Under global warming, extreme weather events and air pollution are becoming increasingly critical challenges. Both pose serious risks to human health, economies, and societal stability, and their complex interactions can further amplify these impacts. Numerical models are essential tools for studying these phenomena; however, traditional low-resolution Earth system models often fail to accurately capture the dynamics of extreme weather and air pollution. This limitation hinders our mechanistic understanding, reduces the reliability of future projections, and constrains the development of effective adaptation strategies. Dynamical downscaling—an approach that uses high-resolution regional models nested within global models—offers a partial solution. However, this method inherits biases from the parent global models and often fails to adequately represent multi-scale and cross-sphere interactions involving the atmosphere, land, and oceans. These shortcomings underscore the growing need for developing and applying high-resolution Earth system models that can more comprehensively and accurately depict land–sea–atmosphere interactions, including heat and material exchanges and their spatial heterogeneity. This article explores the current challenges, recent advances, and future opportunities in understanding the interplay between extreme weather events and air pollution, with a focus on the critical role of high-resolution modeling.

● CS structure overestimates ρ _eff by nearly six times at externally mixed states.

Measurements studies have shown that the absorption of radiation by black carbon (BC) increases as the particles age. However, there are significant discrepancies between the measured and modeled absorption enhancement (E_abs), largely due to the simplifications used in modeling the mixing states and shape diversities. We took advantage of chamber experiments on BC aging and developed an efficient method to resolve the particle shape based on the relationship between the coating fraction (∆D_ve/D_ve,0) and fractal dimension (D_f), which can also reflect the variations of D_f during the whole BC aging process. BC with externally and partly mixed states (0 ≤ ∆D_ve/D_ve,0 ≤ 0.5) can be considered to be uniformly distributed with the D_f values of 1.8–2.1, whereas the D_f values are constrained in the range 2.2–2.8 for fully mixed states (∆D_ve/D_ve,0 > 0.5). The morphological parameters (i.e., the effective density and the dynamic shape factor) were compared with the measured values to verify the simulated morphology. The simulated mean deviations of morphological parameters were smaller than 8% for the method resolving the particle shape. By applying a realistic shape and refractive index, the mass absorption cross for fully mixed states can be improved by 11% compared with a simplified core–shell model. Based on our understanding of the influence of D_f and ∆D_ve/D_ve,0 on E_abs, we propose a two-stage calibration equation to correct the E_abs values estimated by the core–shell model, which reduces the simulation error in the Mie calculation by 6%–14%.