A Study On Dynamic Changes Of Forest Resources And Nature Closeness Evaluation Based On Remote Sensing

This thesis studied the dynamic changes of forest resources and appraised natural closeness based on RS in Huoditang forest farm. In the paper, dynamic changes of resources and the natural closeness on stand level were investigated by taking remote sensing images in 1956 and the SPOT5 image in 2004, the relief map and the results of investigating for woodland types. With the remote sensing (RS) technology as the primary methods,supplemented by geographic information system (GIS) technology,the changes of forest resources and the construction of stands between wildwoods and today's forest have been researched. The driving mechanisms that lead to changing of resources and stands had also been studied, and an indexes system was built which shows the driving factors affecting natural closeness of stand. Through the analysis above, the forest stand natural closeness was integrative appraised automatically using Analytic Hierarchy Process (AHP) method and Con function of ArcGis software. The main results as follows:(1) Based on the software of Erdas, the remote sensing images were retricted by the relief map strictly, and a standard of image in 1956 was also worked out. Select the result image of K-L amalgamation by contrasting the four methods of image amalgamation, namely, HIS, K-L, ratio of Brovey and small wave band on SPOT5 image.(2) The images were classificated and the areas of each forest resource on forestry farm were computed in 1956 and 2004 by remote sensing images in two periods. Through evaluating the classification accuracy,the results showed that the accuracy were 83.93% and 92.2%, respectively, which show that the method above is proper and the precise achievement can be obtained.(3) Based on GIS, the dynamic changes and bi-directional index of each forest types have been calculated, and the effect of terrain on forest types'change have been incarnated. The result shows that elevation has a big influence on the change of forest types. Further more, the biggest changes taken place in the low elevation region.(4) Using AHP method, the effect of different landform coefficients on the stand of natural closeness was calculated and the forest stand natural closeness was integrative valued by the changing of proportion in different types, which can be a base for the forest natural closeness evaluation.(5) Based on ARCGIS software, the forest stand natural closeness has integrative calculated and evaluated automatically. The result showed that the stand of natural forest natural closeness of forestry farm is distributed from 0.68 to 1, and the forest stand structure has a favorable states.