRR SST product, the error of the LST retrieved by the multi-channel method for high emissivity surfaces from Metop-a/IASI data is 0.43 K. Then we extended the developed multi-channel method to natural land surfaces. Finally, we developed an improved physics-based temperature unmxing (TUM) method to jointly retrieve subpixel temperatures and abundances from hyperspectral thermal infrared data for heterogeneous surfaces with large variation of subpixel temperatures. The developed TUM method was analyzed extensively using simulation data.

This thesis was supervised by Zhao Liang Li, senior researcher in the ICube laboratory of Strasbourg.

The PhD defense (in english) will be held on Thursday 23th June 2017 at 10.00am in the B248 room of the pole API building in Illkirch.