基于GIS的月平均气温空间化方法的比较研究  

Study on the Comparison of SSpatial Methods for Monthly Mean Temperatures Based on GIS

作　　者：陆琦[1,2] 邱新法[3] Lu Qi;Qiu Xinfa(School of Geography and Remote Sensing,NUIST,Nanjing 210044,China;School of Applied Meteorology,NUIST,Nanjing 210044,China;Jurong Meteorological Bureau of Jiangsu Province,Jurong 212400,China)

机构地区：[1]南京信息工程大学地理与遥感学院,南京210044 [2]江苏省句容市气象局,句容212400 [3]南京信息工程大学应用气象学院,南京210044

出　　处：《科技通报》2017年第3期43-48,共6页Bulletin of Science and Technology

基　　金：国家自然科学基金项目(41175077;41330529);南京信息工程大学科研基金资助项目

摘　　要：使用1961-2000年全国743个气象台站常规气象观测资料,利用IDW、Kriging以及Spline这3种常用空间插值方法,以及复杂地形下月平均气温分布式模型,生成全国的月平均气温空间分布图,并同时与中国气象数据网提供的中国地面气温月值0.5°×0.5°格点数据集进行比较,结果表明:3种插值方法(IDW、Kriging、Spline)、格点数据集与气温分布式模型的绝对误差分别为1.59℃、1.54℃、1.99℃、1.40℃、0.56℃,气温分布式模型的精度最高,而且其空间分辨率最高,模拟的稳定程度较好,能够很好地体现气温随地形的变化特征。因此,气温分布式模型对于平均气温的模拟性能最好。Based on the conventionally observational data from743weather stations in China from1961to2000,the distribution maps of monthly mean temperatures in China were generated by3common spatial interpolation methods(IDW,Kriging and Spline)and distributed model of monthly mean temperatures over the rugged terrain,to compare with the0.5°×0.5°grid data set of monthly air temperatures in China provided by China Meteorological Administration.The results indicate that the absolute errors of3spatial interpolation methods(IDW,Kriging and Spline),grid data set and distributed model are1.59°C,1.54°C,1.99°C,1.40°C and0.56°C.The accuracy of the distributed model of mean monthly meantemperature is the highest with the best spatial resolution and good stability of simulation,and the simulation of distributed model is capable of reflecting terrain-influencing characteristics.As a result,the distributed model over the rugged terrain give the best simulation of monthly mean temperatures.

关 键 词：分布式模型 插值方法 格点产品 平均气温 性能评价 

分 类 号：P423.3[天文地球—大气科学及气象学]