Assessment Of Ecological Recovery Based On BP Neural Network Model

Ecological degradation, as a hot issue, has been a serious threat to human survival, sustainable social and economic development. It is an important action for human to deal with ecological degradation through recovery and reconstruction of ecosystem structure and function that has been damaged. Nowadays, Ecological restoration and reconstruction is main in qualitative evaluation, quantitative research is shortage. The concept of "ecological recovery degree" is proposed, the ecological recovery evaluation system is established, and the ecological recovery degree is quantitative evaluated by using BP neural network in this paper. The research try to provide reference for later ecological restoration and reconstruction work of the small watershed.The small watershed of Zhuxi, which is a typical representative of red eroded soil in Southern China, is the subject of this study. A lot of foundation data were gotten by field sampling and indoor experiment test, ASTER image data, SPOT5image data of the small watershed of Zhuxi. The ecological recovery degree in the small watershed is gotten by using BP neural network model, which set up the ecological restoration with the support of ENVI4.2, ArcGIS9.2, MATLAB7.0, to analyze and evaluate the relationship between the different factors.The results are as follows:1) According to the management concept of ecosystem succession, the concept of ecological restoration degree is proposed, which divide the level of the ecological restoration of between0and1, quantitative evaluating ecological recovery process.2) The precision of simulating and forecasting ecological restoration degree in BP neural network is high. The spatial distribution of ecological restoration degree in the small watershed of Zhuxi is consistent with the field situation.3) By analysis the relationship of factors and ecological restoration degree in the small watershed of Zhuxi, we found the certain relations between factors and the ecological restoration degree. The relationship between ecological recovery degree with species diversity, vegetation coverage, soil erosion intensity respectively are obvious.