违反了 PRIMARY KEY 约束 'PK_t_counter'。不能在对象 'dbo.t_counter' 中插入重复键。 语句已终止。 基于西南地区台站降雨资料空间插值方法的比较-Comparison of spatial interpolation methods based on monthly precipitation in Southwest China
doi:  10.3878/j.issn.1006-9585.2017.17076
基于西南地区台站降雨资料空间插值方法的比较

Comparison of spatial interpolation methods based on monthly precipitation in Southwest China
摘要点击 78  全文点击 22  投稿时间:2017-04-28  修订日期:2017-08-02
查看HTML全文   下载PDF   查看/发表评论  下载PDF阅读器
基金:  国家重点研发计划项目课题(2016YFA0602401)
中文关键词:  降雨量 空间插值 空间自相关 交叉验证 变异函数模型
英文关键词:  Precipitation,Spatial interpolation, Spatial autocorrelation,Cross validation,Semivariogram model
     
作者中文名作者英文名单位
李金洁lijinjie中国科学院大气物理研究所
王爱慧wangaihui中国科学院大气物理研究所
引用:李金洁,王爱慧.2018.基于西南地区台站降雨资料空间插值方法的比较[J].气候与环境研究
Citation:lijinjie,wangaihui.2018.Comparison of spatial interpolation methods based on monthly precipitation in Southwest China[J].Climatic and Environmental Research(in Chinese)
中文摘要:
      摘 要: 以西南地区93个台站降雨观测资料中1996-2000年的月均降雨量为基础数据,对各月降雨量进行空间自相关性,变异特征等空间分析后,采用反距离加权法(IDW)和以三种(指数模型、球面模型和高斯模型)变异函数模型为基础的普通克里金(0-Kriging)两种方法进行空间插值,通过交叉验证结果进行两种方法的分析比对。结果表明:①西南地区月均降雨量存在明显的空间集聚现象,并且具有显著的空间自相关性和变异特征,可对该研究区域降雨资料进行空间插值研究。②在0-Kriging插值时,变异函数选用指数模型的效果最好,球面模型次之,高斯模型最差。③两种方法对月均降雨量、月降雨量极大和极小值插值时,0-Kriging的插值误差均小于IDW,并且插值误差总体上与降雨量呈正相关关系。对月均降雨量剔除各月降雨量极大值较为集中的站点后进行插值,两种方法插值结果的误差均明显降低。④对研究区域整体来说,0-Kriging的插值效果优于IDW,但就单个站点来看,结果并非如此。因此,本研究建议对某个特定的站点,需要用两种方法进行插值后才能比较方法的优劣。在降雨量的空间插值中,由于研究区域和时间尺度的不同,并不存在绝对的最优方法,应比较实际效果后选择较好方法。
Abstract:
      Abstract: Based on the monthly precipitation observations at 93 meteorological stations in the Southwest China from 1996 to 2000, this study investigates the spatial interpolation results from the Weighted Inverse Distance Weighting (IDW)and O-Kriging interpolation methods. Firstly, we analyze the spatial autocorrelation and spatial variability character of monthly average precipitation data; Secondly, IDW and 0-Kriging based on three semi-variograms (exponential, spherical and Gaussian model) are used to spatially interpolate monthly precipitation. Finally, the interpolation results are compared and discussed in terms of the cross-validation method. The conclusions are:(1)the monthly precipitation distribution in the Southwest of China show clearly spatial aggregation with the high spatial autocorrelation and variation,which favors for the spatial interpolation. (2) Compared to the three semi-variograms used in 0-Kriging interpolation method, the best performance is from exponential model, while the worst is from Gaussian model. (3) When O-Kriging and IDW are used in the spatial interpolation of monthly average, maximum and minimum precipitation, the former one perform better than the latter method. The errors between interpolation and observation are overall increase with monthly precipitation magnitude, and the errors from both interpolation methods are obviously reduced after removing the maximum monthly precipitation points. (4) For the study area as a whole, interpolation effect of 0-Kriging is better than that of IDW, but at a single site, this is not the case. Therefore, this study suggests that two interpolation methods have to be tested and compared with observation before we choose a relatively optimal interpolation method at a particular station. There is no absolute optimal method in the spatial interpolation of precipitation for every the study area and all time scales. The optimal interpolation method depends on the actual demands and applications.
主办单位:中国科学院大气物理研究所 单位地址:北京市9804信箱
联系电话: 010-82995048,010-82995049传真:010-82995048 邮编:100029 Email:qhhj@mail.iap.ac.cn
本系统由北京勤云科技发展有限公司设计
京ICP备09060247号