Dynamic Changes and Driving Factors of Ecosystem Carbon Storage in Inland River Basins Under Land Use Change

ABSTRACT Under the “carbon peaking and carbon neutrality” goals, the extent to which land use restructuring shapes regional carbon storage (CS) in inland river basins remains insufficiently quantified. Using the Tarim River Basin as a case study, we integrated multiple datasets at 1 km resolution for five time points from 2000 to 2023, and coupled the InVEST carbon model, PLUS scenario simulations, and GeoDetector analysis within a “pattern‐process‐projection” framework (PLUS accuracy: 94.31%). From 2000 to 2023, cropland expanded by 1.69 × 10 4 km 2 and construction land nearly doubled, whereas grassland and forestland decreased by 0.89 × 10 4 km 2 and 0.15 × 10 4 km 2 , respectively, accompanied by intermittent shrinkage of water bodies. CS across the basin increased slightly from 56.53 × 10 8 t to 56.83 × 10 8 t, indicating overall stability, but with clear spatial contrasts: gains occurred along oasis margins and riparian corridors, while losses emerged in transitional zones converted to cropland and construction land. GeoDetector identified fractional vegetation cover, soil erosion, and soil type as dominant drivers ( q > 0.25). Interactions related to fractional vegetation cover (FVC) strengthened after 2010, whereas GDP and population density exerted weaker effects. By 2030, the ecological protection scenario yielded the highest CS (57.10 × 10 8 t), the economic development scenario showed limited gains (+0.08 × 10 8 t), and the natural development scenario approached net neutrality. By integrating multi source data, scenario constraints, and interaction informed driver diagnostics, this study delineates carbon sensitive corridors that are highly accessible and quantifies the carbon benefits of controlling fragmentation, stabilizing cropland density, and optimizing water allocation. The findings provide scientific guidance for land use planning and coordinated water and carbon governance in arid regions.