Prediction of degree of riming within a bulk microphysical scheme

Bulk microphysical schemes typically represent snow particles in a broad-brush, simplistic way. One of the more important growth mechanisms for precipitation is the collection of cloud water by snow (“riming”). Bulk schemes typically represent the effects of riming by apportioning the collected mass to the graupel and snow categories, according to the rate of riming mass growth compared to other mechanisms. However, the properties of the snow hydrometeor species are not altered to account for the different physical characteristics of rimed versus unrimed snow, because no information is retained on the fraction of snow mass that came from riming, as compared to that which came from depositional or other growth mechanisms. Thus, the precipitation is forced to behave either as unrimed snow or as graupel, with no option for a hybrid behavior characteristic of moderately to densely rimed. We describe in this presentation an attempt to resolve this situation, by introducing another prognostic variable into the bulk scheme that tracks the rimed mass of snow. Inclusion of this fully prognostic variable allows for the specification of variable properties of snow on a sliding scale between that of unrimed snow and that of graupel. The properties of greatest interest are the assumed relationships of particle fall velocity and mass as a function of diameter, both of which enter into nearly every microphysical production term in the scheme that involves snow. The variable riming capability is found to make a substantial differences in precipitation behavior in a test case from the IMPROVE-2 field study, particularly in the region of orographic precipitation where snow forming in deep frontal precipitation falls through orographically enhanced cloud liquid water. The interaction of the variable riming capability with our recently developed snow habit prediction scheme is also described.