Understanding the Spatiotemporal Dynamics and Driving Factors of the Carbon‐Water Relationship in the Qinghai Lake Basin: Past Trends, Mechanisms, and Future Projections

ABSTRACT Global climate change and intensified human activities have introduced significant uncertainties to carbon‐water systems. Understanding the spatiotemporal dynamics and nonlinear driving mechanisms of these systems is crucial for ecological restoration and sustainable management. This study quantified the Carbon‐Water Coupling Coordination Degree (CWCCD) in the Qinghai Lake Basin using an improved coupling model; it revealed spatiotemporal patterns via spatial mapping and time‐series clustering, employed the LightGBM‐SHAP model to identify nonlinear driving effects, and conducted multi‐scenario simulations by integrating CMIP6 data with the PLUS model. The results are as follows: (1) CWCCD shows an overall trend of “first rising and then declining”; spatially, coordination is stronger in the northern and eastern regions. (2) The evolutionary patterns of CWCCD are categorized into four types: Continuous Decline, Decline after Rapid Improvement, Decline after Stabilization, and Rapid Decline after Improvement. (3) Climatic and biophysical conditions are the core dimensions governing CWCCD, and each influencing factor exhibits a complex nonlinear relationship with CWCCD. (4) Scenario projections indicate that CWCCD is the highest under the SSP126 scenario, highlighting the importance of ecological protection and emission control. Based on the identified temporal evolution patterns, this study proposes zoned ecological management and planning strategies, which provide support for the monitoring and sustainable management of carbon‐water systems.