Research on Remote Sensing Geochemical Modeling Based on Surface to Surface Model

Abstract

Abstract Remote sensing geochemistry is a simple, fast and economical advanced prospecting method, which carries out inversion and prediction of surface element content using the empirical model by regression or machine learning. The key problem faced by quantitative remote sensing is the low inversion accuracy of the model due to the mismatch of “point surface” information. How to overcome this problem? This paper proposes a “surface to surface” modeling method, which converts point data into surface data through Kriging interpolation to solve this problem. This paper uses geochemical interpolation data of Cu elements at different scales in the Qishitan gold mine area, Xinjiang, and ASTER remote sensing data to conduct geochemical modeling. In order to test the effect of Kriging on decreasing the scale effect, five sets of experiments were designed for comparison. The first four sets of sample data were interpolated according to different cell sizes, and the last set of data was not interpolated. The results show that the Kriging interpolation based on the ground resolution of the remote sensing image can effectively improve the accuracy of the remote sensing quantitative inversion model. When the square interpolation is close to the ground resolution of the used remote sensing data, the modeling accuracy gets the best value. This paper provides a new idea for improving the accuracy of remote sensing geochemical modeling.