| In this dissertation, I investigate volatile chemical partitioning during hydrometeor freezing (termed ‘freezing retention’). Three-dimensional convective cloud simulations of soluble tracer and reactive chemical redistribution were performed for one field project storm. Two bounding parameterizations of freezing retention, (1) complete solute degassing from the solid hydrometeor and (2) complete solute retention, were compared. Retention during freezing led to more scavenging, more ground deposition, and less upward vertical transport of moderately to highly soluble species (Henry's constants greater than 10 5 M/atm). Sedimentation in hail and rain played a significant role in species redistribution. By analyzing the hydrometeor-scale processes involved in retention for non-rime freezing, dry-growth riming, and wet-growth riming, I investigated the factors that control it. For non-rime freezing, I developed a theoretical dimensionless number to indicate retention, derived its dependence on conditions and chemical properties, and calculated its value for several freezing cases for SO2, H2O2, NH3, and HNO3. Retention is apparently highly chemical specific, controlled largely by the effective Henry's constant (and hence the drop pH for dissociating chemicals). Chemicals with high effective Henry's constants (HNO3) will be fully retained during freezing, while chemicals with lower effective Henry's constants (SO2) will undergo some loss. For chemicals that undergo loss, the degree of retention depends on freezing conditions. It likely increases with decreasing temperature and exhibits a maximum at intermediate drop sizes and ventilation. For dry-growth riming, I extended my development to predict retention and compared predicted values to experimental data from several measurement studies. The model agrees well with the data and provides the first quantitative explanation for the differences in measured retention. For wet-growth riming, I developed a steady-state retention model. It suggests retention is dependent on the fraction unfrozen water in the riming hydrometeor. Finally, I developed and demonstrated a one-dimensional time-dependent numerical model of freezing retention. The model results represent freezing and solute transfer in a physical manner, consistent with available data. Results from this dissertation improve our understanding of freezing retention, provide theory-based hypotheses regarding its dependence on physical factors and chemical properties, and provide a framework for robust parameterization in cloud models. |
