Spatiotemporal changes in land use and identification of driving factor contributions in the Chengdu-Chongqing Economic Circle based on the random forest model.

As urbanization accelerates and ecological protection demands increase, the evolution of land use patterns and their driving mechanisms have become critical issues for regional sustainable development. This study focuses on the Chengdu-Chongqing Economic Circle as the research area, utilizing land use data from 2000 to 2022. By employing the land use transfer matrix, dynamic degree and stability index, standard deviational ellipse method, and the PLUS model integrated with the random forest algorithm, we systematically analyze the spatiotemporal evolution characteristics of land use and the contribution of driving factors. The findings are as follows: first, land use patterns have significantly evolved, with a continuous decrease in cultivated land, primarily converted to forestland and construction land; forestland shows scattered expansion, construction land exhibits clustered expansion around urban agglomerations, grassland area has contracted, while water bodies and unused land have slightly increased. Second, the intensity of land change has increased, and stability has decreased, with the comprehensive dynamic degree rising from 0.17 to 0.24%, indicating increasingly active land use activities. Third, driving factors vary by type, with natural factors (e.g., elevation, slope, distance to rivers) mainly influencing the expansion of cultivated land, forestland, and water bodies, while socio-economic factors such as GDP and population density dominate the expansion of construction land and unused land. This study suggests strengthening the protection of cultivated land at urban fringes, optimizing the spatial layout of construction land, and enhancing management and ecological restoration in ecologically fragile areas to achieve coordinated regional ecological protection and economic growth.