Abstract: Based on data derived from Landsat images in 2001, 2006, 2010, 2014, and 2018, we calculated the weight of selected indices using the entropy weight method and  the remote sensing ecological index (RSEI) using the index weighted sum method. Combined with GM (1,1) model, we analyzed the dynamics of eco-environmental quality and its causes in Mentougou District of Beijing and predicted the future trend. The results showed that the RSEI of Mentougou in 2001, 2006, 2010, 2014 and 2018 were 0.5993, 0.5451, 0.5592, 0.6113 and 0.716, respectively, displaying a trend of “decline first and then rise”. The areas with “excellent” and “good” eco-environmental quality were mainly forestland and shrub land. “Inferior” and “poor” areas were dominated by bare land, urban land use and roads. Mining activity was the main cause for the decline of RSEI from 2001 to 2006. Since 2006, urbanization had negative impacts on the eco-environmental quality of Mentougou District. Some key projects had improved the ecoenvironmental quality. The Beijing-Tianjin sandstorm source control project had a remarkable positive effect on the eco-environmental quality of Mentougou. The visual expression of eco-environmental quality can better reflect the changes of eco-environmental quality and provide better support for the maintenance of ecological balance.

Key words: entropy weight method, remote sensing ecological index, GM (1,1) model, Mentougou district.