Table: Ship-induced changes in the mean shortwave (SW) cloud radiative effect (CRE) at the top of atmosphere and cloud optical thickness τ. Ship simulations were run at resolutions listed, whilst in the Dil simulations merely the ship particle flux was coarse grained to resolutions listed in table.
In a preceding study we demonstrated that the physical parameterisations were capable of simulating the formation and evolution of ship tracks within a stratocumulus-topped shallow boundary layer at the kilometre scale. Here we investigate whether the agreement between simulations and observations can be maintained at resolutions approaching the horizontal grid spacing of current global climate models (GCMs).
The simulations showed that the CRE triggered by ship exhaust was increasingly overestimated (almost up to a factor of three) at coarser resolutions. About one third of this overestimation can be attributed to the plane-parallel albedo bias. The remainder of the bias is due to the underestimated vertical mixing of the aerosols at coarse resolutions which leads to enhanced formation of activation sized particles in highly concentrated regions of the exhaust plume.
In summary, these results show that the negligence of subgrid-scale aerosol variability may induce large cloud radiative biases in GCM estimates of aerosol-cloud-radiation interactions in low-lying stratocumulus clouds.