Investigation and modelling of anchor ice formation and release processes at Clark Lake
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Anchor ice impacts are commonly observed at the outlet of Clark Lake, which is in northern Manitoba upstream of the Keeyask Generating Station (GS) operated by Manitoba Hydro. Manitoba Hydro experiences challenges in accurately forecasting the inflow reaching the Keeyask GS during winter months due to the variable formation and release of anchor ice at the outlet of Clark Lake, which can result in potential revenue losses. This work improves the understanding of anchor ice and other ice processes at this location by performing historical analysis and conducting an on-site field monitoring program, and provides an empirical model to predict the timing of anchor ice intended for Manitoba Hydro to use in operation.
There were 88 definite ice events identified over the 2003/04-2022/23 winter seasons at the outlet of Clark Lake, with 81 being anchor ice. Analysis of the ice events identified proved that different types of ice events occurred, and they were further divided into categories based on their dominant ice type, anchor ice duration type, timing, and release type. There were general trends found between different ice event categories related to their timing, size, and duration. The energy budget trends for the largest formation and release events in both the field monitoring program winters and the historical winters were investigated in more detail. It was found that the energy budget typically decreased surrounding major formation events and increased surrounding major release events, with the sensible heat flux being the dominant heat flux.
Dynamic threshold models were developed to predict anchor ice formation and release events, independently, using the change in the sensible and evaporative heat flux as the predictor variables. The final threshold models developed had a 73% and 72% weekly accuracy, for formation and release, respectively, with a higher accuracy for major events. Future work should focus on developing a model to predict the magnitude and duration of the ice impacts in addition to timing.