The Beijing-Tianjin-Hebei major mineral belt represents a significant economic development area in China. Effective monitoring and assessment of the regional landscape ecological risk can provide a scientific basis for an ecological protection strategy for the environmental protection of the Beijing-Tianjin-Hebei major mineral belt. In this study, a landscape ecological risk index was constructed based on land use/land cover, and the spatial and temporal variations of landscape ecological risk were subsequently analyzed. Furthermore, the contribution of the main driving factors of landscape ecological risk was quantified in the Beijing-Tianjin-Hebei major mineral belt. The results demonstrate that: (1) The land use types within the study area underwent significant changes from 1985 to 2022. The predominant type of transfer was cropland, which was primarily converted to construction land, grassland, and woodland. (2) The landscape ecological risk in central-northern and western parts of the Beijing-Tianjin-Hebei major mineral belt is higher, while the landscape ecological risk in the southwest parts is lower. Using 2015 as the time point, the landscape ecological risk in the study area was found to change, with the average value of landscape ecological risk for all classes of landscape ecological risk within 2015–2022 being lower than that of 1985–2015, with the exception of the high-risk area. The mean annual landscape ecological risk is obviously higher during the 1985–2015 period in comparison to the 2015–2022 period, with the exception of regions exhibiting high risk. (3) There is a significant positive spatial correlation between landscape ecological risks in different periods. The pattern of landscape ecological risk exhibits both ‘high-high’ aggregation and ‘low-low’ aggregation. The ‘high-high’ aggregations are primarily located in the northern, central and western parts of the study area, while the ‘low-low’ aggregation zones are mainly located in the southeastern study region. (4) The spatial distribution of landscape ecological risk is predominantly shaped by population density and slope. In the context of interactive factor detection, the positive interaction between slope and average annual temperature, night-time illumination and slope, population density and annual precipitation were identified as exerting a more significant influence on the observed spatial differentiation of landscape ecological risk. It was found that the interaction of multiple drivers had a more pronounced impact on landscape ecological risk than any single factor. The findings of the research project provide a scientific rationale and reference for future land use, resource optimization, landscape ecological risk differential management and ecological restoration. Furthermore, the findings are of considerable importance in terms of maintaining ecological security patterns.

Atmospheric aerosol pollution has seriously affected the ecological environment and human health in recent years. There are great differences in aerosol optical properties and types due to the influence of environmental conditions, meteorology, industrial and agricultural activities, and other factors of each ecological functional zones. Using MODIS aerosol products (including MCD19A2 and MOD04_3K), this study discussed the temporal and spatial distribution of aerosol optical depth (AOD), Ångström wavelength index (AE) and aerosol types in the Beijing–Tianjin–Hebei region (BTH region) based on the ecological functional zones from 2015 to 2020. The results showed as follows: (1) The AOD in BTH region showed an obviously spatial pattern of low in the north and high in the south, while the spatial pattern of AE was opposite to that of AOD. In addition, the dominant aerosol type of the north part was clean aerosol, the dominant aerosol type of the middle part was biomass burning or urban-industrial aerosol, while that of the other part was mixed aerosols. (2) The seasonal changes of AOD and AE indexes in each ecological functional area had obvious seasonal changes, and the AOD and AE values were highest in summer. At the same time, the proportion of biomass combustion or urban industrial aerosol was the highest in summer. (3) The ecological functional areas with fewer human activities were dominated by clean aerosols, with lower AOD and higher AE value. The ecological functional areas dominated by cities were dominated by mixed aerosols, with higher AOD. The ecological functional areas dominated by agriculture and heavy industry were dominated by biomass combustion or urban industrial aerosols, with the largest AOD. (4) Compared with 2015, the average AOD of each ecological functional area decreased significantly to 2020, and biomass combustion or urban industrial aerosols changed to mixed aerosols.

High-precision prediction of near-surface PM_2.5 concentration is a significant theoretical prerequisite for effective monitoring and prevention of air pollution, and also provides guiding suggestions for the prevention and control of PM_2.5-related health risks. It has been acknowledged that existing PM_2.5 prediction models predominantly rely on variables influenced by near-surface factors. This inherent limitation could hinder the comprehensive exploration of the continuous spatio-temporal characteristics associated with PM_2.5. In this study, an optimal 7-day prediction model for PM_2.5 concentration based on the Stacking algorithm was constructed based on multi-source data mainly including atmospheric environment ground monitoring station data, MODIS remote sensing-derived aerosol optical depth (AOD) daily data and meteorological factors. The findings indicated that the PM_2.5 forecasting outcomes derived from this integrated RF-LSTM-Stacking model exhibited a superior fit, with R², RMSE, and MAE values of 0.95, 7.74 µg/m³, and 6.08 µg/m³, correspondingly. This approach enhanced the accuracy of prediction to a degree of approximately 17% in comparison with a solitary machine learning model. The findings of this study demonstrated that the integration of the LSTM-RF model with the fusion-based Stacking algorithm led to a substantial enhancement in the accuracy of PM_2.5 predictions. This model was found to serve as an effective reference for the monitoring of PM_2.5 prediction and early warning systems.