HR: 11:35h
AN: V42A-05 [Abstracts]
TI: Lattice Boltzmann Simulation of Kinetic Isotope Effect During Snow Crystal Formation
AU: * Lu, G
EM: gplu@lbl.gov
AF: Lawrence Berkeley National Laboratory, 1 Cylcotron Road, ESD, Berkeley, CA 94720,
United States
AU: DePaolo, D J
EM: depaolo@eps.berkeley.edu
AF: Lawrence Berkeley National Laboratory, 1 Cylcotron Road, ESD, Berkeley, CA 94720,
United States
AU: DePaolo, D J
EM: depaolo@eps.berkeley.edu
AF: University of California at Berkeley, EARTH & PLANETARY SCIENCE, 307 McCone Hall,
Berkeley, CA 94720, United States
AU: Kang, Q
EM: qkang@lanl.gov
AF: Los Alamos National Laboratory, EES Divison P.O. Box 1663, Los Alamos, NM 87545, United States
AU: Zhang, D
EM: donzhang@usc.edu
AF: University of Southern California, Department of Civil and Environmental Engineering, Los
Angeles, CA 90089, United States
AB:
The isotopic composition of precipitation, especially that of snow, plays a special role in the global hydrological
cycle and in reconstruction of past climates using polar ice cores. The fractionation of the major water isotope
species (HHO, HDO, HHO-18) during ice crystal formation is critical to understanding the global distribution of
isotopes in precipitation. Ice crystal growth in clouds is traditionally treated with a spherically-symmetric steady
state diffusion model, with semi-empirical modifications added to account for ventilation and for complex crystal
morphology. Although it is known that crystal growth rate, which depends largely on the degree of vapor over-
saturation, determines crystal morphology, there are no quantitative models that relate morphology to the vapor
saturation factor. Since kinetic (vapor phase diffusion-controlled) isotopic fractionation also depends on growth
rate, there should be direct relationships between vapor saturation, crystal morphology, and crystal isotopic
composition. We use a 2D lattice Boltzmann model to simulate diffusion-controlled ice crystal growth from vapor-
oversaturated air. In the model, crystals grow solely according to the diffusive fluxes just above the crystal
surfaces, and hence crystal morphology arises from the initial and boundary conditions in the model and does
not need to be specified a priori. Crystal growth patterns can be varied between random growth and deterministic
growth (along the maximum concentration gradient for example). The input parameters needed are the isotope-
dependent vapor deposition rate constant (k) and the water vapor diffusivity in air (D). The values of both k and D
can be computed from kinetic theory, and there are also experimentally determined values of D. The deduced
values of k are uncertain to the extent that the condensation coefficient for ice is uncertain. The ratio D/k is a length
(order 1 micron) that determines the minimum scale of dendritic growth features and allows us to scale the
numerical calculations to atmospheric conditions. Our calculations confirm that the crystal/vapor isotopic
fractionation approaches the equilibrium value, and the crystals are compact (circular in 2D) as the saturation
factor approaches unity (S= 1.0). However, few natural crystals form under such conditions. At higher
oversaturation (e.g. S = 1.2), dendritic crystals of millimeter size develop on timescales appropriate to cloud
processes, and kinetic effects control isotopic fractionation. Fractionation factors for dendritic crystals are similar
to those predicted by the spherical diffusion model, but the model also gives estimates of crystal heterogeneity.
Dendritic crystals are constrained to be relatively large, with dimension much greater than about 20D/k. The most
difficult aspect of the modeling is to account for the large density difference between air and ice, which requires
us to use a fictitious higher density for the vapor-oversaturated air and scale the crystal growth time accordingly.
An approach using a larger scale simulation and the domain decomposition method can provide a vapor flux for a
nested smaller scale calculation. The results clarify the controls on crystal growth, and the relationships between
saturation state, growth rate, crystal morphology and isotopic fractionation.
DE: 0320 Cloud physics and chemistry
DE: 0515 Cellular automata
DE: 0545 Modeling (4255)
DE: 0736 Snow (1827, 1863)
DE: 1041 Stable isotope geochemistry (0454, 4870)
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2007 Fall Meeting