Characterization of Bulk Entrainment Proxy using CloudSat Observations

Abstract

Deep tropical clouds account for a substantial portion of the vertical transport of momentum, heat, and trace gases from the lower troposphere to the upper troposphere and stratosphere. Entrainment or the mixing of the environmental air through the sides of the clouds modulate the vertical extent and strength of these clouds. Therefore, accounting for the entrainment processes is crucial for the realistic estimate of the vertical transport by the deep convective clouds. This study aims to provide a better characterization of entrainment associated with deep tropical convective clouds using CloudSat observations. The deep convective clouds are identified by the deep convective core with an attached anvil. We estimate the bulk entrainment proxy using the difference in the level of neutral buoyancy (LNB) based on CloudSat observation and the same based on the parcel theory. It is found from the diagnostic analysis that the bulk entrainment proxy of deep convective clouds over the land and ocean are significantly different. Convective clouds over the land are associated with lower entrainment rates, and the same over the ocean are associated with relatively higher entrainment rates. In addition to that, the significant diurnal variations in entrainment rates are found to be associated with convective clouds over the land. For example, the entrainment over land is higher during daytime compared to the same during nighttime. Sensitivity analysis showed that the bulk entrainment rate estimation based on CloudSat observations is sensitive to the choice of the parcel source level and the set of constraints imposed on the upward motion of the parcel from the lifting condensation level (LCL) to LNB. For example, the bulk entrainment proxy estimated using the pseudoadiabatic based parcel ascent gives a lower entrainment rate than the same using the reversible moist adiabat based parcel ascent. However, the land-ocean contrast in the bulk entrainment rate is insensitive to the choice of parcel source level and the model of parcel ascent.