Simulation of future land use/cover change (LUCC) in typical watersheds of arid regions under multiple scenarios.

The rapid development of the social economy has promoted a continuous increase in the intensity and scale of land use by humans, which has seriously affected the sustainable development of the region. It is important to understand the land use/cover change (LUCC) in the arid region and its future development trends and to make reasonable planning recommendations for the sustainable development of the ecological environment. This study validates the patch-generating land use simulation (PLUS) model in a typical arid region, the Shiyang River Basin (SRB), and analyzes the applicability of the model in arid regions. On this basis, the PLUS model is combined with the scenario analysis method to design four scenarios including no policy intervention, farmland protection, ecological protection and sustainable development to analyze the dynamic changes in past and future land use in the SRB and to make corresponding planning recommendations for the development of each type of land use in the arid region. The results showed that the PLUS model had a better simulation effect in the SRB (its overall accuracy reached 0.97). Coupled models obtain better simulation results than quantitative and spatial models by comparing the mainstream models, with PLUS model that combines CA model and patch generation strategy showing better simulation results in the same category. From 1987 to 2017, the spatial centroid of each LUCC in the SRB moved to varying degrees due to a continuous increase in human activities. The spatial centroid of water bodies had the most obvious change, with a moving speed of 1.49 km/a, while the moving speed of built-up land increased year by year. The spatial centroid of farmland, built-up land and unused land all shifted toward the middle and lower plains, which is a further indication of increased human activity. Due to different government policies, the development trend of land use was also different under different scenarios. However, the four scenarios all showed that the area of built-up land will be increasing exponentially from 2017 to 2037, which would seriously affect the surrounding ecological land and have a negative impact on the local agro-ecological environment. Therefore, we proposed the following planning recommendations: (1) Land leveling work should be carried out on scattered farmland located at high altitudes and with slopes over 25°. Additionally, the land use of low-altitude areas should strictly adhere to basic farmland, increase the diversification of cropping patterns and improve the efficiency of agricultural water. (2) The relationship between ecology, farmland and cities should be reasonably coordinated and the existing idle built-up land should be efficiently used. (3) Forestland and grassland resources should be strictly protected and the ecological redline should be strictly observed. This study can provide new ideas for LUCC modeling and prediction in other parts of the world and provide a strong basis for ecological management and sustainable development in arid areas.