摘要
以温州地区1991～2001年90个观测站的月平均降雨为例,在GIS支持下,对样条函数法、克立格内插法及距离平方反比法这3种插值方法进行了对比研究;并在90个观测站中筛选出最少量的、能达到稳定结果的站点数;还考虑了插值的空间分辨率对结果的影响。对上述3点进行大量的计算、统计分析的结果表明:①在选择距离平方反比法及样条函数法对月平均降雨量插值效果均较好;②利用这两种插值方法对不同站点密度选取不同空间分辨率插值后得出,选取50站误差趋于稳定,利用距离平方反比法插值结果误差较小,而空间分辨率对结果几乎没有影响。利用距离平方反比法,选择50个站得到了温州地区1km×1km面雨量空间扩展结果,为水文预报工作提供一种较精确的面雨量估算和输入方法。
Taking the monthly mean precipitation of Wenzhou （90 stations） from 1991 to 2001 as an example,with the aid of GIS,a comparison is conducted of three interpolation methods：the spline,Kriging,and inverse distance square methods. The different station densities are used to determine the minimum number of the stations that can make the results stable,and the effect of spatial resolution on interpolation results is considered. After a great deal of calculation and statistic analysis,the results are as following：（1） the inverse distance square and spline methods are better for calculating monthly mean precipitation; （2） compared the interpolation results for the two methods applied to different station density cases with different spatial resolutions,it is shown that when selecting 50 stations,the inverse distance square method is most stable with the minimum error,and the effect of spatial resolution is very little,can be neglected. The inverse distance square method and 50 stations are selected finally to calculate the precipitation of the 1km×1km area. This method offers a good areal precipitation estimating method for hydrology prediction service.
出处
《气象科技》
北大核心
2010年第1期9-14,共6页
Meteorological Science and Technology
基金
重庆市自然科学基金项目(2007BB2193)
中国气象局武汉暴雨研究所开放基金项目(IHR2007G03)
中国气象局成都高原气象研究所开放基金项目(LPM200814)资助
关键词
面雨量
插值
空间扩展
areal precipitation
interpolate method
spatial extension
