不同空间插值方法的比较研究--以气候舒适期为例
首发时间:2015-05-25
摘要:气象相关要素的空间插值能够大大提高资料获取、数据处理和地图测绘的效率,对生态环境治理、资源管理以及全球变化研究显得尤为重要。本文利用地理信息系统软件ArcMap(10.1版本),分别运用普通克里金插值法和反距离权重插值法,在考虑或者剔除海拔高程对气象要素空间分布的影响的两种情况下,对中国大陆780个气象站点多年(1951-2009)的年平均气候舒适期进行空间插值处理,并运用交叉检验对各插值结果的精度进行比较分析,结果表明:考虑了海拔高度影响的反距离权重插值方法的精度提高得比较明显,平均误差从-0.44天变为-0.21天,平均绝对误差从14.65下降到了12.86天,均方根误差从23.52天下降到了20.08天。而对于普通克里金法,考虑海拔影响后的插值结果误差反而大于不考虑海拔影响时的误差,平均误差从-0.06天变为-0.08天,平均绝对误差从11.49天升高到了12.68天,均方根误差从18.25天升高到了19.75天。从误差检验结果来看,四种方法中不考虑海拔高程的普通克里金法是最优的空间插值方法。
关键词: 空间分析 普通克里金 反距离权重 气候舒适期 海拔高程
For information in English, please click here
Comparative research of different spatial interpolation methods --taking climate comfort days for example
Abstract:Spatial interpolation of meteorological elements can greatly improve the efficiency of data acquisition, data processing and ground mapping, especially for ecological environment management, resource management, and global change research. This paper used geographic information system software ArcMap 10.1, respectively, in both cases of without considering the effect of altitude on the spatial distribution of meteorological elements and considering the impact of the elevation, with the use of Ordinary Kriging interpolation method (OK) and Inverse Distance Weighting interpolation method (IDW), conducted a spatial interpolation to the annual average climate comfort days ( mean climate comfort days of 1951-2009 ) of 780 weather stations in China, and used cross-validation methods to assess the accuracy of various interpolation methods, test results showed that: the accuracy of Inverse Distance Weighting interpolation method considering the impact of altitude is enhanced obviously, the mean error (ME) dropped from -0.44 days to -0.21 days, the mean absolute error (MAE) dropped from 14.65 days to 12.86 days, the root mean square error (RMSE) dropped from 23.52days to 20.08 days. However, for Ordinary Kriging interpolation, the accuracy of the interpolation taking the impact of elevation into account was higher than not considering the impact of elevation, the mean error (ME) rose from -0.06 days to -0.08 days, the mean absolute error (MAE) rose from 11.49 days to 12.68 days, the root mean square error (RMSE) rose from 18.25 days to 19.75 days. In these four kinds of interpolation methods, the Ordinary Kriging interpolation method without considering the impact of altitude is the optimal spatial interpolation method.
Keywords: spatial analysis ordinary kriging inverse distance weighing climate comfort days altitude
基金:
论文图表:
引用
No.4643765106311214****
同行评议
勘误表
不同空间插值方法的比较研究--以气候舒适期为例
评论
全部评论