当前位置: 首页 > 文章 > 普通克立格法在昆虫生态学中的应用 应用生态学报 2003,14 (1) 90-92
Position: Home > Articles > Application of ordinary Kriging method in entomologic ecology. Chinese Journal of Applied Ecology 2003,14 (1) 90-92

普通克立格法在昆虫生态学中的应用

作  者:
张润杰;周强;陈翠贤;王寿松
单  位:
中山大学生物防治国家重点实验室/昆虫学研究所
关键词:
普通Kriging法;区域化变量;变差函数;地统计学
摘  要:
地统计学是以区域化变量为基础 ,以变差函数为主要工具 ,分析空间相关变量结构的统计方法 .在对波动较大的实验变差函数进行拟合时 ,虽无法获得最优拟合 ,但运用人机对话的拟合方法来灵活选取参数 ,可以得到较理想的变差函数模型的参数 .本文运用加权多项式回归法以及人机对话的方法 ,得到了较理想的 1级与 2级球状模型拟合结果 ,同时利用直线函数对实验变差函数进行了拟合 .最后利用普通Kriging法 ,对待估计点进行各理论模型的最优、线性、无偏内插估计 ,得出克立格内插权重 .将此方法应用于广东省四会市大沙镇富溪乡试验田稻飞虱观测数据 ,由待估点周围若干观测点的数据 ,有效地估计出待估点的昆虫分布密度 ,并讨论比较了不同理论模型的拟合效果以及估计误差 .结果表明 ,2级球状模型的拟合最好 ,一级球状模型次之 ,直线函数的拟合最差 ,但直线函数计算最为简便 .
译  名:
Application of ordinary Kriging method in entomologic ecology.
作  者:
ZHANG Runjie, ZHOU Qiang, CHEN Cuixian, WANG Shousong (Institute of Entomology and State Key Laboratory for Biocontrol, Zhongshan University, Guangzhou 510275,China) .
关键词:
Ordinary Kriging method, Local variable, Variograms, Geostatistics.
摘  要:
Geostatistics is a statistic method based on regional variables and using the tool of variogram to analyze the spatial structure and the patterns of organism. In simulating the variogram within a great range, though optimal simulation cannot be obtained, the simulation method of a dialogue between human and computer can be used to optimize the parameters of the spherical models. In this paper, the method mentioned above and the weighted polynomial regression were utilized to simulate the one-step spherical model, the two-step spherical model and linear function model, and the available nearby samples were used to draw on the ordinary Kriging procedure, which provided a best linear unbiased estimate of the constraint of the unbiased estimation. The sum of square deviation between the estimating and measuring values of varying theory models were figured out,and the relative graphs were shown. It was showed that the simulation based on the two-step spherical model was the best simulation, and the one-step spherical model was better than the linear function model.
计量
文章访问数: 10
HTML全文浏览量: 0
PDF下载量: 0

所属期刊