Due to a series of human activities like deforestation and land degradation, the concentration of greenhouse gases has risen significantly. Forest vegetation is an important part of forest ecosystems with high carbon sequestration potential. Estimates of the carbon sequestration rate of forest vegetation in various provinces and districts are helpful to the regional and global Carbon cycle. How to build an effective carbon sequestration potential model and reveal the spatiotemporal evolution trend and driving factors of carbon sequestration potential is an urgent challenge to be solved in carbon cycle simulation and prediction research. This study characterized the carbon sequestration status of forest vegetation using the modified CASA (Carnegie-Ames Stanford Approach) model and estimated the carbon sequestration potential from 2010 to 2060 using the FCS (Forest Carbon Sequestration) model combined with forest age and biomass under the four future Shared Socioeconomic Pathways (SSP) scenarios: SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5, then proposes natural, social, and economic perspectives. This study found that the average NPP of the forest vegetation in Chongqing from 2000 to 2020 was 797.95 g C/m², and the carbon storage by 2060 was 269.94 Tg C. The carbon sequestration rate varied between <0.01 Tg C/a and 0.20 Tg C/a in various districts and counties. Over time, forest growth gradually slowed, and carbon sequestration rates also decreased. Under the four future climate scenarios, the SSP5-8.5 pathway had the highest carbon sequestration rate. Natural factors had the greatest influence on changes in carbon sequestration rate. This result provides data support and scientific reference for the planning and control of forests and the enhancement of carbon sequestration capacity in Chongqing.

Analysis of the relationship between future land use change and agricultural non-point source pollution (ANPSP) evolution is vital to promoting sustainable regional development. By simulating future land use types, we can identify and analyze the evolution trend of ANPSP. This study takes Chongqing as a case study to establish an integrated solution based on the PLUS model, output coefficient model, and GIS technology. The solution can simulate data, identify trends, and identify key control areas under future development scenarios. The results show that the PLUS model can simulate land use types at the provincial scale with high accuracy, with a Kappa coefficient of around 0.9. The land use type changes show that urban expansion has occupied a large amount of cultivated land. From 2000 to 2020, the proportion of high-load areas with TN pollution load levels was 4.93%, 5.02%, and 4.73%, respectively. Under the two scenarios in 2030–2050, the number of high-load areas decreased, and the average load level decreased from west to east. Sensitivity analysis found that risk changes are more sensitive to the increase in fertilizer application. When the TN and TP output coefficients are increased, the number of towns with increased levels is greater than those with decreased levels when the output coefficients are decreased. Sensitivity analysis can better identify key pollution control areas. The areas sensitive to changes in farmers’ behavior are mainly the Hechuan District, Nanchuan District, Qijiang District, Jiangjin District, and Bishan District. This study provides data and decision-making support for rural green development and water environment improvement.

Accurate forecasting of carbon emissions at the county level is critical to support China’s dual-carbon goals. However, most current studies are limited to national or provincial scales, employing traditional statistical methods inadequate for capturing complex nonlinear interactions and spatiotemporal dynamics at finer resolutions. To overcome these limitations, this study develops and validates a high-resolution predictive model using advanced gradient boosting algorithms—Gradient Boosting Decision Tree (GBDT), Extreme Gradient Boosting (XGBoost), and Light Gradient Boosting Machine (LightGBM)—based on socioeconomic, industrial, and environmental data from 2732 Chinese counties during 2008–2017. Key variables were selected through correlation analysis, missing values were interpolated using K-means clustering, and model parameters were systematically optimized via grid search and cross-validation. Among the algorithms tested, LightGBM achieved the best performance (R² = 0.992, RMSE = 0.297), demonstrating both robustness and efficiency. Spatial–temporal analyses revealed that while national emissions are slowing, the eastern region is approaching stabilization, whereas emissions in central and western regions are projected to continue rising through 2027. Furthermore, SHapley Additive exPlanations (SHAP) were applied to interpret the marginal and interaction effects of key variables. The results indicate that GDP, energy intensity, and nighttime lights exert the greatest influence on model predictions, while ecological indicators such as NDVI exhibit negative associations. SHAP dependence plots further reveal nonlinear relationships and regional heterogeneity among factors. The key innovation of this study lies in constructing a scalable and interpretable county-level carbon emissions model that integrates gradient boosting with SHAP-based variable attribution, overcoming limitations in spatial resolution and model transparency.