By analyzing the quantitative changes in land use types, identifying driving mechanisms, and examining land cover transition patterns along with spatial proximity relationships in Changsha from 2000 to 2020, this study aims to reveal the underlying trends of land-use transformation in the city.

Using a PLUS model-based mask extraction approach on raster datasets, this study examined the correlations and underlying mechanisms of various driving factors influencing land use change patterns in Changsha from 2000 to 2020. The land-use expansion module employs adaptive relationships to simulate spatial evolution, enabling analysis of dynamic processes within the study area. Based on the land-use spatiotemporal change data of Changsha from 2000 to 2020, a linear regression model was fitted to investigate the temporal and quantitative relationships among different land use types. The optimal fitting line was obtained by minimizing the sum of squared residuals between observed values and model predictions. The 12 selected driving factors were normalized to generate raster spatial distribution maps, contribution rates, and root mean square errors. Using the random forest algorithm, development probabilities for various land types were calculated. Based on three research objectives-natural development, ecological benefit maximization, and urban economic growth-three future scenarios (natural development, ecological conservation, and urban expansion) were established to simulate future land changes.

Changes in land use categories showed a high correlation with time sequences, with linear regression coefficients averaging over 0.85. The main driving factors and their contributions vary among different land categories. The expansion area of urban land was mainly affected by elevation. The contribution rate of elevation reached 0.16 during the period from 2000 to 2010, and the RMSE value was 0.23. During the period from 2010 to 2020, the contribution rate of elevation decreased to 0.15, and the RMSE value was 0.24. Future land use simulations under the ND (Natural development), EP (Ecological protection), and US (Urban spread) scenarios predict urban expansion areas of 145.22 km², 46.93 km², and 317.64 km², respectively, highlighting the significant role of government policies and measures in influencing the spatiotemporal evolution of urban land use.

From 2000 to 2010, urban expansion demonstrated a steady growth trend in both scale and duration, while the decline rates of arable land and forest land accelerated during the same period. From 2010 to 2020, the expansion rate of construction land slowed down and stabilized, effectively mitigating the reduction trends of ecological lands such as arable land and forest land. In the influence mechanism, elevation is the primary factor; during this period, number of soil types exhibited the lowest impact intensity. The temporal variation patterns of these driving factors align closely with Changsha's actual development trajectory: extensive and rapid urban expansion from 2000 to 2010, followed by controlled expansion scale and slower growth rates from 2010 to 2020.