Sonar detection and measurement of ice in a freezing river II: Observations and results on frazil ice

Results from 2004-2008 BC Hydro SWIPS (Shallow Water Ice Profiling Sonar) monitoring programs on the Peace River are presented and analyzed with respect to their content of information on the properties and behaviour of frazil ice suspended in water column. Strong distinctions were made in analyses applied to data collected, respectively, prior to and following local stabilization of the seasonal ice cover. This choice is shown to reflect both the behaviour and origins of the frazil ice present in these two intervals. Pre-freeze-up frazil tends to coat submerged surfaces, grows rapidly while in suspension and is only rarely and episodically present. It shows time dependences quite different from those associated with frazil detected prior to freeze-up. Frazil under a stationary ice cover is ubiquitous, does not adhere to other materials and appears to closely associated with the ice cover undersurface. The return intensity profiles observed in pre-freeze-up frazil events increased roughly linearly with height in the water column. Profile comparisons with earlier simulation results allowed inferences on particle size distributions and their controlling processes. The bulk of data analysis efforts was devoted to clarifying the time dependences of the more abundant post-stabilization SWIPS intensity data and its underlying connections to changes in the major river, atmosphere and ice cover environmental parameters. The obtained results showed unlagged or weakly negatively-lagged, positive correlations between such intensities and both river water levels and air temperatures (or solar radiation input) on, at least, diurnal and shorter time scales. The water level parameter was the stronger of these two influences and was believed to be representative of a physical link to river flow speed. Lower frequency connections to the physical state of the ice cover were also apparent. Interpretations of these results were offered in terms of a local equilibrium between the suspended frazil population and a dynamic slush layer at the bottom of the ice cover. Initial steps were taken toward developing and testing analysis tools potentially capable of supporting combined analysis and modelling of this equilibrium. A simple interpretative approach was applied to data acquired at a single acoustic frequency based upon combined use of a Rouse Law-based Inverted sediment model of uniformly-sized spherical frazil particles and volume backscattering coefficients measured in post-freeze-up intervals. Results suggested consistency with the basic form of return intensity profiles at most but not all times. For suitable measurement periods, the approach yielded reasonable estimates of the major particle size and concentration parameters. The deficiencies of this approach are discussed and compared with expectations and results from recently successful simultaneous measurements at two acoustic frequencies. Priorities are identified for validation, and refinement of single and multifrequency approaches along with integrated study of the lower ice cover as part of a broader program to understand frazil ice growth in freezing rivers.