Characterizing Shifts in Major Land Use Types and the Response of Water Yield in a Catchment with Widespread Peaty Wetlands

Landscape patterns have changed substantially in many countries over the last few decades which have profound impacts on hydrological processes. Wetlands have great potential in increasing watershed water retention capacity, so it is considered an important nature-based solution to improve water management. Yet, the impacts of wetland area alongside climate change on water yield are not adequately reported, especially in headwater catchments with widespread wetland distributions. In this study, we carried out analysis in the Zoige Plateau where owns the largest peaty wetlands in China. After reclassification of the land use land cover (LULC) types by applying a machine learning algorithm on Landsat imageries, we found an overall increase in wetland area during 1991-2022, mostly converted from grassland. Wetland experienced large dynamic changes in the region, distinguishable in two phases, i.e., dramatic degradation before 2009 because of overgrazing, and recovery afterwards after the policy implementation for wetland restoration. Furthermore, the impacts of climate factors and LULC types on water yield were quantified using Structural Equation Model and Multiple Linear Regression methods. Results showed that precipitation was evidently the dominant factor with a contributing factor of over 0.7, followed by wetland area (similar to 0.2). In comparison, forests and grassland change and drought conditions played a much weaker role. The analysis indicates that in the context of wetting and warming tendency in the region, increasing the wetland areas can play a critical role in modulating hydrological processes for sustainable water supply to the downstream areas.