River ice processes are among the most important subjects of study for hydrotechnical engineers in cold regions. This is because extremes of both minimum flow (impacting fish habitat and the concentration and transport of pollutants) and maximum water levels (impacting channel geomorphology and the flooding of human infrastructure) often occur during the ice-affected season. However, there is a dearth of data describing many facets of ice cover formation and evolution because river ice processes are often logistically challenging to measure. Nevertheless, these data are essential for developing conceptual models of river ice processes and the predictive numerical models that are based on them.

This research project has improved our understanding of unsteady (i.e. time-varying) river ice processes in a variety of environments. This includes anchor ice processes in small headwater streams, ice jam release processes in single-channel river reaches, and ice jam evolution processes in multi-channel river systems. The primary variables monitored in these studies were water level and ice condition and each was observed continuously for up to six months. Each environment was studied for multiple seasons, and multiple examples of each process were observed. This resulted in an unprecedentedly complete picture of each process, and allowed for the development of new conceptual models of these river ice processes.

This thesis presents several key new results. This work confirms that for these streams, thermal processes are an important control on anchor ice release, and that a linear heat transfer approach can be used to predict anchor ice release. It presents several fundamental observations of anchor ice processes, such as variations in ice accumulation morphology, event duration, effect on water level, modes of incorporation into seasonal ice cover, modes of release, and growth rates.

This thesis provides one of the most complete pictures of anchor ice processes ever compiled.This thesis presents the first-ever series of simultaneous observations of the water waves and ice runs that emanate from an ice jam release. It shows how the water waves and ice runs advanced downstream together and then separated due to differing celerities. These observations were taken over a channel distance longer than 10,000 undisturbed flow depths, much longer than can be practically accommodated in laboratory physical models. These data provide important validation data for numerical models of ice run and water wave propagation.

This work also qualitatively and quantitatively describes how ice and water moved through a multi-channel river reach and described the underlying mechanisms of ice jam movement at dividing channel junctions. These descriptions allowed for the development of a new conceptual model that describes how unsteady flow conditions and ice cover momentum are particularly important in multi-channel environments. In addition, this thesis presents a new mode of ice jam release, whereby the release is caused by a water wave that emanates from melting and creeping consolidation of the ice jam itself.