The impact of grid resolution on global dust emission potential

Due to its radiative effects, mineral dust constitutes a critical component in global aerosol climate models. However, the representation of dust emissions currently remains a substantial source of uncertainties in dust model simulations. Convective systems are major contributors to dust emission. Moist convection, however, is still a sub-grid scale process in most climate models, which has to be parameterized. Recent comparison studies between high-resolution, convection-resolving simulations and models with horizontal resolutions, that do not allow for considering moist convection explicitly, have revealed the model resolution as a key driver for the model uncertainties.  To further evaluate the impact of model resolution on dust emission, we conducted an analysis based on surface winds from two distinct modeling frameworks: (i) the coarse-resolution CMIP6 model ensemble, where convection is parameterized, and (ii) high-resolution ICON simulations from the DYAMOND (DYnamics of the Atmospheric general circulation Modeled On Non-hydrostatic Domains) project, which explicitly resolve moist convection. An indicator of dust emissions is the so-called dust emission potential, which is calculated offline for these different datasets and systematically evaluated for key global source regions. The analysis reveals pronounced regional and seasonal differences in the magnitude and characteristics of the modeled dust emission proxy. To investigate the origins of these uncertainties, we further compare the model outputs with high-resolution regridded data and analyze the diurnal cycle of dust emissions in selected source regions with a special focused investigation of the Central Asian dust sources. The results highlight the necessity of using high-resolution emission modeling in specific dust source regions to more accurately represent dust-generating processes and their climate impacts.