Review of prediction methods for breakup ice jams

Breakup ice jams often occur suddenly, with little warning. Severe flooding or ice-related damage can result from rapid rises in upstream water levels associated with breakup ice jams. Breakup jam prediction methods that can be used to increase response time are desirable to minimize flood damage, including potential loss of life. A variety of hydrologic and hydraulic models exist to predict open-water flooding, whether resulting from rainfall, snowmelt, or catastrophic events such as dam breaches. However, breakup ice jams result from a complex series of physical processes that cannot currently be described with analytical or deterministic models, hindering the development of prediction methods. Those which do exist are highly site specific and range from simple empirical models to an artificial intelligence formulation. To date, no one model exhibits a clear advantage over the others. This paper provides examples of existing breakup ice jam prediction methods and discusses their potential advantages and disadvantages.