Modeling Decadal Organosulfates in Particulate Matter to Improve Estimations of Aerosol Hygroscopicity in the United States

Abstract Organosulfates (OSs) can comprise a substantial fraction of secondary organic aerosol and exhibit significantly different hygroscopicity than inorganic sulfates, yet their spatiotemporal prevalence and climatic impacts remain poorly constrained because speciated measurements are sparse and common chemical transport models do not explicitly represent OS formation. This study develops a flexible and robust modeling framework to quantify isoprene-derived OS (iOS) concentrations and OS impacts on aerosol hygroscopicity (κ) across the contiguous United States (U.S.) for summers spanning 2013–2022. Using the Community Multiscale Air Quality (CMAQ) model with updated iOS chemistry, simulations reproduce Interagency Monitoring of Protected Visual Environments (IMPROVE) network-derived organosulfur and inorganic sulfate and agree with speciated iOS measurements from filter analyses. OS concentrations are elevated in the eastern U.S., with summertime mean values up to 0.51 μg m–3. OS formation produces a net 5.6% decrease in κ (direct OS contribution, +2.8%; indirect, −8.4% via sulfate reduction). We apply an interpretable machine-learning surrogate to identify predictive features associated with OS variability and enable efficient estimation without rerunning CMAQ. These results provide the first observation-validated, high-spatial-coverage OS reconstruction and nationwide quantification of OS-driven κ changes, underscoring the importance of representing OS formation and expanded observations to improve aerosol-cloud radiative effect assessments.