Abstract: The reservoir wetland of Beijing, constitutes one of the important eco-systems in Beijing. The driving factors index system of Beijing reservoir wetland landscape evolution in the study area was built in the two aspects of the natural environment and socio-economy. Natural driving factors include precipitation, temperature, entry water and groundwater depth; social economic driving factors include the resident population, urbanization rate and per capita GDP. Using TM images from 1984 to 2010 to extract reservoir wetland's spatial distribution information of Beijing, we analyzed the area of reservoir wetland change laws in nearly 30 years. The driving mechanism of reservoir wetland evolution in the study area was explored by the Logistic regression model in different periods. The results indicated that in different phases, the driving factors and their influence on reservoir wetland evolution had certain differences. During 1984-1998, the leading driving factors were annual average precipitation and entry water index with the contribution rate of Logistic regression being 5.78 and 3.50, respectively, which was mainly affected by natural environmental factors; from 1998 to 2004, the impact of human activities intensified and man-made reservoir wetland reduced, and the main driving factors were the number of residents, groundwater depth and urbanization rate with the contribution rate of Logistic regression 9.41, 9.18, and 7.77, respectively. During 2004-2010, reservoir wetland evolution was impacted by both natural and socio-economic factors, and the dominant driving factors were urbanization rate and precipitation with the contribution rate of 6.62 and 4.22, respectively.

Key words: reservoir wetland, driving force, Logistic regression model, Beijing