Volatile chemical partitioning during cloud drop freezing and its effects on tropospheric chemical distributions: Modeling studies

Amy Stuart

Stanford University

Convective clouds impact airborne pollutant concentrations and precipitation chemistry through redistribution of air and hydrometeors containing trace chemicals, and by providing a multi-phase environment for chemical phase changes and reactions. Interactions between ice-containing cloud hydrometeors and chemicals, and their effects on tropospheric chemistry, are not well understood. Laboratory and field measurement of chemical partitioning during freezing provide greatly varying estimates of the retention efficiency of volatile solutes. The reasons for this variance are as yet unknown. In this work, we use modeling at the storm and drop scales to elucidate the processes involved in this partitioning and its likely effects on post-storm chemical distributions and deposition.

We first present results of storm-scale modeling used to investigate the range of effects on post-storm chemical fate of

partitioning during microphysical processes that convert liquid hydrometeors to ice, snow, or hail. For simulations we used the COllaborative Model for Multiscale Atmospheric Simulation (COMMAS) augmented to include gas- and aqueous-phase chemistry. The model is a three dimensional, time varying, and nonhydrostatic. Two bounding parameterizations of chemical partitioning to ice were implemented; the first assumes complete evaporation of volatile solutes during liquid-to-ice hydrometeor transformation and the second assumes complete retention of volatiles by the ice-containing hydrometeor.

Modeling was performed to simulate the conditions of the July 10, 1996 STERAO field study storm. Results indicate that entrapment in the ice-phase may have significant impacts on chemical spatial and phase distributions and the flux of chemical mass into the upper troposphere and to the ground. Allowing entrapment in ice hydrometeors led to losses of mass of moderately soluble trace species from the cloud anvil region and increases of mass deposited to the ground, in comparison with allowing species to degas during hydrometeor freezing. Chemical transfer to hail and precipitation of hail and rain (formed from melting hail) were largely responsible for these differences.

Second, we present the development and preliminary results from a numerical model of chemical transfer during drop freezing. The model simulates the time-dependent heat and chemical mass transfer in the liquid and solid phases between radial shells of a supercooled spherical drop of adjustable size. Freezing is initiated at the center by a spherical nuclei of adjustable size. The kinetics of freezing are simulated using a temperature dependent prognostic equation, developed to represent dendritic growth characteristics in a bulk manner. Latent heat release due to freezing is assumed to be a volume source in each shell and heat redistribution after each freezing timestep is determined diagnostically using enthalpy conservation equations. Prognostic diffusion equations for radial heat and mass transfer are solved for each phase and include transfer between phases. Concentration dependent interphase mass transfer during freezing is calculated using an equilibrium distribution coefficient at the ice/water interface and allows for trapping if shell freezing is complete. Heat and chemical mass transfer to/from the drop environment are calculated using time-dependent flux equations that account for turbulent enhancement to transfer due to drop velocity in air. Preliminary results of model simulations indicate that it represents the processes involved in chemical retention during freezing in a physical manner. Freezing times calculated by the model agree well with bulk freezing calculations.

We plan to apply the model to simulate retention under a variety of conditions relevant to freezing in natural clouds and for specific cases representative of experimental and field study conditions. We hope to elucidate the differences between available retention measurements and move toward developing a robust parameterization of partitioning to ice for cloud modeling purposes.