A study on forest site type classification and site quality evaluation using remote sensing techniques

This paper deals with a new method to accomplish the forest site type classification, site quality evaluation, and automatic mapping in Beijing region using the remote sensing, GIS, expert system and other relevant mathematical statistics analysis. First, we determine the principles and system of classification according to the requests of application, the results of qualitative analysis to the features of landscape, climate, landforms, terrain, soil and vegetation, etc., qualitative and quantitative analysis to the first type and the second type survey data. Next, we set up the thematic maps and attributes of data bases of all factors of different classification level using GIS as platform, and accomplish the pre-classification of every level by overlaying the thematic maps. Then, with the aid of experts knowledge stored in the knowledge base and the inference mechanism of ES, we combine the pre-classification maps in the light of Dempster Shafer´s information synthetical theory. Finally, we assess the site quality of every site type using scores of the factor indices given by the experts, and create the evaluation maps.