Ice jam modelling and field data collection for flood forecasting in the Saint John River, Canada

Abstract

AbstractIce jams cause major flooding and severe damages to communities and infrastructure along the Saint John River. There is a growing need to develop capability in forecasting and analysing ice‐jam‐related flood events. This capability is also essential in anticipating the potential for increased ice jam damages as a result of a changing climate. The well‐known and user‐friendly Hydrologic Engineering Centre's River Analysis System (HEC‐RAS) model, which can simulate ice jam configuration under steady‐state conditions, has been calibrated for operational application along the international Saint John River from Dickey, Maine, USA, to Grand Falls, New Brunswick. Examples of model results are presented, and the modelling experience gained to date is outlined. Surprisingly, model output begins to deteriorate when the spacing of cross sections is less than a site‐specific threshold. Once calibrated, the model generates good results, but model parameters change from site to site. Inconsistencies relative to current ice jam understanding and potential improvements are identified. A dangerous consequence of jamming is the sharp wave (jave for short) that is generated upon ice jam release. At present, dynamic aspects of breakup can best be assessed by measurement. Specially designed portable loggers were deployed in 2009 to capture various javes as well as the spring flood that typically arrives after ice clearance. The results enabled a comprehensive comparison between the characteristics of javes and runoff waves, whereas the runoff data were also used to test the unsteady flow routine of HEC‐RAS, which can be used to develop forecasts for runoff floods. Copyright © 2012 Her Majesty the Queen in right of Canada. Published by John Wiley & Sons, Ltd.