安徽省热量资源空间分布模型及插值方法研究

Establishment of spatial distribution models and spatial interpolation for thermal resources in Anhui Province

【目的】对安徽省热量资源进行栅格化处理，解决无测站地区的热量资源估算问题，为区域农业气候资源评估提供依据。【方法】利用安徽省1971～2000年气象站地面资料，在地理信息系统技术（GIS）支持下，采用多元回归结合空间插值方法，建立安徽省热量资源要素空间分布模型，并比较反距离权重法（IDW）、克里格和样条函数3种空间插值方法的精度。【结果】基于GIS平台，建立了安徽省250m×250m分辨率的温度和积温栅格化数字地图。对空间插值方法的交叉验证结果表明，3种空间插值方法的平均绝对误差、平均误差平方的平方根排序均为IDW克里格法〈样条函数法，IDW法效果最优。【结论】应用GIS空间内插技术，构建高分辨率的栅格化热量资源要素集，可有效提高农业气候资源分析和利用的精细化程度。

[Objective] Spatial interpolation of climate data is frequently required to provide regional agricultural climate resource assessment and it can effectively solve the estimation of thermal resource in the area without observation station. [Method] Making full use of the meteorological observation data from 1971 to 2000,the spatial analysis models of heat factors; were built up in virtue of SPSS. The heat factors included the annual mean temperature, and the accumulated temperature with characteristic temperature. [Result] On the basis of ArcGIS 9.2,the digital grid maps were formed on the scale of 250 m×250 m grid by applying the spatial distribution model. In this paper, based on regression analysis combined with spatial grid interpolation, the spatial interpolation methods of Inverse Distance Weight（IDW）, Kriging, Spline and combined method were utilized for the comparison studdy on spatial interpolation of heat factors from 1971 to 2000 in Anhui province. Based on the mean absolute errors（MAE） and Root Mean Squared Interpolation Error （RMSIE） from cross-validation tests,the methods were ranked as IDW〈Kriging〈Spline for interpolating heat factors. [Conclusion] On the whole, IDW brought lowest errors. The high-resolution digital grid maps of heat factors were formed with the help of GIS spatial technology and could improve precision and accuracy of agricultural climate resource evaluation.