Evaluation of a physically-based dust emission scheme in the Community Earth System Model (CESM)
Thursday, 18 December 2014
The complex nature of dust aerosol emission makes it difficult to accurately represent the dust cycle in global circulation models (GCMs). Indeed, both measurements and the new physically-based dust emission parameterization of Kok et al. (2014) indicate that many GCMs underestimate the dust flux’s sensitivity to soil erodibility. This finding can explain why dust cycle simulations in many GCMs are improved by using an empirical preferential sources function that shifts emissions towards the most erodible regions. To both test this hypothesis and evaluate the performance of the new Kok et al. (2014) dust emission scheme, we implemented this scheme in the Community Earth System Model (CESM) and compared its results against measurements. We find that the physically-based scheme produces even better agreement against measurements most representative of dust emission, specifically aerosol optical depth measurements near dust source regions, than simulations that use a source function. This apparent improvement in the representation of dust emission in CESM propagates into a modest improvement against measurements of surface concentration and dust deposition, which tend to be further from source regions and depend on a range of other uncertain model processes besides dust emission. These results indicate that the need to use a source function is at least partially eliminated by the additional physics accounted for by the Kok et al. (2014) parameterization, and the increased sensitivity to soil erodibility that it produces. Since soil erodibility is affected by climate changes, our results further suggest that many models might underestimate the climate sensitivity of the global dust cycle.