基于雨量计的城区降雨插值不确定性分析  

Uncertainty Analysis of Urban Rainfall Spatial Interpolation Based on Gauges Data

作　　者：戴培培[1] 靖常峰[1,2] 杜明义[1,2] 

机构地区：[1]北京建筑大学测绘与城市空间信息学院,北京100044 [2]北京建筑大学现代城市测绘国家测绘地理信息局重点实验室,北京100044

出　　处：《人民黄河》2015年第4期38-42,共5页Yellow River

基　　金："十二五"国家科技支撑计划项目(2012BAJ14B03);北京市自然科学基金资助项目(8122017)

摘　　要：以北京市西城区2013年6月23日自动雨量计实测数据为研究样本,采用反距离权重法、普通克里金法、径向基函数法等空间插值方法及交叉验证检验方法,以最小均方根误差为插值精度评价指标,分析了站点数量、站点密度、站点分布差异以及不同插值方法对城区降雨插值精度的影响。结果表明:1在城区插值分析中,站点数量及站网密度对插值精度影响较大,而站点分布影响较小,当雨量站点采样率不低于50%、站点密度选择中度分布时可得到较稳健的插值结果;2普通克里金法更适用于城区降雨空间插值。This paper took Beijing Xicheng District as a study region and adopted the real-time rainfall data,which was collected by rain gauges on June 23,2013,as samples to do the experiment. The performances of interpolation methods(inverse distance weighting,ordinary kriging and radial basis functions)were assessed and compared by using the root mean square error for cross validation. The study analyzed the impact on urban rainfall interpolation precision by rain gauge number,density and distribution and analyzed the uncertainty in urban rain-fall interpolation. The results show that:a)the rain gauge number and density have a greater influence on the precision of urban spatial in-terpolation;otherwise,rain gauge distribution has little influence. When the sample rate is larger than 50% and density is medium deploy-ment,we can get a more precise and robust interpolation accuracy;b)the Kriging is more suitable for urban rainfall spatial interpolation.

关 键 词：反距离权重法 普通克里金法 径向基函数法 不确定性分析 插值分析 降雨 城区 

分 类 号：P332.1[天文地球—水文科学]