Refining simulated mineral dust composition through modified size distributions: dual validation with mineral-specific and elemental observations

Abstract. The JATAC2022 campaign in Cape Verde provided a unique opportunity to collect mineral dust aerosols from multiple Saharan source regions and characterize their composition. Mineral dust aerosols comprise a complex assemblage of minerals with distinct physico-chemical properties, leading to differentiated climatic impacts through interactions with radiation, cloud microphysics, and atmospheric chemistry. A crucial physical property governing these interactions is the particle size distribution (PSD), which strongly influences aerosol optical properties, transport, and deposition. Although contemporary atmospheric models have begun integrating mineralogical data into their dust aerosol representations, implementation faces complications due to variations in dust emission parameterizations, making some models more compatible with existing soil mineralogical databases than others.

This work addresses the challenges encountered when incorporating mineralogical information into the COSMO5.05-MUSCAT atmospheric model, which employs a dust emission scheme based on Marticorena and Bergametti (1995). We present an improved approach that refines the translation of mineralogical soil PSDs into emitted aerosol PSDs. The revised implementation is evaluated using historical Saharan dust measurements and new mineralogical observations from the JATAC2022 and DUSTRISK2022 campaigns. Model performance is assessed using a dual validation framework considering both mineral-resolved and elemental composition. The elemental validation approach provides complementary constraints that expose discrepancies in internal mixing assumptions and reveal limitations invisible to mineral-only comparisons. Results indicate that the proposed modification substantially improves representation of phyllosilicates, quartz, and feldspar, while biases in iron, calcium, and magnesium highlight fundamental challenges in representing the heterogeneous internal structure of natural dust particles.