Winter Storms and Orographic Clouds

Mesoscale aspects of winter storms are reviewed by Businger [1995] in this issue. In the following we consider microphysical aspects of winter storms.

A number of winter storm projects were conducted just prior to or during this four year period, including the Winter Icing and Storms Project (WISP) in Colorado (Rasmussen et al. [1992]), the Lake Ontario Winter Storms Project (LOWS; Reinking et al. [1993]), and the University of Illinois Winter Precipitation Project in Illinois (UNIWIPP; Ramamurthy et al. [1991]). These projects focussed on mesoscale aspects of winter storms, including the formation and evolution of mesoscale snowbands, the production and depletion of supercooled liquid water, and the formation of freezing rain. A hazardous winter storm that crossed the continental United States in mid-February 1990 was observed by all three projects, with freezing drizzle and snow occurring in Colorado, and freezing rain in Illinois and near Lake Ontario (Martner et al. [1992]). Estimated impacts from this storm were 9 deaths, 27 injuries, and $120 million dollars worth of damage. The STORM (Stormscale Operational and Research Meteorology) Field Experiment and Systems Test (STORM-FEST) was conducted during winter 1992; only a few results from this study have appeared in the literature to date.

Winter storms in the western United States typically interact with topography during their eastward propagation. Ongoing and past weather modification programs have collected a wealth of data during the passage of winter storms over the topography of the west. A series of observational papers on a storm passing over the Tushar mountains in southwestern Utah (Long et al. [1990], Sassen et al. [1990], Campistron et al. [1991]) showed that the storm above the barrier was dominated by barrier level orographic clouds or propagating mesoscale cloud systems. The orographic cloud component consisted of weakly supercooled liquid water (-3 to -7C) in the form of an extended barrier wide cap cloud. The spatial supercooled liquid water distribution was linked to local topographic features such as abrupt rises and ridges. This orographic cloud produced precipitation primarily through riming of particles sedimenting from aloft, and also to some extent through a Hallett-Mossop ice multiplication process during graupel growth. In contrast, mesoscale precipitating bands associated with a slowly moving cold front generated much more significant amounts of snowfall. These precipitation bands periodically disrupted the shallow orographic supercooled liquid water clouds. This series of papers provides a relatively complete study of the passage of a winter storm through a region of complex terrain.

An observational and modeling study of a winter storm passage over the mountains of northern Arizona (Bruintjes et al. [1994]) also showed the development of a shallow orographic cloud with supercooled liquid water close to the barrier, and in addition showed that gravity waves produced by the complex orography produced enhanced regions of supercooled liquid water and ice crystals. As a result, cloud and precipitation development were sensitive to wind speed and direction as a result of gravity wave excitation. The gravity waves extended through deep layers of the atmosphere with substantial vertical velocities in excess of 5 ms-1. Abbs and Jensen [1993] show similar results for the Baw Baw Plateau in Australia with a shallow cap cloud forming over the plateau, standing lee waves downstream of the plateau, and a region of high liquid water extending 50 km upstream of the plateau. The production of gravity waves was also shown to occur in the Sierra Nevada mountain range as a result of the interaction of the northerly barrier jet with local east-west ridges in a modeling study by Rasmussen et al. [1988]. The above results show the importance of gravity waves produced by the flow of air over local ridges and valleys in the production of clouds and precipitation over complex terrain.

Peterson et al. [1991] showed through observations and numerical modeling that the low-level decoupled flow created as a result of flow blocking upstream of the Rocky Mountains causes part of the orographic lift of the mountain barrier to be experienced well upstream of the barrier. This changes the location of condensate production which in turn shifts the precipitation formation upstream. Rasmussen and Smolarkiewicz [1993] show this to be true upstream of the island of Hawaii as well in association with flow blocking effects created by the low Froude number flow past the tall volcanoes on the island.

Super and Huggins [1993] analyzed supercooled liquid water flux and precipitation amounts over four different mountain barriers in Arizona, Colorado, and Utah and found that none of the datasets supported the concept that large precipitation producing storms are highly efficient in converting supercooled liquid water to snowfall. While phases of storms producing heavy snowfall contained relatively little supercooled liquid water, other phases contained relatively little snowfall and relatively high amounts of supercooled liquid water.

Marwitz and Toth [1993] analyzed the Colorado Front Range Blizzard of 1990 and showed that melting graupel and snow led to the development of a barrier jet and a mesoscale front. Both of these features were important to the production of the observed heavy snowfall amounts in this storm.

The above studies have revealed some of the complex interactions of winter storms with topography and show that the use of three-dimensional models can be extremely useful in determining the type of interaction moist air flows can have with complex topography.

A number of studies have focussed on the precipitation structure of winter storms over relatively flat terrain (Martner et al. [1993], Rauber et al. [1994], Rasmussen et al. [1993], Wesley et al. [1995], Rasmussen et al. [1995] and Sousounis [1993]). Martner et al. documented and identified signatures of melting aloft that may be useful for the prediction of freezing rain using remote sensing instruments. Rauber et al. [1994] investigated the causes for a freezing rain storm in Illinois and showed that convective instability was the primary process leading to the freezing rain. They also showed that the melting and sublimation of the ice on the ground kept the surface temperature near 0C for an extended period of time. Rasmussen et al. [1995] documented the structure and evolution of the shallow upslope cloud forming along the Colorado Front Range during the passage of the 13-15 February 1990 winter storm through Colorado. During the initial stages of the upslope cloud over 30 hours of continuous supercooled liquid water was present in the Front Range. This supercooled liquid water was produced by: 1) upglide over the gently sloping terrain leading into the Front Range, 2) uplift associated with local ridges, 3) uplift associated with the Rocky Mountain barrier, 4) low-level convergence, and 5) boundary layer turbulence. Depletion of supercooled liquid water occurred by growing ice crystals. Freezing drizzle occurred in the southern portion of the domain as a result of uplift over a stationary front at that location.

Wesley et al. [1995] showed that topography-induced circulations in the lee of a local ridge can ennhance snow formation as a result of enhanced low-level convergence.

Rasmussen et al. [1993] described the formation of snowbands during the 15 November 1987 aircraft accident at Denver Airport and showed that the snowbands formed following the passage of a cold front. The bands were located above the cold front and moved in the opposite direction of the cold front movement. The bands formed by a convective instability in the weak shear layer above the cold front and moved with the mean flow in the layer. The snowfall during the time of the crash occurred when one of these bands stalled over the airport as a result of the cloud layer winds shifting to band parallel as the upper-level wave moved eastward.

Sousounis [1993] showed through a modeling study that snowfall amounts from lake effect snowstorms depend critically on wind speed. Largest snow accumulations on the lee shore of the lake were produced by moderate wind speeds between 4 - 6 m s. Higher wind speeds yielded very little, if any, significant snowfall on the lee side of the lake.