基于GIS的降雨空间插值研究——以湖南省为例
阳宽达1,3, 谢红霞1, 隋 兵2, 周 清1, 刘 沛1, 王海涛1

(1.湖南农业大学 资源环境学院, 长沙 410128; 2.气象防灾减灾湖南省重点实验室, 长沙 410007; 3.中南大学 地球科学与信息物理学院, 长沙 410083)

普通克里金插值; 协同克里金插值; 降雨; 插值精度

Research on Spatial Interpolation of Rainfall Based on GIS-A Case Study of Hunan Province
YANG Kuanda1,3, XIE Hongxia1, SUI Bing2, ZHOU Qing1, LIU Pei1, WANG Haitao1

(1.College of Resources and Environment, Hunan Agricultural University, Changsha 410128, China; 2.Hunan Key Laboratory for Meteorological Disaster Prevention and Reduction, Changsha 410007, China; 3.School of Geosciences and Info-Physics, Central South University, Changsha 410083, China)

ordinary Kriging interpolation; CoKriging interpolation; rainfall; interpolation accuracy

备注

降水数据是区域水文模拟、水资源分析和管理及地质灾害预警等方面的基础数据,提高降雨数据的插值精度具有重要的理论和现实意义。以湖南省为研究区,结合站点观测降雨数据、TRMM雷达降雨数据及DEM数据,采用克里金插值和协同克里金插值两种方法进行了降雨插值,通过对比研究降雨插值精度。结果 表明:(1)引入与降雨空间信息相关的TRMM,DEM数据的协同克里金插值能够提高插值精度,以TRMM,DEM为辅助变量的协同克里金插值的平均相对误差较克里金插值降低了0.02%,0.23%;(2)降雨观测数据插值结果对于TRMM,DEM这两种协变量表现出不同的灵敏度,以DEM为协变量的克里金插值的插值精度要高于以TRMM为协变量的克里金插值,研究区的降雨量与地形因素有更高的相关性。

Rainfall data are the basic for regional hydrological simulation, water resources analysis and management, geological hazard warning, etc. It is of great theoretical and practical significance to improve the interpolation accuracy of rainfall data. Hunan Province is taken as the research area. The station data, TRMM data and DEM data are used for rainfall interpolation by the ways of Kriging and CoKriging. The interpolation results are compared to study the accuracy of rainfall interpolation. The results show that:(1)CoKriging interpolation using the TRMM and DEM as data source which is related to rainfall spatial information can improve interpolation accuracy; the average relative errors of CoKriging interpolation with TRMM and DEM as auxiliary variable are 0.02% and 0.23% lower than Kriging interpolation, respectively;(2)The interpolation results of station data show different sensitivity to the auxiliary variables of TRMM data and DEM data; The interpolation accuracy of CoKriging with DEM as auxiliary variable is higher than CoKriging with TRMM as auxiliary variable. The rainfall in the study area has a higher correlation with the topographic factors.