The development of Microwave Vegetation indices according to WindSat data

As an important vegetation indicator, vegetation indices have become a widely used tool in vegetation parameters retrieval and condition monitoring. A newly developed MVI was deduced and evaluated using WindSat data. During the deduction, the w-τ model was utilized as the theory foundation. The emission from ground can be rearranged into a two component model including the vegetation emission component and the vegetation transmission component. In order to characterize the frequency dependence of surface emission signals on the objective of minimizing the effects of the ground surface emission signals, we have built a simulation database for the configurations of WindSat using the Advanced Integral Equation Model (AIEM) at 6.8, 10.7, and 18.7 GHz, dual-polarization (v and h) and the corresponding incidence angles. Unlike previous MVIs, this simulation contains both Gaussian and Exponential correlation surfaces. On the basis of simulation data analysis, we found that bare soil emissivity at two adjacent WindSat frequencies has a linear relationship, which makes it possible to minimize the surface emission signal and maximize the vegetation signal. As a result, brightness temperature at a higher frequency can be a function of the adjacent lower frequency at the same polarization, whose slope and intercept are the newly developed Microwave Vegetation Index (MVI) from WindSat data. The new MVI shared the same vegetation distribution pattern as AMSR-E based MVIs and was also negative to NDVI.