Modeling Ice Storm Climatology

Abstract

Extreme weather events such as ice storms cause significant damage to life and property. Accurately forecasting ice storms sufficiently in advance to offset their impacts is very challenging because they are driven by atmospheric processes that are complex and not completely defined. Furthermore, such forecasting has to consider the influence of a changing climate on relevant atmospheric variables, but it is difficult to generalise existing expertise in the absence of observed data, making the underlying computational challenge all the more formidable. This paper describes a novel computational framework to model ice storm climatology. The framework is based on an objective identification of ice storm events by key variables derived from vertical profiles of temperature, humidity, and geopotential height (a measure of pressure). Historical ice storm records are used to identify days with synoptic-scale upper air and surface conditions consistent with an ice storm. Sophisticated classification algorithms and feature selection algorithms provide a computational representation of the behavior of the relevant physical climate variables during ice storms. We evaluate the proposed framework using reanalysis data of climate variables and historical ice storm records corresponding to the north eastern USA, demonstrating the effectiveness of the climatology models and providing insights into the relationships between the relevant climate variables.