The existence of anchor ice in supercooled water can have a profound impact on the management of water resource infrastructures in cold regions. For example, it can raise a tailrace water level and cause significant losses in generation revenue. So far, there have been limited studies on anchor ice, therefore, many problems still exist and much more study is needed. In the present research, experimental and mathematical studies of anchor ice were carried out. Experiments were conducted in a counter-rotating flume, located in a cold room at the University of Manitoba. The experiments were mainly focused on anchor ice evolution around rocks and on gravel beds under different hydro-meteorological conditions. The results are compared to a mathematical model developed herein and some important parameters such as anchor ice porosity and frazil ice deposition coefficient are examined. The growth process of anchor ice was monitored by two CCD cameras. A digital processing program was developed to analyze anchor ice images and determine the growth rate of anchor ice. In addition, anchor ice density, an important factor when studying anchor ice, was estimated and the effect of air temperature, Froude number and Reynolds number is explored. By analyzing torque load signals from the counter-rotating flume, the variation of bed roughness with the growth of anchor ice is elucidated. The deposition coefficient of anchor ice growth was also determined from the experiments. A mathematical model was developed based on a two-stage method to simulate the process of frazil ice transportation and deposition. Both frazil ice attachment and heat transfer between the supercooled water and ice crystals are considered in the model. Four governing equations related to the distribution of velocity and frazil ice transportation and deposition inside and outside the roughness layers were built. A fourth-order Runge- Kutta numerical method was used and programmed in Matlab to solve the governing equations. The growth rate of anchor ice under different hydro-meteorological conditions can be simulated by this numerical model. The proposed experimental and mathematical studies of anchor ice are presented intuitively in this paper and the results from this study contribute to a better understanding of the anchor ice growth mechanism. This study will help to develop better management strategies to mitigate ice related complications associated with hydroelectric generating stations and other hydraulic structures in cold regions.

The existence of anchor ice in supercooled water can have a profound impact on the management of water resource infrastructures in cold regions. For example, it can raise a tailrace water level and cause significant losses in generation revenue. So far, there have been limited studies on anchor ice, therefore, many problems still exist and much more study is needed. In the present research, experimental and mathematical studies of anchor ice were carried out. Experiments were conducted in a counter-rotating flume, located in a cold room at the University of Manitoba. The experiments were mainly focused on anchor ice evolution around rocks and on gravel beds under different hydro-meteorological conditions. The results are compared to a mathematical model developed herein and some important parameters such as anchor ice porosity and frazil ice deposition coefficient are examined. The growth process of anchor ice was monitored by two CCD cameras. A digital processing program was developed to analyze anchor ice images and determine the growth rate of anchor ice. In addition, anchor ice density, an important factor when studying anchor ice, was estimated and the effect of air temperature, Froude number and Reynolds number is explored. By analyzing torque load signals from the counter-rotating flume, the variation of bed roughness with the growth of anchor ice is elucidated. The deposition coefficient of anchor ice growth was also determined from the experiments. A mathematical model was developed based on a two-stage method to simulate the process of frazil ice transportation and deposition. Both frazil ice attachment and heat transfer between the supercooled water and ice crystals are considered in the model. Four governing equations related to the distribution of velocity and frazil ice transportation and deposition inside and outside the roughness layers were built. A fourth-order Runge- Kutta numerical method was used and programmed in Matlab to solve the governing equations. The growth rate of anchor ice under different hydro-meteorological conditions can be simulated by this numerical model. The proposed experimental and mathematical studies of anchor ice are presented intuitively in this paper and the results from this study contribute to a better understanding of the anchor ice growth mechanism. This study will help to develop better management strategies to mitigate ice related complications associated with hydroelectric generating stations and other hydraulic structures in cold regions.

Streams in cold regions are characterized by unique hydrological processes that control flow regime and water levels. One of the most important processes is the formation, growth and melting of different types of ice in and around the stream channel during winter. River ice controls major hydrologic events such as winter floods with magnitudes and frequencies often greater than those created by open-water conditions. While river management in northern countries has already recognized high risk of ice damages, the focus of the risk assessment has been mostly aimed towards the local economy; the ecological role of river ice has been less acknowledged. Along rivers in boreal Sweden, riparian vegetation has developed specific zonation with height and age of the plants increasing the further away they are from the stream channel. On lower levels the vegetation is often comprised of short-lived plants, such as annuals and biennials whereas more permanent woody vegetation is found at higher levels. This zonation has most often been explained by the resilience of different growth forms to the inundation regimes, such as the spring flood in northern systems. Within this framework, I investigated which factors drive the ice formation and how ice and ice-induced floods affect riparian and in-stream vegetation. A 3-year survey was conducted of ice formation and vegetation along 25 stream reaches and a set of experiments were used to evaluate ice as a disturbance agent. Reaches far away from lake outlets which had a low input of groundwater and a high velocity and stream power were most prone to form anchor ice, but many other factors also influenced ice formation. Streams with anchor ice experienced more frequent flooding of the riparian vegetation during winter. Our findings suggests that ice and winter floods favour diversity and create habitat heterogeneity for riparian species. On a community level, woody plants such as evergreen dwarf shrubs are eliminated when flooded during winter, opening up patches for other species to colonize, creating a dynamic riparian understory community. Significant changes in river ice conditions could develop with projected changes in climate which would have important geomorphologic, ecological and socio-economic impacts. One implication of climate change could be less ice disturbance and consequently a riparian vegetation in cold regions that slowly changes from forb to dwarf-shrub dominated with a subsequent decrease in species richness. Changes in species diversity and abundance of groups of species related to changes in ice formation could potentially cascade into riparian and in-stream processes such as nutrient cycling, litter decomposition and organism dispersal.

Riparian ecosystems are of great importance in the landscape, connecting landscape elements longitudinally and laterally and often encompassing sharp environmental gradients in ecological processes and communities. They are influenced by fluvial disturbances such as flooding, erosion and sediment deposition, which create dynamic and spatially heterogeneous habitats that support a high diversity of species. Riverine ecosystems belong among the world’s most threatened systems. In rivers throughout the world, human alterations to fluvial disturbance regimes have resulted in degraded ecosystems and species loss. For example, in Sweden, watercourses of all sizes have been channelized to facilitate timber floating, but in the last 10–20 years the impacts in some of the affected rivers have been reduced by restoration actions. The objectives of this thesis are to evaluate how riverine ecosystems in general, with specific focus on riparian communities, are affected by (1) restoration of channelized reaches by boulder replacement, (2) ice formation, and (3) restoration of in-stream wood abundance in the stream channel. Objective (1) was assessed by quantifying the retention of plant propagules in channelized and restored stream reaches and by evaluating effects on riparian plant and bryophyte communities in disconnected and re-opened side channels. Retention of plant propagule mimics was highest at low flows and in sites where boulders and large wood had been replaced into the channel. Propagules are however unlikely to establish unless they can be further dispersed during subsequent spring high flows to higher riparian elevations suitable for establishment. Thus, immigration to new suitable sites may occur stepwise. Our study demonstrates that restoration of channel complexity through replacement of boulders and wood can enhance retention of plant propagules, but also highlights the importance of understanding how restoration effects vary with flow. We detected no differences in riparian diversity between re-opened and disconnected side channels, but we did observe significant differences in species composition of both vascular plant and bryophyte communities. Disconnected sites had more floodplain species, whereas restored sites had more species characteristic of upland forest. This suggests that the reopening of side channels resulted in increased water levels, resulting in new riparian zones developing in former upland areas, but that the characteristic floodplain communities have not had time to develop in response to the restored fluvial regime. Objective (2) was approached by evaluating the effect of both natural anchor ice formation and experimentally created ice in the riparian zone. Riparian plant species richness and evenness proved to be higher in plots affected by anchor ice. Plants with their over-wintering organs above the ice sheet suffered from the treatment but the overall species richness increased in ice-treated plots. Objective (3) was evaluated by studying wood recruitment and movement, channel hydraulics, propagule retention and fish abundance in streams restored with large wood. Only one stream experienced reduced velocities after large wood addition. The large size and reduced velocity were probably also the reasons why this stream proved to be the best one in trapping natural, drifting wood. Increased retention and decreased mechanical fragmentation in large wood sites will lead to decreased loss of detritus from the site and therefore higher availability of coarse particulate organic matter which can result in more species rich shredder communities. Our study did not show that the occurrence of large wood had an important role in controlling density or biomass of brown trout.

Water resource infrastructure in cold regions of the world can be significantly impacted by the existence of river ice. Major engineering concerns related to river ice include ice jam flooding, the design and operation of hydropower facilities and other hydraulic structures, water supplies, as well as ecological, environmental, and morphological effects. The use of numerical simulation models has been identified as one of the most efficient means by which river ice processes can be studied and the effects of river ice be evaluated. The continued advancement of these simulation models will help to develop new theories and evaluate potential mitigation alternatives for these ice issues. In this thesis, a literature review of existing river ice numerical models, of anchor ice formation and modeling studies, and of aufeis formation and modeling studies is conducted. A high level summary of the two-dimensional CRISSP numerical model is presented as well as the developed freeze-up model with a focus specifically on the anchor ice and aufeis growth processes. This model includes development in the detailed heat transfer calculations, an improved surface ice mass exchange model which includes the rapids entrainment process, and an improved dry bed treatment model along with the expanded anchor ice and aufeis growth model. The developed sub-models are tested in an ideal channel setting as somewhat of a model confirmation. A case study of significant anchor ice and aufeis growth on the Nelson River in northern Manitoba, Canada, will be the primary field test case for the anchor ice and aufeis model. A second case study on the same river will be used to evaluate the surface ice components of the model in a field setting. The results from these cases studies will be used to highlight the capabilities and deficiencies in the numerical model and to identify areas of further research and model development.

Water resource infrastructure in cold regions of the world can be significantly impacted by the existence of river ice. Major engineering concerns related to river ice include ice jam flooding, the design and operation of hydropower facilities and other hydraulic structures, water supplies, as well as ecological, environmental, and morphological effects. The use of numerical simulation models has been identified as one of the most efficient means by which river ice processes can be studied and the effects of river ice be evaluated. The continued advancement of these simulation models will help to develop new theories and evaluate potential mitigation alternatives for these ice issues. In this thesis, a literature review of existing river ice numerical models, of anchor ice formation and modeling studies, and of aufeis formation and modeling studies is conducted. A high level summary of the two-dimensional CRISSP numerical model is presented as well as the developed freeze-up model with a focus specifically on the anchor ice and aufeis growth processes. This model includes development in the detailed heat transfer calculations, an improved surface ice mass exchange model which includes the rapids entrainment process, and an improved dry bed treatment model along with the expanded anchor ice and aufeis growth model. The developed sub-models are tested in an ideal channel setting as somewhat of a model confirmation. A case study of significant anchor ice and aufeis growth on the Nelson River in northern Manitoba, Canada, will be the primary field test case for the anchor ice and aufeis model. A second case study on the same river will be used to evaluate the surface ice components of the model in a field setting. The results from these cases studies will be used to highlight the capabilities and deficiencies in the numerical model and to identify areas of further research and model development.