A general hydrodynamic theory for riming and aggregation was developed by Bohm [1992] and includes a general semi-empirical solution for the collision efficiencies and kernels for riming and aggregation that is in fair agreement with previous studies. The collision efficiencies for aggregation include the effect of variation in fall speed.
Kajikawa [1992] observed free-fall patterns of unrimed ice crystals in both the vertical and horizontal and found that dendritic crystals have an appreciable horizontal velocity of order 5 - 20% of the vertical velocity. He suggests that the variation of the horizontal velocity for platelike crystals may play an important role in the random aggregation of these crystals, especially those having nearly the same size and shape.
Aircraft observations with a new high volume probe by Lawson et al.
[1993] showed the formation of extremely large snowflake aggregates up to
5 cm in diameter within an isothermal layer near 0
C. Individual
crystals in the aggregates consisted of dendrites. Gamache [1990] also
found evidence for enhanced aggregation near 0
C in stratiform
clouds observed in the tropics. The above results are consistent with
previous studies showing that the aggregation of ice crystals is enhanced
near the melting level.
Mitchell [1991] investigated theoretically the influence of assumed super or sub-exponential size spectra for aggregates, and found significant differences in the calculated collision efficiencies and aggregation rate constants.
Recent studies of melting ice have shown that the thermal impact and associated dynamic impact on the atmosphere can be significant. Gedzelman et al. [1993] showed that melting and evaporating snowflakes can have a significant impact on surface air temperatures during cyclonic storms. Marwitz and Toth [1993] showed that melting snow and graupel was the dominant process leading to the development of a barrier jet and mesoscale front during the 6 March Colorado Front Range blizzard of 1990. This jet significantly effected the precipitation production of the storm by creating a region of enhanced lifting east of the Rocky Mountains, resulting in over 1 m of snow falling in the foothills of northeast Colorado. Straka and Anderson [1993] showed through a three-dimensional modeling study that wet microbursts are stronger and deeper when the effects of cooling due to melting are included. These studies support the growing realization that microphysical processes such as melting can often have a significant impact on both the thermal and dynamical structure of the atmosphere.
Johnson and Rasmussen [1992] examined the transition between wet (unfrozen water on the surface) and dry hail growth theoretically, and showed that the transition from dry hail growth to wet growth may often require higher water contents than transitions from wet growth to dry growth. This means that wet growth may be relatively hard to initiate, but once it begins it will tend to continue, even in the face of environmental conditions that would not have been adequate to initiate wet growth to begin with. This hysteresis effect is primarily associated with differences in drag and heat transfer coefficients for smooth and rough particles.