基于GIS的贝叶斯统计推理技术在印度野牛生境概率评价中的应用  

Application of Bayesian Statistics Inference Techniques Based on GIS to the Evaluation of Habitat Probabilities of Bos Gaurus Readei

作　　者：张洪亮[1] 李芝喜 王人潮[2] 张军[3] 孟鸣[3] 

机构地区：[1]西南林学院资源学院,云南昆明650224 [2]浙江大学环境与资源学院,浙江杭州310029 [3]云南省地理研究所,云南昆明650223

出　　处：《遥感学报》2000年第1期66-70,T001,共5页NATIONAL REMOTE SENSING BULLETIN

基　　金：云南省科委资助项目!( 项目编号:98C013Q)

摘　　要：目前，ＧＩＳ技术已被广泛应用在野生动物生境研究中。但是，作为空间数据分析和处理工具，ＧＩＳ缺乏进行启发式推理的能力。因此，与擅长于此的贝叶斯统计推理技术相结合则是解决这一问题的重要途径。以西双版纳纳板河流域生物圈保护区为试验区，综合应用ＧＩＳ技术和多元统计技术建立印度野牛生境的两个逻辑斯蒂多元回归模型：趋势表面模型和环境模型，第一个模型的自变量是位置坐标，第二个模型的自变量是一组环境因子，然后应用贝叶斯统计合并这两个模型产生贝叶斯综合模型。结果表明，贝叶斯综合模型优于环境模型，可应用于野生动物生境概率评价。At present, GIS has been widely applied to the study of wildlife habitat. However, GIS, which is a tool of spatial data analysis and processing, lacks of the capacity of heuristic reasoning. Therefore, it is an important way to solve this problem by the integration of Bayesian statistics inference with GIS. in this article, the Naban river nature reserve of Xishuangbanna was taken as an experimental area, GIS and multivariate statistical techniques were applied to the development of two logistic multiple regression models for Bos gaurus readei habitat: trend surface model and environmental model. Independent variables were locational coordinates in the first model, and a set of environmental factors in the second model. Bayesian statistics were then used to integrate the two models into a Bayesian integrated model. The results show that the Bayesian integrated model is superior to the environmental model and can be applied to wildlife habitat evaluation.

关 键 词：贝叶斯统计推理 生境 西双版纳 印度野牛 GIS 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]