Improving model performance in mapping cropland soil organic matter using time-series remote sensing data

作     者：Xianglin Zhang Jie Xue Songchao Chen Zhiqing Zhuo Zheng Wang Xueyao Chen Yi Xiao Zhou Shi Xianglin Zhang;Jie Xue;Songchao Chen;Zhiqing Zhuo;Zheng Wang;Xueyao Chen;Yi Xiao;Zhou Shi

作者机构：Institute of Applied Remote Sensing and Information TechnologyCollege of Environmental and Resource SciencesZhejiang UniversityHangzhou 310058China Department of Land ManagementZhejiang UniversityHangzhou 310058China ZJU-Hangzhou Global Scientific and Technological Innovation CenterZhejiang UniversityHangzhou 311215China Institute of Digital AgricultureZhejiang Academy of Agricultural SciencesHangzhou 310021China Key Laboratory of Spectroscopy SensingMinistry of Agriculture and Rural AffairsHangzhou 310058China 

出 版 物：《Journal of Integrative Agriculture》 (农业科学学报（英文版）)

年 卷 期：2024年第23卷第8期

页      面：2820-2841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：the National Natural Science Foundation of China(U1901601) the National Key Research and Development Program of China(2022YFB3903503) 

主　　题：cropland soil organic matter digital soil mapping machine learning feature selection model averaging 

摘      要：Faced with increasing global soil degradation,spatially explicit data on cropland soil organic matter(SOM)provides crucial data for soil carbon pool accounting,cropland quality assessment and the formulation of effective management *** a spatial information prediction technique,digital soil mapping(DSM)has been widely used to spatially map soil information at different ***,the accuracy of digital SOM maps for cropland is typically lower than for other land cover types due to the inherent difficulty in precisely quantifying human *** overcome this limitation,this study systematically assessed a framework of“information extractionfeature selection-model averagingfor improving model performance in mapping cropland SOM using 462 cropland soil samples collected in Guangzhou,China in *** results showed that using the framework of dynamic information extraction,feature selection and model averaging could efficiently improve the accuracy of the final predictions(R^(2):0.48 to 0.53)without having obviously negative impacts on *** the dynamic information of the environment was an efficient way to generate covariates that are linearly and nonlinearly related to SOM,which improved the R^(2)of random forest from 0.44 to 0.48 and the R^(2)of extreme gradient boosting from 0.37to *** recursive feature selection(FRFS)is recommended when there are relatively few environmental covariates(500).The Granger-Ramanathan model averaging approach could improve the prediction accuracy and average *** the structures of initial prediction models are similar,increasing in the number of averaging models did not have significantly positive effects on the final *** the advantages of these selected strategies over information extraction,feature selection and model averaging have a great potential for high-accuracy soil mapping