Trouble in the natural laboratory: Suppression of aerosol-induced entrainment enhancement in ship tracks

Covariation of aerosol loadings and meteorological conditions present a fundamental challenge to understanding the effects of aerosols on clouds. Studies of sufficiently narrow aerosol plumes provide a natural laboratory in which aerosol effects can be isolated from the larger scale meteorological variability. Ship tracks have been studied for decades for this very reason, with on the order of 100 tracks intercepted by instrumented aircraft and a much larger sample studied using passive remote-sensing from polar orbit (AVHRR, MODIS). The remote-sensing studies indicate that, on average, liquid water path (LWP) in ship tracks is reduced compared to the surrounding clouds, while there is a lack of statistically significant change in the more limited sample of in-situ measurements. Modeling studies have suggested that reductions in LWP result from enhanced entrainment of dry overlying air induced by a reduction in the size of cloud droplets, while increased LWP is associated with moist air aloft and reduction of drizzle reaching the surface. However, the passive remote-sensing studies find no statistical change in cloud-top altitude between ship tracks and the surrounding clouds, as would be expected from enhanced entrainment. Perhaps significantly, the model studies that found enhanced entrainment in polluted clouds were not simulations of ship tracks embedded within cleaner clouds. Rather than designed to study ship tracks specifically, they were designed to consider the effects of aerosols on clouds more generally, and thus were based on simulations in which the specified aerosol concentration was uniform throughout the model domain, and that aerosol concentration was increased over a sequence of simulations.

To overcome this deficiency, we have modeled ship tracks embedded within fields of ambient clouds using large-eddy simulations with bin microphysics. To distinguish the ship tracks from ambient, unpolluted clouds in our simulations, we emulate the method that Coakley and co-workers devised for MODIS radiances, applying it to near-infrared reflectance averaged over 0.5 x 0.5 km areas. In simulations of overcast stratocumulus, based on a case study from the FIRE-I field campaign, we find that the entrainment rate in the ship track is negligibly enhanced, in contrast to simulations of equivalent clean and polluted domains in which the entrainment is enhanced by over 50% in the polluted case relative to a clean case. The difference in behavior suggests a negative feedback that inhibits any increase of entrainment over the ship track, resulting from the negative buoyancy of cold, boundary-layer air over the ship track relative to the warm air above the surrounding, ambient clouds. We also find that the entrainment over the ship track is somewhat increased when the air aloft is cooler or moister (with downwelling longwave flux fixed), as expected from the thermodynamic arguments made by Randall (1984), but the increase is insufficient to result in a substantial enhancement of entrainment over the ship track. Thus far we have investigated only overcast, drizzling cases in which LWP increases in the ship tracks. We will also present overcast cases in which LWP decreases, as well as ship tracks arising from fields of broken clouds, the latter of which we will compare and contrast with the recent simulations of ship tracks in open cellular convection by Wang and Feingold (2009).