The African continent continuously experiences extreme aerosol load conditions, during which the World Health Organization clean air standard of 10 μg/m³ of PM_2.5 mass is systematically exceeded. Africa holds the world largest source of desert dust emissions, undergoes strong industrial growth, and produces approximately a third of the Earth's biomass burning aerosol particles. Sub‐Saharan biomass burning is driven by agricultural practices, such as burning fields and bushes in the postharvest season for fertilization, land management, and pest control. Thus, these emissions are predominantly anthropogenic. Here we use global atmospheric composition, climate, and health models to simulate the chemical composition of the atmosphere and calculate the mortality rates for Africa by distinguishing between purely natural, industrial/domestic, and biomass burning emissions. Air quality‐related deaths in Africa rank within the top leading causes of death in Africa. Our results of ~780,000 premature deaths annually point to the extensive health impacts of natural emissions, high mortality rate caused by industrialization in Nigeria and South Africa, and a smaller extent by fire emissions in Central and West Africa. In Africa, 43,000 premature deaths are linked to biomass burning mainly driven by agriculture. Our results also show that natural sources, in particular windblown dust emissions, have large impacts on air quality and human health in Africa.

We investigate and assess how well a global chemical transport model (GEOS‐Chem) simulates submicron aerosol mass concentrations in the remote troposphere. The simulated speciated aerosol (organic aerosol (OA), black carbon, sulfate, nitrate, and ammonium) mass concentrations are evaluated against airborne observations made during all four seasons of the NASA Atmospheric Tomography Mission (ATom) deployments over the remote Pacific and Atlantic Oceans. Such measurements over pristine environments offer fresh insights into the spatial (Northern [NH] and Southern Hemispheres [SH], Atlantic, and Pacific Oceans) and temporal (all seasons) variability in aerosol composition and lifetime, away from continental sources. The model captures the dominance of fine OA and sulfate aerosol mass concentrations in all seasons. There is a high bias across all species in the ATom‐2 (NH winter) simulations; implementing recent updates to the wet scavenging parameterization improves our simulations, eliminating the large ATom‐2 (NH winter) bias, improving the ATom‐1 (NH summer) and ATom‐3 (NH fall) simulations, but producing a model underestimate in aerosol mass concentrations for the ATom‐4 (NH spring) simulations. Following the wet scavenging updates, simulated global annual mean aerosol lifetimes vary from 1.9 to 4.0 days, depending on species. Aerosol lifetimes in each hemisphere vary by season, and are longest for carbonaceous aerosol during the southern hemispheric fire season. The updated wet scavenging parameterization brings simulated concentrations closer to observations and reduces global aerosol lifetime for all species, indicating the sensitivity of global aerosol lifetime and burden to wet removal processes.

The evolution of planetary boundary layer (PBL) was investigated using observations from a laser ceilometer, an eddy covariance system, and an automatic meteorological station in the north of Nanjing city during an air pollution episode in 2016–2017 winter. Based on 7‐day observation under clean to polluted day, we recorded the temporal variations of backscatter signals observed by the ceilometer, and then intercompared the planetary boundary layer height (PBLH) retrieved from individual methods. The results show that backscatter signal gradient, standard deviation, and wavelet transform analysis methods generated similar PBLH values and PBL diurnal variation patterns. Moreover, the PBL structure varied diurnally, with distinct patterns corresponding to clean and polluted days. Based on these measurements, the relationships between PBLH, weather conditions, and contaminants were analyzed. Results show that on clean days, strong surface turbulence exchange makes the PBL fully developed and makes the PBLH increase sharply after sunrise, with a maximum of 1,483 m; on polluted days, stable synoptic conditions and the weaker wind speeds facilitated the accumulation of air pollutants, leading to smaller net surface radiation and weaker turbulence. Consequently, these conditions during polluted days led to lower PBLH values, which were typically less than 900 m.

Stratospheric injections of sulfur dioxide from major volcanic eruptions perturb the Earth's global radiative balance and dominate variability in stratospheric sulfur loading. The atmospheric component of the GFDL Earth System Model (ESM4.1) uses a bulk aerosol scheme and previously prescribed the distribution of aerosol optical properties in the stratosphere. To quantify volcanic contributions to the stratospheric sulfur cycle and the resulting climate impact, we modified ESM4.1 to simulate stratospheric sulfate aerosols prognostically. Driven by explicit volcanic emissions of aerosol precursors and non‐volcanic sources, we conduct ESM4.1 simulations from 1989 to 2014, with a focus on the Mt. Pinatubo eruption. We evaluate our interactive representation of the stratospheric sulfur cycle against data from Moderate Resolution Imaging Spectroradiometer, Multi‐angle Imaging SpectroRadiometer, Advanced Very High Resolution Radiometer, High Resolution Infrared Radiation Sounder, and Stratospheric Aerosol and Gas Experiment II. To assess the key processes associated with volcanic aerosols, we performed a sensitivity analysis of sulfate burden from the Mt. Pinatubo eruption by varying injection heights, emission amount, and stratospheric sulfate's dry effective radius. We find that the simulated stratospheric sulfate mass burden and aerosol optical depth in the model are sensitive to these parameters, especially volcanic SO₂ injection height, and the optimal combination of parameters depends on the metric we evaluate.

Accurate wind speed prediction is crucial for the safe and efficient utilization of wind resources. However, current single-value deterministic numerical weather prediction methods employed by wind farms do not adequately meet the actual needs of power grid dispatching. In this study, we propose a new hybrid forecasting method for correcting 10 m wind speed predictions made by the Weather Research and Forecasting (WRF) model. Our approach incorporates variational mode decomposition (VMD), principal component analysis (PCA), and five artificial intelligence algorithms: deep belief network (DBN), multilayer perceptron (MLP), random forest (RF), eXtreme gradient boosting (XGBoost), light gradient boosting machine (lightGBM), and the Bayesian optimization algorithm (BOA). We first predict wind speeds using the WRF model, with initial and lateral boundary conditions from the Global Forecast System (GFS). We then perform two sets of experiments with different input factors and apply BOA optimization to tune the four artificial intelligence models, ultimately building the final models. Furthermore, we compare the aforementioned five optimal artificial intelligence models suitable for five provinces in southern China in the wintertime: VMD-PCA-RF in December 2021 and VMD-PCA-lightGBM in January 2022. We find that the VMD-PCA-RF evaluation indices exhibit relative stability over nearly a year: the correlation coefficient (R) is above 0.6, forecasting accuracy (FA) is above 85 %, mean absolute error (MAE) is below 0.6 m s⁻¹, root mean square error (RMSE) is below 0.8 m s⁻¹, relative mean absolute error (rMAE) is below 60 %, and relative root mean square error (rRMSE) is below 75 %. Thus, for its promising performance and excellent year-round robustness, we recommend adopting the proposed VMD-PCA-RF method for improved wind speed prediction in models.