A multi‐scale spatial framework for simulating land use carbon risks and ecological compensation under future development scenarios

Land‐use change creates spatial mismatches between carbon emissions, coefficient‐based potential carbon uptake, and ecological compensation responsibilities. Using land‐use, energy‐consumption, and socioeconomic data from 2000 to 2024, this study develops a Prediction–Assessment–Zoning framework for the Yellow River Basin (YRB), China. The framework integrates land‐use carbon accounting, a Carbon Emission and Absorption Coupling Risk (CEACR) index, bivariate LISA, PLUS‐based land‐use simulation, GM‐LSTM‐based energy‐emission prediction, and K‐means clustering for carbon compensation zoning. The estimated uptake in this study refers to coefficient‐based potential carbon uptake rather than observed net ecosystem productivity. The results show that the mean CEACR value increased from 0.0193 in 2000 to 0.0279 in 2024, representing a 44.56% increase, while the proportion of extremely high‐risk areas rose from 1.92% to 2.79%. High‐risk clusters expanded mainly in urban‐industrial areas of the middle and lower reaches, indicating clear cross‐boundary spillover of land‐use carbon risk. Under the 2030 urban development scenario, the mean CEACR value reached 0.0347, 22.6% higher than that under the natural development scenario, whereas low‐carbon development and cropland protection constrained risk growth. The low‐carbon scenario reduced total land‐use carbon emissions by 15.05% compared with the natural development scenario. Based on ECC, ESC, and spatial risk patterns, Shandong, Henan, and Ningxia were identified as payment zones; Qinghai, Sichuan, and Inner Mongolia as compensation zones; and Shanxi, Shaanxi, and Gansu as balanced zones.