Simulating wetland changes under different scenarios based on integrating the random forest and CLUE-S models: A case study of Wuhan Urban Agglomeration

Abstract

Wetlands are one of the most productive ecosystems and play an important role in supporting a wide range of biodiversity and providing various kinds of ecosystem services. Rapid urbanization and climate change, however, have resulted in the disappearance of large amounts of wetlands. In this research, we developed a spatial allocation model by coupling random forest regression and the CLUE-S algorithm to simulate the spatial dynamics of wetlands in the Wuhan Urban Agglomeration. Then, the calibrated model was used to predict the wetland distributions from 2015 to 2040 under three scenarios, i.e. natural increase scenario (NIS), economic development scenario (EDS), and wetland protection scenario (WPS). The results showed that: (1) the natural wetlands exhibit a slight decreased trend under EDS, and a slight increased trend under WPS. But regardless of scenarios, the natural wetlands will suffer from degradation in densely built-up areas during 2015–2040; (2) The ponds will clearly expand under all scenarios, and most of their expansion distribute in Xiantao city; (3) The paddy will continue to degradation under three scenarios, and the area decreases most under EDS with value of 2866.94 km2. Most of the paddy degradation are located in Wuhan, Xiantao and Ezhou city. The proposed framework offered an effective tool to explore the urban wetland dynamics in the future, and revealed wetland distribution under different scenarios, which could provide support for the protection of urban wetlands and sustainable future develop in the Wuhan Urban Agglomeration.