Temporal and Spatial Analysis of Freezing Rain over Eastern North America

Freezing rain is one of the least understood and understudied types of precipitation. While responsible for billions of dollars in damage over the last 50 years and posing a significant risk for both urban and forested areas alike, limitations in reporting and analysis continue to persist. This effect is attributed to the relatively short event periods, as well as confusion with similar precipitation types; namely rain and sleet, in reporting such events. Although with a limited temporal resolution of 1 hour, the ERA5 reanalysis freezing rain data has shown consistency when compared to localized weather stations.

This study examines the relationships between freezing rain events and atmospheric conditions and synoptic patterns over the eastern half of the United States and Canada. ERA5 freezing rain data were collected from 1979-2020 and sorted into distinct freezing rain events, wherein events were further sorted and filtered by total size, duration, and location. Synoptic maps were collected for each of these events at their peak time, defined as the hour in the event where the largest area was affected. These maps were then loaded developed into different regimes using Self-Organizing Maps (SOM). These regimes were developed by using multivariate SOM using standardized anomalies of four key meteorological variables (500mb height, surface and 850mb temperature, sea level pressure) to derive the initial synoptic types, followed by k-means clustering for final classification We will present the our final regimes for seven distinct regions across Eastern North America, and discuss the underlying meteorology behind each regime, as well as evaluate differences for each region. This research can be used to further develop and understand freezing rain patterns and develop how these patterns are changing over a changing climate.