Land surface temperature vs. soil spectral reflectance fractional approach and fractional differential algorithm

Abstract

Land surface temperature plays an important role in studying the radiation and energy exchanges between Earth's surface and atmosphere. There is little litera-ture on the relationship between the land surface temperature and the soil spec-tral reflectance. This paper adopts the Grunwald-Letnikov fractional differential algorithm to reveal their relationship. The simulation results elucidate that a higher land surface temperature will result in a higher spectral reflectance. The variation characteristics of spectral reflectance curves with different land surface temperatures have basically a same trend. Fractional differential order is a good index for spectral change during the derivation process. An optimal fractional order is 6/5, which corresponds to the band of 1710 nm. As the increase of the fractional order, the number of bands increases, but when it reaches a threshold, an opposite trend is found. This study provides a new perspective for adequately excavating spectral data information, and it also can be used as a reference for precision agriculture.