STUDY ON A SINGLE INTERPOLATION FUSION ALGORITHM FOR MULTISOURCE REMOTE SENSING DATA OF SOIL MOISTURE

Abstract

Remote sensing of soil moisture can provide important data for monitoring large-scale agricultural drought. Due to differences between the various sensors and inversion methods, remote sensing data from different sources are unsuitable for direct comparison and analysis. Data fusion has become an area of active research regarding the application of remote sensing data. Based on the principle of cumulative distribution function matching, this study proposed a continuous relationship establishment algorithm for multisource remote sensing soil moisture data. Using this new algorithm, soil Moisture and Ocean Salinity (SMOS) and Climate Change Initiative (CCI) satellite data from the Songnen Plain as test data were fused to a long time series product of real-time remote sensing soil moisture data. This application validation of this new method to SMOS and CCI indicated that this Lagrange interpolation continuous fusion algorithm could improve the fusion accuracy of multisource remote sensing soil moisture data significantly. The low-value region of the cumulative probability distribution curve is a crucial data segment for characterization of agricultural drought. Through implementation of the proposed continuous fusion algorithm, fused SMOS and CCI data were found to have high coincidence at each quantile in the low-value region of the curve.