应用“3S”技术进行北京市森林立地分类和立地质量评价的研究

Application of 3S Technology to Forest Site Type Classificationand Site Quality Evaluation in Beijing

本文针对目的森林立地分类和立地质量评价研究中存在的问题。试图从立地分类和评价的方法进行技术上新的探讨。研究中以生态学和景观生态学理论为基础，以遥感（ＲＳ）、地理信息系统（ＧＩＳ）和专家系统（ＥＳ）作为主要的技术手段，借助相关数学分析，对北京市森林进行立地类型划分和在此基础上的立地质量评价及多目标动态决策。通过对北京地区景观特征及其气候、地貌、地形、土壤、植被等各种因子的定性分析，以及一、二类调查数据的定性定量分析，结合实际应用，确定分类原则和分类系统。应用ＧＩＳ建立各级分类的各因子专题数据库和图形库，通过各分类层次的主导因子及辅助因子的专题图叠加完成初分类。在此基础上，应用知识库中的专家知识及ＥＳ的推理机制对初分类结果进行综合，根据可信度理论，得出各级符合生产、造林规划和经营实际的立地类型图。对于最低一级的分类结果，采用专家打分的方法进行各类型立地质量的多因子综合评价，并编制立地质量评价图。在分类和评价的基础上，可以根据森林的生态效益、经济效益、环境效益和社会效益的综合考虑选择目标集，进行面向目标的动态决策。

In view of the existing problems in the study of forest site type classification and site quality evaluation ,based on the ecology and landscape ecology theory, this paper deals with a new method to accomplish the forest site type classification,site quality evaluation, and automatic mapping in Beijing region using Remote Sensing(RS), GIS, Expert System(ES) and other relevant mathematical statistics analysis. First, the principles and system of classification are determined, according to the requests of application, the results of qualitative analysis to the features of landscape, climate, landforms, terrain, soil and vegetation, qualitative and quantitative analysis to the first type and the second type survey data. Secondly, setting up the thematic maps and attributes data bases of all factors of different classification levels using GIS as platform, and accomplishing the pre-classification of every level by overlaying the thematic maps. Thirdly, with the aid of expert knowledge stored in the knowledge base and the inference mechanism of ES, compiling the pre-classification maps in the light of the conference theory; Finally, assessing the site quality of every site type using scores of the factor indices given by the experts, and creating the evaluation maps. Based on the classification and evaluation results, the users can select different sets of objects which can integrate the ecological, economic, environmental and social benefit of forest, and implement the object-oriented dynamic decision-making.