Spatiotemporal Dynamics and Driving Factors of Carbon Storage in the Weigan River Basin, an Arid Inland River Basin of Northwest China

Clarifying carbon storage (CS) patterns in space and over time, together with the factors influencing these patterns, is important for achieving carbon neutrality in arid regions. Many basin-scale assessments overlook spatially differentiated dynamics within heterogeneous landscapes. To address this issue, the Weigan River Basin (WRB) was categorized into spatial clusters, and the InVEST model was integrated with GeoDetector to evaluate driver effects over 1990–2020. In the same interval, cultivated land expanded by 2,244.85 km 2 and unused land increased by 3,090.07 km 2 , largely through conversion from grassland, which declined by 4,368.42 km 2 . Total CS decreased by 2.34×10 7 t. Higher values occurred in northern and southern areas, whereas the central area showed lower values, and Global Moran’s I confirmed spatial dependence with statistical significance. In low-CS clusters, nighttime light (NL) and population density (PD) exerted stronger influences, whereas climatic factors played a leading role in high-CS clusters. Interaction analysis further demonstrated that interactions among socioeconomic, climate, and topography provided greater explanatory strength than individual variables, highlighting differentiated controls across heterogeneous zones.