[Coupled SD-PLUS Model for Land Use Change and Multi-scenario Dynamic Simulation in Shandong Province]

Multi-scenario simulation of land use can provide theoretical support for regional land space optimization and resource allocation. Based on the land use data of Shandong Province in 2000， 2010， and 2020， combined with 13 driving factors， four constraint scenarios of inertial development， cultivated land protection， ecological protection， and economic development were established. The land use change trend and driving mechanism of Shandong Province in 2030 were predicted and analyzed by the coupling system dynamics （SD） model and patch-level land use change simulation （PLUS） model. The results showed that： ① Under the inertial development scenario， the construction land increased significantly by 5 232.36 km²， the cultivated land area decreased by 5 363.11 km²， and the conversion of agricultural land to construction land showed an accelerating trend. Under the scenario of cultivated land protection， the proportion of cultivated land increased significantly， reaching 69.19%， which was the highest in the four scenarios. ② The ecological protection scenario emphasized the protection and restoration of natural ecosystems such as forest land， grassland， and waters and promoted the increase of forest land， grassland， and waters by 938.02， 1 027.43， and 983.81 km²， respectively， but the area of cultivated land was reduced by 4 226.47 km². The economic development scenario led to the expansion of construction land to 26.33% （a net increase of 6 835.61 km²） and the reduction of cultivated land by 6 488.91 km². ③ The driving mechanism analysis revealed that the combined contribution rate of GDP and population density to the expansion of construction land was 42.3%， and the leading role of DEM and slope on the distribution of ecological land was more than 35%. The verification of the model showed that the historical simulation error of the SD module was less than 8% （the cultivated land error was only 0.27%）， the Kappa coefficient of the PLUS module was 0.8， and the simulation accuracy was significantly better than that of the single model. By setting and comparing the four scenarios， the research results can provide scientific basis and decision support for future land use planning and management.