Research On Construction Of Land Surface Temperature/Vegetation Index Feature Space

The land surface temperature/vegetation index feature space has important application in quantitative soil remote sensing inversion and drought monitoring, water resources management, such as soil water content, evapotranspiration. However, the study of its feature space construction method is still relatively lacking. In this study, we take the Oklahoma state of the United States as an example, the fitting method of the dry and wet edges of the land surface temperature/vegetation index feature space is carried out, and the linear and index, logarithm, polynomial, and power functions are used to fit the dry and wet edges, respectively, and the fitting results were evaluated by the measured soil water content data. We found that the results by polynomial function fitting, r-squared is the highest in the five fitting modes, and r-squared is more than 0.66 in the dry and wet edges of the feature space; and the water content of soil surface was compared with that of soil moisture content, and the root mean square error value is the smallest. In conclusion, these results strongly suggest that the polynomial function fitting the dry and the wet edges is the best way to construct the land surface temperature/vegetation index feature space.