An integrative method for land surface component temperature inversion

Based on the matrix formula L(thetas_k)_Ktimes1 =W_KtimesJLb(T)_Jtimes1, the component temperature can be inversed by inverse matrix W_KtimesJ ^-1. But the correlation of multi-angle infrared radiance was obvious, the retrievable parameters and inversion result were unsteady, it was influenced by various physical and structural parameters. In this paper, we chose the row crop as the example to discuss the problem of component inversion in detail. When the physical and structural parameters were fixed, the structural pattern of row crop such as the layered mode of canopy also affected the weight of different components significantly, and the retrievable parameters were changed correspondingly. So the appropriate structural pattern should be chosen according to the retrieving parameters and components of object. The iterative method was established to inverse component temperature with the combining inverse matrix using above-mentioned research. First, we calculated the initial value by inverse matrix method, and then gradually gained the real solution by the iterative formula. The result of simulation and field experiment showed that the method can improve the retrieving accuracy and stability remarkably