Graduation Year


Document Type




Degree Granting Department

Engineering Science

Major Professor

Amy L. Stuart, Ph.D.

Committee Member

Jeffrey Cunningham, Ph.D.

Committee Member

Jennifer M. Collins, Ph.D.

Committee Member

Maya A. Trotz, Ph.D.


Atmospheric chemistry, Riming, Retention, Ice microphysics, Cloud modeling


Phase partitioning during freezing of hydrometeors affects the transport and distribution of volatile chemical species in convective clouds. Here, the development, evaluation, and application of a mechanistic model for the study and prediction of partitioning of volatile chemical during steady-state hailstone growth are discussed. The model estimates the fraction of a chemical species retained in a two-phase growing hailstone. It is based upon mass rate balances over water and solute for constant accretion under wet-growth conditions. Expressions for the calculation of model components, including the rates of super-cooled drop collection, shedding, evaporation, and hail growth were developed and implemented based on available cloud microphysics literature. A modified Monte Carlo simulation approach was applied to assess the impact of chemical, environmental, and hail specific input variables on the predicted retention ratio for six atmospherically relevant volatile chemical species, namely, SO2, H2O2, NH3, HNO3, CH2O, and HCOOH. Single input variables found to influence retention are the ice-liquid interface supercooling, the mass fraction liquid water content of the hail, and the chemical specific effective Henry's constant (and therefore pH). The fraction retained increased with increasing values of all these variables. Other single variables, such as hail diameter, shape factor, and collection efficiency were found to have negligible effect on solute retention in the growing hail particle. The mean of separate ensemble simulations of retention ratios was observed to vary between 1.0x10-8 and 1, whilst the overall range for fixed values of individual input variables ranged from 9.0x10-7 to a high of 0.3. No single variable was found to control these extremes, but rather they are due to combinations of model input variables.