Thermal inversion based on separating the radiance contribution of atmosphere and land surface

For thermal inversion, observations of TIR are usually used to generate land surface temperature (LST) and atmosphere thermal and moisture profiles. Observations of VNIR are regarded as with little relationship with thermal parameters, as a result, few people use VNIR in thermal inversion. Zhao first used VNIR in thermal inversion and proved that VNIR data is helpful in thermal inversion. However, for land surface researchers, land surface temperature (LST) plays an important role in many applications. A physical algorithm based on quasi-Newton iteration is used on the retrieval of LST. We combined SAS (separation of land surface and atmospheric contribution) algorithm and the physical algorithm, which retrieved LST and atmosphere parameters based on pixel together. In this way, a new algorithm with both microcosmos (pixel) and macroscopy (image) considered is proposed in the paper. The retrieved result showed that the precision of LST retrieval is much higher than the result with original observations