收藏设为首页 广告服务联系我们在线留言

干旱区地理 ›› 2017, Vol. 40 ›› Issue (3): 555-563.

• 气候与水文 • 上一篇    下一篇

秦岭主脊区年降水量空间插值最优方法研究

卓静1, 朱延年2   

  1. 1 陕西省农业遥感信息中心, 陕西 西安 710014;
    2 陕西省气象科学研究所, 陕西 西安 710014
  • 收稿日期:2016-12-30 修回日期:2017-03-13 出版日期:2017-05-25
  • 通讯作者: 朱延年.Email:yannianzhu@gmail.com
  • 作者简介:卓静(1978-),女,硕士,高级工程师,从事生态环境遥感监测应用研究.Email:79506610@qq.com
  • 基金资助:

    国家自然科学基金面上项目(41575136)

Spatial interpolation methods of annual average precipitation on Qinling Mountains

ZHUO Jing1, ZHU Yan-nian2   

  1. 1 Shaanxi Remote Sensing Information Center for Agriculture, Xi'an 710014, Shaanxi, China;
    2 Metrological Institute of Shaanxi province, Xi'an 710016, Shaanxi, China
  • Received:2016-12-30 Revised:2017-03-13 Online:2017-05-25

摘要: 利用ARCGIS软件,以研究区多年平均年降水量为数据源,开展以下研究:(1)采用反距离权重法、普通(泛)克里金法、样条函数法和趋势面法进行直接内插。(2)利用降水量随海拔递增率,再次利用以上方法进行插值。(3)分析站点数量的变化对插值精度的影响。(4)分析插值空间尺度变化对插值精度的影响。针对插值结果,使用交叉验证法来评估插值方法的优劣,旨在找出研究区年降水量空间插值的最优方法。结果表明:(1)研究区多年降水量与纬度呈现出很好的负相关,与其他各地理环境因子相关性很差或无相关性。(2)考虑降水随海拔的递增率后,研究区南坡西段和研究区海拔<600 m和>800 m的大部分观测站的空间内插精度得到有效提高,其它区域以及其它海拔区间的插值精度反而降低,因此使用单一线性数据作为整个研究区降水随海拔的变化率不够科学和严谨。(3)在缺少研究区不同剖面降水观测资料的情况下,权重为0.001 的正规则样条函数法是最优插值方法。(4)像元尺度在50~1 000 m间的变化对降水插值的不确定影响甚微;插值精度与选取的插值点个数无明显相关性,当站点个数为20个时,插值精度最高。

关键词: 秦岭, 年降水量, 空间插值, 最优方法

Abstract: Qinling Mountains are a major east-west mountain range in southern Shaanxi Province,China. It extending about 1 500 kilometers across central China from Gansu-Qinghai border in the west to central Henan Province in the east,form a natural dividing line between China's sub-tropical and warm-temperate zones. It also provide a natural boundary between North and South China,and support a huge variety of plant and wildlife, some of which is found nowhere else on the Earth. It plays one of the most important climate boundaries between North and South China. The orographic effects on climate and climatic impacts on hydrological regimes are very important to surrounding areas. Therefore,its precipitation pattern plays an important role on regional and even national vegetation productivity,ecosystem carbon cycle and the ability of water conservation. So far,the studies of optimal interpolation method evaluation are mostly focus on other area,lack of research on complex terrain of Qinling Mountains. In this paper,22 meteorological stations within Qinling Mountains and 19 nearby stations, the annual precipitation of from 2000 to 2015 combined their geographical information and digital terrain model were analyzed with Arcgis spatial analysis package as follows:(1)Direct spatial interpolations of precipitation using five different methods which including Inverse Distance Weighting(IDW),Ordinary Kriging,Universal Kriging,Spline,and Trend Surface Method;(2)Applying the listed interpolation methods again based on the fact that the precipitation increases with increasing elevation;(3)Analyzing the influence of the change of the number of stations on the accuracy of interpolation of precipitation data;(4)Analyzing the influence of interpolation spatial scale on interpolation precision and discussed the optimal interpolation method. In order to find out the optimal method of spatial interpolation of annual precipitation of Qinling Mountains,the mean absolute error (MAE),mean relative error(MRE)and root mean square error(RMSE)of above listed interpolation methods were compared by cross validation based on the interpolation results. The results show as follows:(1)Annual precipitation pattern is irrelevant to topographic but shows a negative correlation with the latitude. Therefore,this paper does not consider multiple regression method;(2)Owing to the terrain complication,the southern slope and the northern slope have their own significant climatic characteristics,currently only considering the precipitation increases with increasing elevation,the interpolation results improved in the west part of south slope of main ridge of Qinling Mountains but not in other areas. The precipitation is not linearly increases with increasing elevation; we need to define the increasing rates on different altitude. The results show that after considering the terrain influence,areas with elevation less than 600 meters or greater than 800 meters have significant improvement. It indicates that only considering the terrain is not sufficient to analyze the precipitation distribution. Due to the complex topography,the difference between north and south slope or the illuminated and shaded slope,further more different increasing rate with increasing elevation need to be taken under consideration for future research. (3)Currently,Qinling Mountains area does not have sufficient precipitation observations,under such circumstances the Spline interpolation with weight of 0.001 provided the ideal results which tested by the meteorological observations;(4)The spatial resolution has minor effect on precipitation interpolation when it between 50 m and 1 000 m. Increasing the number of interpolation stations would not improve the interpolation results. The best result was found when interpolate with 20 stations. The results of the paper would satisfy the researchers who need the spatial information of precipitation in the region,and also provide basic support for ecological model, hydrological model and land surface process researches.

Key words: Qin Mountains, Annual Precipitation, Spatial Interpolation

中图分类号: 

  • P426.61