The resolution dependence of cloud effects and ship‐induced aerosol‐cloud interactions in marine stratocumulus

Abstract Measures of aerosol‐cloud interactions in stratocumulus have been shown to depend on the resolution of the applied data set. In order to contrast resolution with emission dilution effects in models, a regional numerical weather prediction model is used to simulate ship tracks at a range of spatiotemporal resolutions ranging from the global climate modeling scale (Δx = 50 km, Δt = 180 s) to the convection‐resolving scale (Δx = 1 km, Δt = 20 s). The background simulations without ship emissions display a high degree of similarity in the planetary boundary layer and cloud properties at all spatiotemporal resolutions. Simulations assessing the impact of emission dilution show an increasing overestimation of the shortwave (SW) cloud radiative effect (CRE) with degenerating emission resolution. Although mean perturbations in the activation‐sized aerosol number concentration (Nact) are similar for all dilution experiments, the variability in Nact is increasingly lost with stronger emission dilution. The enhanced Nact homogeneity in turn leads to an overestimated SW CRE. We show that emission dilution alone accounts for 47% of the overestimated SW CRE simulated at low resolutions. The remainder of the differences is attributed to a combination of locally enhanced aerosol concentrations due to weaker vertical mixing simulated at coarse resolutions, in combination with a faster conversion rate of Aitken to accumulation mode particles by redistribution in these regions.