| Abstract | The St. Lawrence River is an important trade corridor with significant vessel traffic and hundreds of infrastructure assets, such as ports, docks, dams, hydroelectric facilities and locks. The safe and effective operation of existing seaway assets, and the design of new infrastructure, depend on many environmental factors – one of which is ice. An ice cover forms seasonally along most of the marine corridor from the Great Lakes to the Gulf of St. Lawrence. River ice can cause issues (e.g., ice jam-induced flooding, frazil ice clogging of water intakes of power generation stations and potable water plants). One way to mitigate these issues is to promote the formation of a stable ice cover by decreasing river flow. Once a stable ice cover has formed, flow can be increased with less risk of breaking the ice cover.
When planning future operations and construction on the St. Lawrence River, the expected effects of climate change must be considered. Over the past 70 years, the average air temperature has increased in all regions of Canada, and further warming is predicted in all seasons. Climate change is also projected to increase the intensity of some extreme weather events. Changes in ice conditions in the future could affect, favourably or adversely, the operations of infrastructure systems that depend on a stable ice cover.
Two target sites were addressed in this project: Canal Beauharnois just upstream of Montreal, and Lake St. Lawrence in the international section of the river. The objectives were to examine historical trends in the river ice and quantify plausible future ice scenarios, including future season length and number of freeze-up/break-up cycles. The project began with a review of river ice fundamentals, of previously existing methodologies to investigate river ice, and of available data relevant to river ice modelling. No existing methods or models were identified that could be used to meet the project objectives. Next, the historical and future border ice thicknesses were assessed, as a first step to understanding the influence of climate on the ice. That method only predicts the growth of ice along shorelines – not the freeze-up and break-up dates at the target sites. To investigate these parameters, a site-specific ice cover prediction model was then developed and validated against historical data, followed by its extension to future climate scenarios.
The available historical data on border ice thickness and ice cover presence point to a declining trend: overall, ice is getting thinner, and ice seasons are starting later and ending earlier. Modelling of future ice cover and ice thickness scenarios was performed using two sets of air temperature input data (RCP 4.5 and 8.5), over two periods (2040-2060 and 2080-2100). The future simulations are generally consistent with trends observed in historical data. Ice thickness modelling indicated that some border ice will likely still form at the end of this century, but at reduced thicknesses; inter-annual variability in maximum ice thickness will continue or increase. Ice cover modelling showed continuing trends of a reduction of ice season length, mostly due to progressively later season starts. Under both climate scenarios, it is predicted that Canal Beauharnois will be ice-free in some years during the period 2040-2060, and in most years by 2080. For the seasons in the period 2040-2060 for which the Canal is expected to have ice, the model predicts a modest increase in the average number of freeze-up/break-up cycles compared to observations of historical seasons. For Lake St. Lawrence, ice is predicted to be present in less than half of the years from 2040-2060, and there may be no ice cover at all after the mid-2080s. There is no evidence for increased cycles at Lake St. Lawrence for the year 2040 onward.
Further work could be done to test the ice cover model against additional historical ice data, and with different future climate inputs. The model could also be tested as a forecasting tool to support the daily operations of the International Lake Ontario - Saint Lawrence River Board (ILOSLRB). The present model utilized only air temperature; other factors such as water temperature, precipitation and flow could be incorporated. Case studies could be investigated with a physics-based model.
As part of this project, the University of Waterloo carried out a preliminary investigation on river ice classification from satellite imagery. The ice classification method was successfully used to create a timeseries of percentage of ice cover on Canal Beauharnois from 2016 to 2021. Additional work could be done to generate historical ice coverage data for additional years, and the method could be improved by integrating environmental parameters. It would also be useful to explore the role of satellite data in enhancements of the NRC’s ice cover model, as well as to support winter operations of the ILOSLRB. |
|---|
