Confronting structural uncertainty in aerosol-cloud interactions through process-level benchmarking

Aerosol interactions with clouds and radiation are a large contributor inter-model variability in radiative forcing among climate models, and this variability has remained essentially unchanged in the past two decades. Climate-relevant aerosol properties depend critically on the distribution in size, shape, and chemical composition of particle populations, but tracking such particle-level details is computationally impractical for large-scale, long-running climate simulations. Instead, aerosol modules in Earth System Models necessarily simplify the representation of particle characteristics, with different aerosol models different approximations. These differences in the aerosol schemes and error introduced by approximating particle distributions likely contribute to the inter-model variability in aerosol forcing, but structural uncertainty from numerical approximations of particle distributions have not yet been well quantified. To address this gap, we introduce the Aerosol Model Benchmark Repositories and Standards (AMBRS) framework for quantifying error in reduced aerosol schemes. AMBRS provides a set of test cases and analysis code for performing box model comparisons between reduced aerosol schemes and the particle-resolved benchmark model, PartMC-MOSAIC. In this presentation, we will demonstrate our benchmarking framework for a subset of aerosol schemes, with a focus on quantifying the factors that most contribute to errors in cloud condensation nuclei activity of aerosol populations. We hope this presentation will lead to a wider discussion on the role of process-level benchmarking in understanding aerosol-cloud interactions.