For many years, Beijing has suffered from severe air pollution. At present, fine particulate matter (PM_2.5) pollution in the winter and ozone (O₃) pollution in the summer constitute serious environmental problems. In this study, the combination of a comprehensive air quality model with particulate matter source apportionment technology (CAMx-PAST) and monitoring data was used for the high-spatial resolution source apportionment of secondary inorganic components (SNA: SO₄²⁻, NO₃⁻, and NH₄⁺) in PM_2.5; their corresponding precursor gases (SO₂, NO₂, and NH₃); and O₃ in the winter and summer over Beijing. Emissions from residents, industry, traffic, agriculture, and power accounted for 54%, 25%, 14%, 5%, and 2% of PM_2.5 in the winter, respectively. In the summer, the emissions from industry, traffic, residents, agriculture, and power accounted for 42%, 24%, 20%, 10%, and 4% of PM_2.5, respectively. The monthly transport ratio of PM_2.5 was 27% and 46% in the winter and summer, respectively. The regional transport of residential and industrial emissions accounted for the highest proportion of PM_2.5. The regional transport of emissions had a significant effect on the SO₄²⁻ and NO₃⁻ concentrations, whereas SO₂ and NO₂ pollution were mainly affected by local emissions, and NH₄⁺ and NH₃ were mainly attributed to agricultural emissions. Industrial and traffic sources were two major emission sectors that contributed to O₃ pollution in Beijing. The monthly transport ratios of O₃ were 31% and 65% in the winter and summer, respectively. The high-spatial resolution regional source apportionment results showed that emissions from Langfang, Baoding, and Tangshan had the greatest impact on Beijing’s air pollution. This work’s methods and results will provide scientific guidance to support the government in its decision-making processes to manage the PM_2.5 and O₃ pollution issues.

Northern China experienced a severe sand and dust storm (SDS) on 14/15 March 2021. It was difficult to simulate this severe SDS event accurately. This study compared the performances of three dust-emission schemes on simulating PM₁₀ concentration during this SDS event by implementing three vertical dust flux parameterizations in the Comprehensive Air-Quality Model with Extensions (CAMx) model. Additionally, a statistical gusty-wind model was implemented in the dust-emission scheme, and it was used to quantify the gusty-wind contribution to dust emissions and peak PM₁₀ concentration. As a result, the LS scheme (Lu and Shao 1999) produced the minimum errors for peak PM₁₀ concentrations, the MB scheme (Marticorena and Bergametti 1995) underestimated the PM₁₀ concentrations by 70–90%, and the KOK scheme (Kok et al. 2014) overestimated PM₁₀ concentrations by 10–50% in most areas. The gusty-wind model could reasonably reproduce the probability density function of 2-min wind speeds. There were 5–40% more dust-emission flux and 5–40% more peak PM₁₀ concentrations generated by the gusty wind than the hourly wind in the dust-source regions. The increase of peak PM₁₀ concentration caused by gusty wind in the non-dust-source regions was higher than in the dust-source regions, with 10–50%. Implementing the gusty-wind model could help improve the LS scheme’s performance in simulating PM₁₀ concentrations of this severe SDS event. More work is still needed to investigate the reliability of the gusty-wind model and LS scheme on various SDS events.

Solar radiation is the fundamental energy source of climate change, which has a significant impact on the generations of secondary fine particulate matter (PM_2.5) and ozone (O₃) in the atmosphere. Additionally, surface solar radiation is also affected by the concentration of PM_2.5, which in turn affects the generation of O₃. To clarify the relationships among the O₃, PM_2.5 and the total radiation intensity, this study analyzes their temporal and spatial variation trends from 2017 to 2019. Meanwhile, as a common precursor of O₃ and PM_2.5, concentration variations in nitrogen dioxide (NO₂) are discussed as well in this study. The results showed the following: (1) There are significant positive correlations between the O₃-8 h concentrations and the total radiation intensities in critical regions, especially in the “2 + 26” cities, Fen-Wei Plain and Yangtze River Delta. (2) The decrease in PM_2.5 concentrations is in good agreement with the trend of NO₂ concentrations, while the response of O₃ concentration to the NO₂ concentration variation differs in different regions, except in the Pearl River Delta. (3) In addition to the meteorological factors, changes in the concentrations and ratios of precursors such as NO₂ and volatile organic compounds (VOCs) likely contribute to the observed fluctuations in O₃ concentrations in recent years.

We present AI-DUST, a deep learning model for dust forecasting directly driven by AI-generated weather forecasts. Integrating a Multiple Stacked Graph Attention Network with physical constraints and a physics-based emission scheme, AI-DUST captures key atmospheric physical processes without relying on traditional numerical dust modeling chains. The model demonstrates exceptional accuracy in reproducing a traditional dust model, with correlations >0.99 (one-step) and >0.61 (80-step). In real-time forecasts of 2025 spring sand and dust storms (SDS) over East Asia, AI-DUST outperformed operational models, achieving a 27% higher Threat Score (TS) in 48-hour predictions across 14 strong events. Its 10-day forecast TS exceeds 0.22, demonstrating strong long-term capability. The model generalizes well to unseen regions like the Sahara, enabled by its architecture and standardized preprocessing. This work demonstrates the feasibility of building atmospheric environmental forecasting systems directly driven by AI-generated weather forecasts, paving the way for new, efficient AI-driven chemistry and transport models.

The dynamics of cyanobacteria bloom in Lake Taihu, China, are subject to rapid fluctuations under the influence of various factors, with meteorological conditions being particularly influential. In this study, monitoring data on the surface area of cyanobacteria bloom in Lake Taihu and observational data from automatic meteorological stations around Lake Taihu from 2016 to 2022 were utilized. Meteorological sub-indices were constructed based on the probability density distributions of meteorological factors in different areas of cyanobacterial bloom. A stacked ensemble model utilizing various machine learning algorithms was developed. This model was designed to forecast the cyanobacterial bloom area index in Lake Taihu based on meteorological data. This model has been deployed with real-time gridded forecasts from the China Meteorological Administration (CMA) to predict changes in the cyanobacteria bloom area index in Lake Taihu over the next 7 days. The results demonstrate that utilizing meteorological sub-indices, rather than traditional meteorological elements, provides a more effective reflection of changes in cyanobacteria bloom area. Key meteorological sub-indices were identified through recursive feature elimination, with wind speed variance and wind direction variance highlighted as especially important factors. The real-time forecasting system operated over a 2.5-year period (2023 to July 2025). Results demonstrate that for cyanobacteria bloom areas exceeding 100 km², the 1-day lead-time forecast hit rate exceeded 72%, and the 3-day forecast hit rate remained above 65%. These findings significantly enhance forecasting capability for cyanobacterial blooms in Lake Taihu, offering critical support for sustainable water management practices in one of China’s most important freshwater systems.