Modeling spatio-temporal distribution of soil moisture by deep learning-based cellular automata model

作     者：SONG Xiaodong ZHANG Ganlin LIU Feng LI Decheng ZHAO Yuguo YANG Jinling 

作者机构：State Key Laboratory of Soil and Sustainable AgricultureInstitute of Soil ScienceChinese Academy of Sciences 

出 版 物：《Journal of Arid Land》 (干旱区科学（英文版）)

年 卷 期：2016年第8卷第5期

页      面：734-748页

核心收录：

学科分类：09[农学] 0903[农学-农业资源与环境] 090301[农学-土壤学] 

基　　金：supported by the National Natural Science Foundation of China (41130530,91325301,41401237,41571212,41371224) the Jiangsu Province Science Foundation for Youths (BK20141053) the Field Frontier Program of the Institute of Soil Science,Chinese Academy of Sciences (ISSASIP1624) 

主　　题：soil moisture soil moisture sensor network macroscopic cellular automata (MCA) deep belief network (DBN) multi-layer perceptron (MLP) uncertainty assessment hydropedology 

摘      要：Soil moisture content （SMC） is a key hydrological parameter in agriculture,meteorology and climate change,and understanding of spatio-temporal distributions of SMC in farmlands is important to address the precise irrigation ***,the hybrid interaction of static and dynamic environmental parameters makes it particularly difficult to accurately and reliably model the distribution of *** present,deep learning wins numerous contests in machine learning and hence deep belief network （DBN） ,a breakthrough in deep learning is trained to extract the transition functions for the simulation of the cell state *** this study,we used a novel macroscopic cellular automata （MCA） model by combining DBN to predict the SMC over an irrigated corn field （an area of 22 km^2） in the Zhangye oasis,Northwest *** and dynamic environmental variables were prepared with regard to the complex hydrological *** widely used neural network,multi-layer perceptron （MLP） ,was utilized for comparison to *** hybrid models （MLP-MCA and DBN-MCA） were calibrated and validated on SMC data within four months,*** to September 2012,which were automatically observed by a wireless sensor network （WSN） .Compared with MLP-MCA,the DBN-MCA model led to a decrease in root mean squared error （RMSE） by 18%.Thus,the differences of prediction errors increased due to the propagating errors of variables,difficulties of knowing soil properties and recording irrigation amount in *** sequential Gaussian simulation （s Gs） was performed to assess the uncertainty of soil moisture *** with a threshold of SMC for each grid cell,the local uncertainty of simulated results in the post processing suggested that the probability of SMC less than 25% will be difference in different areas at different time *** current results showed that the DBN-MCA model performs better than the MLP-MCA model,and the DBN-MCA model provides a powerful tool for predicti