Precipitation Production in a Large Montana Hailstorm: Airflow and Particle Growth Trajectories

Abstract

Computations of air motion and precipitation growth using winds derived from Doppler radar measurements were analyzed to reveal important flow features that influenced the production of precipitation during the nearly steady phase of a well-observed severe storm in Montana that produced hail as large as 5 cm in diameter. The storm had many features commonly associated with supercells, though it exhibited a gently sloping overhang on its low-level inflow side, rather than the more classical vaulted structure. Formed initially as the right member of a splitting storm pair, it moved slowly eastward while embedded in moderately sheared environmental winds. Characteristic hail growth trajectories and precipitation fallout positions are considered in conjunction with the deduced embryo sources and formation regions. Based on particle growth calculations, measurements by radar and research aircraft, cloud photography and direct hailstone examination, four general sources of hail embryos were apparent: 1) graupel grown along the updraft fringes, 2) a derivative of the former consisting of drops produced by melted graupel, 3) water drops shed from melting hail, and 4) shedding from hailstones that were in wet growth conditions. The graupel embryos were deduced to originate primarily in two columnar regions on the flanks of the updraft core. One column was within the stagnation zone on the west side (upwind with respect to the midlevel flow), and the other one was at the center of the midlevel, mesocyclonic circulation on the cast (downwind) side. The cyclonically streaming branch of the flow transported some graupel from the west flank to the southwest and south sides of the storm where a fraction of them melted completely before entering the strong updraft. The occasional merger of drifting, isolated cumulus congestus clouds with the storm and the ingestion of graupel from them was also documented. Following the embryo growth stage, three types of hail growth trajectories were found: 1) those passing into the southern (cyclonic) branch of the middle-to-upper level airflow, 2) those passing into the northern (anti-cyclonic) branch of this flow, and 3) those passing in a nearly straight line through the updraft core in midlevels (preferentially the northeastern side of the core). Of these the straight-line trajectory produced the largest hail. Growth trajectories were mostly of a simple up-and-down nature, without multiple passes or loops through the main updraft. Particles that followed cyclonic trajectories produced a broad maximum in reflectivity cast of the updraft, while those that followed straight-line trajectories produced a similar broad maximum west of the updraft. Large hail from straight-line trajectories and some graupel and hail from the cyclonic branch passed through the region that otherwise would have been the echo-weak vault. A subset of the cyclonic trajectories appeared able to loop back on themselves in such a way that low-level melting and breakup of graupel following this circuit could have led to a self-sustaining mode of precipitation growth. Precipitation from graupel grown in the western updraft fringes, from drops produced by transport and melting of these graupel, and from graupel grown in the northern updraft fringes was necessary to explain the observed patterns of radar reflectivity, dual-wavelength ratio and specific attenuation. Further, only embryos from the west and south flanks led to large diameter hail near where stones of similar sizes were observed from aircraft and at the ground. Precipitation from the remaining source east of the updraft maximum duplicated only the central portion of fallout from the other sources. Though embryo transport from the upwind side of the updraft core and around its south side was a necessary deduction from aircraft and radar observations, details of the early particle formation within the upwind region were not well documented. This region was characterized as a vigorous cluster of cumuliform cloud without dominant or discrete turrets. For reasonable particle concentrations and liquid water contents near-millimeter particle sizes had to be present to explain the low reflectivities measured there. Presumably graupel growth was initiated in the upper parts of this peripheral weak-echo region with turbulent diffusion mixing particles throughout the region.