Abstract: The single linear regression （SLOPE） and geographically weighted regression （GWR） methods were used to study the variation trend and mechanism of vegetation distribution in the water source area of the middle route of the South-to-North Water Diversion Project by vegetation index （NDVI）, so as to provide theoretical support for guaranteeing water quantity and quality. The result showed that the NDVI and the inflow runoff increased synchronously in the water source area from 2000 to 2019. Vegetation gradually became better. The NDVI increase rate was large before 2007, but fluctuated from 2008 to 2011, and decreased after 2012. The NDVI changed obviously in the northeast, west and around the Danjiangkou reservoir area,these areas had low and flat terrain and concentrated population. The NDVI changed little in middle and high mountain areas. The elevation, slope, population and urbanization were the main factors affecting NDVI distribution in the water source region, while precipitation, temperature and industrial proportion were the secondary factors. The intensity and direction of the influencing factors had obvious heterogeneity. The most significant range affected by natural geographical factors on NDVI were the areas with elevation below 1 090 m and average annual precipitation above 980 mm. The significant range affected by human geographical factors were the areas with high population density and high urbanization rate.

Key words: water source region of the middle route of South-to-North Water Diversion Project, NDVI, spatiotemporal variation