The Development of a Convective Classification Scheme for Mesoscale Snow Events

Mesoscale snow events (MSEs) are short-lived snowfalls that occur on a spatial scale of less than 2000 KM and are common in the winter across the United States. These vary from high-impact systems like snow squalls or "snow bursts" to low-impact isolated snow showers. Past research has mainly focused on high-impact snow squalls or similar events, with little attention paid to the weaker events. The main goal of this project is to investigate the full range of MSE, including these weaker systems, and create a classification system to measure storm strength. To address this, 901 isolated MSEs were identified over the north-central United States during the cool season (October through April) between 2012 and 2022. Hourly regional radar mosaics were used to identify mesoscale precipitation systems distinct from large extratropical systems, and surface temperature and precipitation observations for each potential case were reviewed to verify that snow was the dominant precipitation type.

Three variables were used to categorize these isolated MSE cases into strongly convective, weakly convective, or non-convective strengths. Radar reflectivity factor (dBz) and reflectivity gradient (dBz/KM) were used to characterize precipitation intensity, and most unstable convective available potential energy (MU CAPE) was used to identify environments supportive of upright convection. Events were classified using the 99th percentile values to represent a peak intensity, as maximum values were sometimes contaminated by noise in the radar data. A decision tree was created to classify the isolated MSE cases by the three variables into the strength categories.

After classification, 283 cases were classified as strongly convective, 263 as weakly convective, and 333 as non-convective. Twenty-two cases could not be classified due to reflectivity, reflectivity gradient, and MU CAPE failing to distinguish between the intensity categories unambiguously. The nearly even distribution of categorized cases suggests that the strongest systems, which have been the focus of past work, likely represented about a third of MSEs overall. This has implications when considering the climatology of these events. For example, the strongly convective cases (including the previously studied snow squalls) had a peak frequency in March. MSE cases follow a more gradual seasonal pattern, with a slow increase in events until December and January and a slow decrease until winter's end. MSEs also show a peak in storm initiation in the late afternoon and evening hours, suggesting that daytime heating may drive some events. However, there is also a slight peak in activity in the early morning, indicating these are not strictly diurnally driven. This increase in cases in the early morning is mainly due to non and weakly convective cases.

This classification system, with its potential to improve operational warning and decision-support practices surrounding MSEs, will be helpful in future MSE research by bridging the gap between strong and weak systems. By capturing a complete range of MSE intensity, this classification will facilitate more detailed comparisons between types of MSEs, leading to a better understanding of the nuanced environmental and dynamical differences between high- and low-impact events. The classification system may prove valuable for improving operational warning and decision-support practices surrounding MSEs as setting definitions for the varying storm strengths may allow for automated identification of the strongest storms in real-time observations or model output.