AUTHOR=Zeng Ling , Liu Quanming , Jing Linhai , Lan Ling , Feng Jun 

TITLE=Using Generalized Regression Neural Network to Retrieve Bare Surface Soil Moisture From Radarsat-2 Backscatter Observations, Regardless of Roughness Effect

JOURNAL=Frontiers in Earth Science

VOLUME=Volume 9 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/earth-science/articles/10.3389/feart.2021.657206

DOI=10.3389/feart.2021.657206

ISSN=2296-6463

ABSTRACT=The combined influence of surface soil moisture and roughness on radar backscatters has been limiting SAR’s application in soil moisture retrieval. In the past research, multi-temporal analysis and ANNs inversion of physically-based forward models were regarded as promising methods to decoupled that combined influence. But the former doesn’t ponder soil roughness change over a relatively longer period and the latter are hard to thoroughly eliminate the effect of soil roughness. This study proposes to use generalized regression neural network (GRNN) to derive bare surface soil moisture (BSSM) directly from radar backscatter observations. This method not only can derive BSSM from radar backscatters provided soil roughness unknown in any long period, but also can train models based on small-size sample data so as to reduce the manual error of training data created by simulation of physically-based models. The comparison of cross validations between BSSM-backscatter models and BSSM-roughness-backscatter models both analysed by GRNN show the incorporation of soil roughness cannot raise the prediction accuracy of models and even reduce it, indicating that combined influence thoroughly decoupled when analysed by GRNN. Moreover, BSSM-backscatter models by GRNN are recommended due to their good prediction, even compared to those related models in the past publications.