A systematic evaluation of criteria for river ice breakup initiation using River1D model and field data

Mechanical breakup of river ice cover is often associated with ice runs, ice jams, and fast-rising water levels. It not only poses great flood threats to riverside communities, but also has great implications to environment and river ecosystems. Our current ability to model the onset of mechanical breakup is limited. Existing river ice models either require the user to manually specify the breakup times and locations or they employ empirical criteria to simulate ice cover breakup. This paper presents a systematic evaluation of existing empirical and semi-empirical/physics-based breakup criteria using the University of Alberta's River1D model. Each criterion was evaluated according to its ability to capture the breaking front propagation observed on the Athabasca River and Peace River during three breakup seasons, as well as the transferability of the calibrated parameter(s). The storage release from broken ice cover has been postulated to lead to the formation of a non-attenuating, i.e. self-sustaining wave, which in turn maintains fast ice cover breaking over very long distance. This study shows that the characteristics of self-sustaining wave under natural river conditions are greatly affected by the varying resistance to ice breaking along a river, which is different than previous findings based on numerical studies in prismatic channels. •River ice breakup initiation criteria were systematically evaluated with field events.•Calibrated parameters of empirical criteria appear to be event and site specific.•Boundary constraint criterion is promising require better data of river morphology.•Characteristics of self-sustaining wave under natural river conditions were examined.