Towards a River Breakup Forecast System

• Main Authors: C. Mahabir, F. Hicks, C. Robichaud, A. Robinson Fayek
• Other Authors: The Pennsylvania State University CiteSeerX Archives
• Format: Text
• Language: English
• Subjects: Athabasca River; Fort McMurray
• Online Access: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.550.3574; http://cripe.civil.ualberta.ca/Downloads/12th_Workshop/Mahabir-et-al-2003.pdf

ice jam events have frequently produced the most extreme and dangerous flood events on record, resulting in millions of dollars in associated damages. However, our ability to forecast such events remains quite limited. A good example of this is the Athabasca River at Fort McMurray, Alberta, where severe ice jam events have been documented for over 100 years, and where breakup has been monitored intensively for the past 25 years. Despite these efforts, no reliable flood forecast model is yet available for this site. Here, the use of Regression techniques and Fuzzy Expert Systems are explored to examine their potential for developing short and long lead time ice jam risk forecasts for this site. Model performance is assessed based on data obtained during breakup in 2003. While, both models performed well in 2003, issues regarding model sensitivity and parameter uncertainty require further investigation.