| 土壤厚度的空间插值方法比较——以青海三江源地区为例 |
| |
| 引用本文： | 易湘生,李国胜,尹衍雨,彭景涛.土壤厚度的空间插值方法比较——以青海三江源地区为例[J].地理研究,2012,31(10):1793-1805. |
| |
| 作者姓名： | 易湘生 李国胜 尹衍雨 彭景涛 |
| |
| 作者单位： | 1. 中国科学院地理科学与资源研究所, 北京100101;
2. 农业部规划设计研究院农业资源监测站, 北京100125;
3. 西南大学地理科学学院, 重庆400715 |
| |
| 基金项目： | 国家科技支撑计划资助项目(2009BAC61B01) |
| |
| 摘 要： | 利用青海三江源地区533个土壤剖面中的厚度数据, 在GIS技术的支持下, 采用确定性内插(反距离加权、全局多项式、局部多项式和径向基函数)和地统计内插(普通克里格、简单克里格、泛克里格和协同克里格)两类共八种插值方法对研究区土壤厚度的空间分布进行了预测, 并综合比较了各种插值方法的预测误差、统计特征值和插值结果分布图。结果表明:(1)在地统计内插方法中, 普通克里格方法(一阶)插值效果比普通克里格方法(二阶)要好;在普通克里格方法(一阶)的半方差函数模型中, 球状模型的插值效果优于指数模型和高斯模型;普通克里格方法在四种地统计内插方法中预测误差最小、预测结果准确性最好。(2)确定性内插方法中, 反距离加权(指数为1)法的误差较小, 并且对区域与局部趋势的反映效果最好。(3)从预测误差大小和对区域总体及局部趋势的综合反映效果来看, 有异向性的球状模型普通克里格(一阶)插值方法预测结果最能准确反映青海三江源地区土壤厚度的空间分布。
|
| 关 键 词： | 土层厚度 空间分布 空间插值 三江源地区 |
| 收稿时间： | 2011-12-10 |
Comparison on soil depth prediction among different spatial interpolation methods: A case study in the Three-River Headwaters Region of Qinghai Province |
| |
| Yi Xiang-sheng,Li Guo-sheng,Yin Yan-yu,Peng Jing-tao.Comparison on soil depth prediction among different spatial interpolation methods: A case study in the Three-River Headwaters Region of Qinghai Province[J].Geographical Research,2012,31(10):1793-1805. |
| |
| Authors: | Yi Xiang-sheng Li Guo-sheng Yin Yan-yu Peng Jing-tao |
| |
| Institution: | 1. Institute of Geographic Sciences and Natural Resources Research, CAS, Beijing 100101, China;
2. Agriculture Resource Monitoring Station, Chinese Academy of Agriculture Engineering, Beijing 100125, China;
3. School of Geographical sciences, Southwest University, Chongqing 400715, China |
| |
| Abstract: | Based on the soil depth data from 533 soil profiles in the Three-River Headwaters Region of Qinghai Province,with the help of GIS technology,this paper attempted to predict the spatial distribution of soil depth by using deterministic interpolation methods(Inverse Distance Weighted,Global Polynomial Interpolation,Local Polynomial Interpolation and Radial Basis Function) and geostatistics interpolation methods(Oridnary Kriging,Simple Kriging,Universal Kriging and Co-Kriging).Then it compared the prediction errors,statistical characteristics and interpolation results of different interpolation methods.Some important conclusions were obtained from this research,which mainly contained three aspects as follows.(1) The spatial distribution of soil depth using the first order surface trend of Ordinary Kriging was better than that using the second order surface trend.The spherical model in the Ordinary Kriging(first order) was better than the exponential model and Gaussian model.The Ordinary Kriging was the best of the four geostatistics interpolation methods because of its minimum error and accurately predicted result.(2) The Inverse Distance Weighted(exponent 1) method was the best in deterministic interpolation methods from the prediction errors and comprehensive reflection in general and local trends.(3) In comparison of the prediction errors and spatial distribution in general and local trends,the Ordinary Kriging(first order) of spherical method,which had anisotropy,could be best to reflect the spatial distribution of soil depth in the Three-River Headwaters Region of Qinghai Province. |
| |
| Keywords: | soil depth spatial distribution spatial interpolation Three-River Headwaters Region |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《地理研究》浏览原始摘要信息 |
| 点击此处可从《地理研究》下载免费的PDF全文 |
|
