数据去噪处理在土壤侵蚀模拟分析中的应用  

Application of data denoising on analysis results of soil erosion simulation

作　　者：杨茹珍 张风宝 杨明义 李占斌[2,3] YANG Ru-zhen;ZHANG Feng-bao;YANG Ming-yi;LI Zhan-bin(State Key Laboratory of Soil Erosion and Dryland Farming on the Loess Plateau,Institute of Soil and Water Conservation,Northwest A&F University,Yangling 712100,China;Institute o f Soil and Water Conservation,CAS and MWR,Yangling 712100,China;Institute of Water Resources and Hydro-Electric Engineering,Xi'an University of Technology,Xi'an 710048,China)

机构地区：[1]西北农林科技大学水土保持研究所黄土高原土壤侵蚀与旱地农业国家重点实验室,陕西杨凌712100 [2]中国科学院-水利部水土保持研究所,陕西杨凌712100 [3]西安理工大学水利水电学院,陕西西安710048

出　　处：《泥沙研究》2021年第2期47-53,14,共8页Journal of Sediment Research

基　　金：国家重点研发计划项目(2016YFC0402406);国家自然科学基金项目(41877080,41671281)。

摘　　要：对已发表的土壤侵蚀数据采用信噪比进行去噪处理,对比分析去噪前后数据在研究影响因子对因变量影响的显著性、贡献率及拟合优度等方面的差异,确定去噪方法对土壤侵蚀相关试验的适用性。结果显示:去噪前后的数据都能识别出土壤侵蚀速率、径流速率、流速及径流输移能力等因变量变化的主导因子,但去噪处理能增加影响因子对观测变量的贡献率,降低误差项的贡献率,明显增大预测值与实测值之间的决定系数、Nash-Sutcliffe效率系数及相对均方根误差,减小预测值与实测值之间的绝对误差和相对误差。说明去噪处理能有效提高影响因子对观测因变量的解释程度,减小试验误差对试验结果的影响,确保试验结果的准确性。The signal-to-noise ratio( S/N) was applied in data denoising of published soil erosion data,and the differences in the significance,contribution rate and fitting degree of influence factors on dependent variables obtained from before and after data denoising were compared and analyzed to determine the applicability to soil erosion tests. Results show that the main influence factors,such as erosion rate,runoff rate,flow rate and sediment transport capacity,could be identified by the data before and after denoising. However,the data denoising could increase the contribution rate of influence factors to the observed variables,reduce the contribution rate of the experiment errors,significantly increase the determination coefficient,Nash-sutcliffe efficiency coefficient and relative root mean square error between the predicted and measured values,and significantly decrease the absolute error and relative error between the predicted and observed values. Results show that the denoising method can improve the interpretation degree of influence factors to the observed dependent variables,reduce the influence of test errors on test results,and further ensure the accuracy of test results.

关 键 词：田口方法 信噪比 数据去噪 土壤侵蚀 

分 类 号：S157[农业科学—土壤学]