Time Series Remote Sensing of Land Use Changes and Influences on Runoff and Sediment Yield in Dongjiang River Basin, China

Land use and land cover (LULC) is an important factor affecting runoff and sediment yield of watershed. Dongjiang River is the main source of drink water for Hong Kong and cities in the Pearl River Delta Area, which experienced a rapid urbanization and blooming economy growth. Studying the land use changes using time series remote sensing and their influences on runoff and sediment yield in watershed-scale can provide valuable information for the formulation of land use plans in Dongjiang Watershed. This study constructed a SWAT (Soil and Water Assessment Tool) model of the Dongjiang River basin and analyzed the LULC changes from 2001 to 2012 using MODIS land use data products (MCD12Q1). The influence of LULC changes on the runoff and sediment yield in the Dongjiang basin was analyzed by simulating the runoff and sediment yield under different LULC scenarios. Our results show that the main types of land use in the Dongjiang River basin were evergreen woodland, grassland and agricultural land from 2001 to 2012, accounting for more than 80% of the total area of the basin. During the period of study, the evergreen forest land increased by 14.46%, while grassland, agricultural land and mixed forest decreased by 4.78%, 4.54% and 3.97%, respectively. These changes led to an increase of 0.22 mm in the average monthly runoff in the basin, which reached by 2.39 mm in flood season (April-September). At the same time range, the average monthly sediment yield decreased by 0.79 t/ha with a more significant decreasing by 6.29 t/ha in flood season (From April to September). Our results indicate that long term conversions of grassland, agricultural land and mixed forest to evergreen woodland in watershed will lead to the increasing runoff and decrease of sediments of river, which is important for catchment-scale resource planning and environmental management.