Abstract:

Located at the east of Chongming Island,Dongtan wetland is an ecological sensitive region of global importance. Interfered with human activities,its ecosystem health is faced with threat.Thus， it is necessary to make quantitative research on its health assessment. This study took Dongtan wetland of Chongming Island as a research area,selected indicators like topography and geomorphology,environment,biology and human interference.Based on theory of ecosystem health,this study divided the health condition of research area into five levels and built an index system of wetland health assessment in Dongtan,Chongming Island. Combined with GIS and neural network model technologies,the study took grid as evaluation unit,and then established a neural network model of wetland ecosystem health assessment to quantitatively analyze the health conditions of Dongtan wetland.This model could provide decisionmaking support and scientific basis for rational utilization and protection of wetland sources.The result displayed the health conditions of Dongtan wetland in different areas from space.Generally,Dongtan wetland is comparatively healthy,but areas with healthier condition occupied only 30% and others are in a general condition.Furthermore,the wetland ecosystem health has a descending trend.Therefore,it is necessary to strengthen management of wetland and make a sustainable development of wetland ecosystem.