Towards Convection-Resolving Dust Emission Modelling

Dust emissions are closely associated with wind speed and are affected by a variety of meteorological drivers and factors that have effects across different spatial and temporal scales. Global or regional atmospheric dust models employing parameterized convection often encounter difficulties in accurately replicating observed dust emissions. Recent investigations by Garcia-Carreras et al. (2021) have demonstrated significant discrepancies when modeling Northern African dust emissions across various grid scales using either parametrized convection or resolved convection. In order to further clarify the influence of model resolution on dust emissions, an investigation was conducted employing surface winds from two different model studies: the coarse-resolution CMIP-6 model intercomparison study [Eyring et al. (2016)] with parameterized convection and the high-resolution ICON model  simulation that was part of the DYAMOND project [Stevens et al. (2019)], which was computed with explicit convection. Two different dust products were computed using the modelled surface winds: the Dust Uplift Potential (DUP) derived from wind data and an offline dust emission model based on Tegen et al. (2002), which incorporates soil and vegetation effects to simulate dust emission fluxes utilizing gridded surface wind fields. The dust emissions from the different models are evaluated across various source regions, including Northern Africa, the Arabian Peninsula, Central Asia, the Gobi Desert, and the Taklamakan Desert. Convective events such as haboobs particularly necessitate explicit modeling at convection-resolving resolution, which is e.g an important cause of dust emissions in the southern Sahara in northern hemisphere summer. Other local wind systems can be discerned by both high and low-resolution models, albeit at varying magnitudes. In the Gobi region, there is negligible impact of model resolution on dust emissions. These findings could inform further research on modeling dust emission and  transport by providing a basis for improved dust emission parameterizations in large-scale models.