Tropical anvil characteristics and water vapor of the tropical tropopause layer: Impact of heterogeneous and homogeneous freezing parameterizations

Two isolated deep convective clouds (DCCs) that developed in clean-humid and polluted-dry air masses, observed during the Tropical Pacific Warm Pool International Cloud Experiment (TWP-ICE) and U. K. Aerosol and Chemical Transport in Tropical Convection (ACTIVE) campaigns, respectively, are simulated using a three-dimensional cloud-resolving model with size-resolved aerosol and cloud microphysics. We examine the impacts of different homogeneous and immersion freezing parameterizations on the anvil characteristics and the water vapor content (WVC) in the tropical tropopause layer (TTL) for the two DCCs that developed in contrasting environments. The modeled cloud properties such as liquid/ice water path and precipitation generally agree with the available radar and satellite retrievals and in situ aircraft measurements. We find that anvil microphysical properties such as ice number concentration and ice effective radius (rei) are sensitive to the homogeneous freezing parameterizations (HomFPs) under both the clean-humid and polluted-dry conditions, while upper level convection and WVC in the TTL air are only sensitive to HomFPs under the polluted-dry condition. Specifically, the cloud anvil with the Koop et al. (2000) relative humidity dependent scheme has up to 50% and 70% lower ice number than those with other schemes (temperature dependent) for the clean-humid and polluted-dry cases, respectively. Consequently, the rei is increased by 20-30 mu m in both cases. As a result, extinction coefficient of cloud anvils is reduced by over 30%. Anvil size and evolution are also much affected by HomFPs. Higher immersion-freezing rate (Bigg, 1953) leads to a stronger convective cloud due to larger latent heat release resulted from much higher freezing rates, with larger ice water path but less precipitation in both humid and dry conditions. Consequently, the domain-averaged homogeneous freezing rates are enhanced by over 15%. Also, the higher immersion-freezing rate results in over 1.5 times higher ice number concentrations, much reduced rei in the cloud anvil, and larger anvil size. The moistening effect of deep convection on the TTL clear air is very significant, with increases of a few times relative to the WVC before convection under both humid and dry conditions. Different HomFPs does not make much difference in WVC and upper level convection in the clean-humid case, but in the polluted-dry case, the HomFPs with lower nucleation rates predict about 25% lower WVC relative to the HomFPs with higher nucleation rates. Under both humid and dry conditions, the Bigg (1953) immersion freezing predicts about 25% higher WVC relative to the Vali (1975) parameterization, due to stronger transport and the larger anvil area in the domain.