摘要
缺测降水数据的插补可以有效改善数据系列的完整性,以元江境内的元江、洼垤、因远、街子河、阿支、磨房河等水文和雨量站点逐月及年降水数据为基础,研究缺测降水数据的插补。站点之间月降水数据相关分析表明:各站点之间相关性较差,相关分析难以满足本研究流域内部分月降水数据插补精度,故尝试采用BP神经网络模型对研究流域降水数据进行插补。研究表明:基于本流域降水数据建立的神经网络模型检测样本合格率达到89.6%,具有较好的插补精度,说明神经网络可以用于本研究流域的缺测降水数据插补,为降水数据缺测的插补提供了新的途径。
The interpolation of missing precipitation data can improve the integrity of data series effectively.We did some research on interpolation of missing precipitation data base on hydrological and rainfall station＇s month and annual precipitation data in Yuanjiang,Wadi,Yinyuan,Jiezihe,Azhi and Mofanghe which are in Yuanjiang area.Correlation analysis among all stations showed that： correlation among all stations in the study area was weak;correlation analysis could hardly meet the interpolation precision in some months.So we tried to use the BP neural network model to interpolate the precipitation data in study area.The research showed that： the acceptable quality level of BP neural network sample test reached 89.6%,which showed that the BP neural network could be used to interpolate the missing precipitation data in the area which we had studied and it provided a new way to interpolate the missing precipitation data.
出处
《云南农业大学学报（自然科学版）》
CAS
CSCD
北大核心
2012年第2期281-284,共4页
Journal of Yunnan Agricultural University:Natural Science
基金
水利部公益性行业专项云南旱灾应急响应系统研究(201001044-1)
云南省应用基础研究面上项目(2007D210M)
云南省教育厅科学研究基金重大专项项目(ZD2009010)
云南省教育厅科学研究基金(09J0076)
关键词
BP网络
降水
相关分析
插补
BP neural network
precipitation
correlation analysis
interpolation
