A Theory for Why Some Clouds Produce Snow

Mixed-phase clouds consist of both supercooled cloud-liquid and ice particles.  They are influential for the Earth’s radiation budget.  Snow reaches the ground typically from mixed-phase nimbostratus cloud.  For humanity, deep snowfalls are influential as they cause much disruption (e.g. to transportation), with a cost of billions of euros annually .   

Both the warm rain (coalescence) and ice crystal (vapour growth of crystals, perhaps followed by aggregation) processes of precipitation can co-exist in mixed-phase clouds.  A cloud base that is not too warm, depending on aerosol conditions, is typically needed for the ice crystal process to prevail in precipitation production, because otherwise an abundance of cloud-liquid mass can promote coalescence before parcels become supercooled, as with deep tropical convective clouds.   

Our theory published in 2024 explained why any competition between both cold and warm processes of precipitation in mixed-phase clouds tends to be won by the ice crystal process.  Since the fall-out of snow is slow, boosting its mass aloft, and its low bulk density creates a wide cross-sectional area for riming, the supercooled cloud-liquid mass is kept weak by the ice crystal process.  This then reinforces the ice crystal process by minimizing the liquid water content, favouring snow production. 

The question of why snow reaches the ground, whether intact or as a melted drop, is partly related also to the issue of why graupel or hail is not produced instead.  Snow may rime to produce graupel or hail.  Precipitation particles tend to be defined by the intensity of the ascent.  Snow particles are balanced against stratiform ascent (< 1 m/s) as it is comparable to their fall-speeds.  This is partly why nimbostratus produces deep snowfalls. Graupel/hail tends to fall much faster. But also wintertime deep convection can produce snow at the ground, as sometimes seen in thunderstorms near the Sea of Japan. 

On this topic, Steiner and Smith in 1998 theorized that there is a phase-space of in-cloud vertical velocity and temperature in which a region of predominant riming and supercooled cloud-liquid exists in a ‘wedge’ within the convective ascent.  Steiner and Smith argued that predominant aggregation for snow is restricted to weak stratiform ascent since at faster convective ascent there is predominant riming.

In this presentation we analyse with a single-crystal growth model the conditions of ascent and temperature determining whether snow or graupel fall out from the mixed-phase region.  The model predicts the evolution of a crystal growing first by diffusional growth in various habits and then by aggregation of crystals and riming, with the chance of becoming either graupel or hail.   We reproduce the wedge in the phase-space by Steiner and Smith, and analyse contributions from aggregation, sticking efficiency and riming.  It is predicted that repeated recirculation cycles and aggregation of crystals by snow is needed to explain the wedge.   

In summary, snow reaches the ground partly because the ice crystal process tends to prevail in mixed-phase clouds, while aggregation of ice crystals and related processes (e.g. habit-dependent sticking efficiency) in weak ascent combine to prevent snow from becoming graupel/hail.