Canada Research Chair in Cold Regions Hydrology

Canada’s freshwater resource is considered one of the most abundant in the world. However, there are serious concerns about the future state of northern surface freshwater, Canada’s largest freshwater reserve, because of its vulnerability to climatic warming and human activities. Laurier’s Chair in Cold Regions Hydrology aims at strengthening Canada’s capacity to sustainably manage its northern water resources in face of new challenges and uncertainties from climate warming and unprecedented human disturbance. The Chair’s research team is undertaking fundamental research on processes and parameterisations in Canada’s cold regions. This research is being accomplished through specialised observations in a set of research basins where the key components of the natural water and energy cycles and associated flow and storage processes can be observed and used to develop and test mass and energy flow concepts and numerical descriptions. New knowledge and predictive tools are disseminated through leading international journals and conferences and through interactive training workshops with user communities and stake-holders. The Chair’s research and outreach continue to inform several federal and territorial resource policy initiatives. The Chair also nurtures a rich pedagogical environment for high-quality student training.

Laurier’s Chair in Cold Regions Hydrology participates in the Laurier-GNWT Partnership Agreement, IP3 Research Network, (funded by the CFCAS), International Polar Year (IPY), Natural Sciences and Engineering Research Council (NSERC) Strategic Projects programme, and other national and international projects. The major field study sites include Wolf Creek (Yukon Territory), Scotty Creek (Northwest Territories), Havikpak Creek and Trail Valley Creek near Inuvik (Northwest Territories), with secondary sites in northern Ontario and northern Alberta. New and ongoing and research projects are described below. [numbers in square brackets refer to the Refereed Articles section]. 

Permafrost thaw and Ecosystem change

Global climate warming and its environmental consequences are unlikely to proceed in an easily predictable fashion due to complex responses and feedbacks. Permafrost thaw is one of the most important and dramatic manifestations of climate warming in Canada, and is strongly influenced by feedback processes. It also has the potential to alter other key aspects of ecosystems such as runoff and snowcover, forest composition, biodiversity and habitat for keystone species, surface-atmosphere interactions including greenhouse gas fluxes, forest fire regimes and the quantity and quality flows to downstream ecosystems and the Arctic Ocean. While permafrost and ecosystem responses to warming occur in varying degrees throughout the North, the discontinuous permafrost zone of the subarctic is where the most dramatic permafrost thaw and landscape transformations are currently observed. Forecasted dramatic changes in temperature and moisture are expected to affect the processes governing the release of carbon dioxide and methane from the vast stores of carbon in northern peatlands. There are strong indications that the hydrology and hydrochemistry of the subarctic are changing as a result of permafrost thaw, yet little is known about the interactions and feedbacks among climate, water, biogeochemistry, and ecology of subarctic ecosystems. As a result, the ecosystem consequences of warming cannot be predicted with confidence. The implications of these feedbacks are not only prevalent in the sub-arctic but also to downstream aquatic and terrestrial ecosystems and the Arctic Ocean, downwind as weather systems to the settled regions of southern Canada, and to global climate systems through changes to greenhouse gas fluxes, forest fire regimes and ground surface albedo.

At Scotty Creek, NWT, aerial image analysis of a 2 km2 area indicates that permafrost decreased by ~32% between 1947 and 2008 and by ~12% between 2000 and 2008 [54], as evidenced by the expansion and merger of bogs and disappearance of plateaus. New methods of detecting permafrost degradation were developed that use a combination of aerial photographs, satellite images and LiDAR [45]. Given the correlation between landcover and basin runoff noted above, permafrost loss (i.e. transformation of plateaus to bogs or fens) has the potential to affect the volume and timing of basin runoff [27]. For example, the runoff from all gauged basins in the lower Liard River valley has steadily increased since the mid-1990s [40]. The influence of permafrost loss on basin runoff also has important implications to linear disturbances (e.g. seismic lines, highways, and gas pipelines), since the removal of their tree cover produces linear, permafrost-free corridors that conduct water in a similar way to the channel fens [40]. The density of these features at Scotty Creek is about 6 times greater than the natural drainage density.

The measurement of permafrost thaw combined with field studies on the hydraulic and thermal properties of peat led to the Soil Moisture Feedback concept of ground thaw, which accounts for the spatial patterns of both seasonal thaw and permafrost thaw. This new concept provides the basis for numerical model development of ground thaw [e.g. 51]. The conceptual model identifies thinning of the canopy by natural or anthropogenic processes as leading to increased radiation loading at the ground surface [40]. This leads to a local thaw depression toward which subsurface water drains. As water accumulates, a local area of elevated soil moisture content and, therefore, increased bulk thermal conductivity above the thaw depression. More thermal energy is transferred downward from the ground surface, further deepening the thaw depression and further drawing local subsurface drainage from surrounding areas. As this process continues, the remaining trees above the frost table depression may be unable to survive due to water-logging, in which case the canopy is thinned further or removed, and radiation loading at the ground surface above the thaw depression increases. This sequence of events may lead to a local removal of permafrost. Field observations and SVAT (Soil-Vegetation-Atmosphere-Transfer) modelling runs indicate that this conceptual model also applies to permafrost thaw following canopy removal by human disturbance, such as along seismic lines.

Collaborators include Drs. J. Baltzer (Laurier), L. Chasmer (Laurier), B. Branfireun (Western Ontario), J. Kanigan (INAC) and S. Kokelj (INAC)

Hydrological functioning of wetland-dominated terrains with discontinuous permafrost

This research examines water flow and storage processes in the low relief, wetland dominated basins of discontinuous permafrost in subarctic and boreal regions [18], using hydrometric [11], geochemical [13] and numerical modelling [30, 35] techniques. This study has generated a new understanding of the hydrological functioning of the wetland-dominated discontinuous permafrost terrains that predominate throughout much of the southern extent of permafrost. This is a critical first step toward developing numerical simulations including the predictive tools needed by water scientists and requested by water managers in this region.

From detailed field measurements and high resolution satellite imagery, a hydrologically-based landcover classification was derived [11], as was a new conceptual model of basin runoff. The latter recognises three major ground-cover types, each with a specific hydrological function. Permafrost plateaus function as runoff generators, given their relatively deep snowpack, limited capacity to store water and a hydraulic gradient directed toward the adjacent bogs and fens [27, 40]. Water draining from plateaus enters the permafrost-free wetlands: channel fens, isolated and connected bogs. Water entering channel fens is conveyed downstream toward the basin outlet, as the primary function of these features is lateral flow conveyance along their broad, hydraulically-rough channels. Since isolated bogs are surrounded by the raised permafrost of the plateaus, water entering them is stored until removed by evaporation or groundwater recharge [35]. During most of the year, water entering connected bogs (i.e. those with a surface connection to channel fens) is stored as in the case of the isolated bogs; however, surface runoff into channel fens occurs during the annual snowmelt period and in response to large, late-summer rain events [30]. Chemical and isotopic analysis of surface and subsurface waters in the Scotty Creek, NWT basin indicate that less than half of the water discharged during the spring freshet is snowmelt water, suggesting a large amount of water is stored during winter in the connected bogs and channel fens [13]. Annual runoff of four wetland-dominated basins with discontinuous permafrost in the lower Liard River valley, was positively correlated with the percentage of the basin covered by channel fens, and negatively correlated with the percentage covered by bogs [17].

Collaborators include Drs. M. Hayashi (University of Calgary), J. Craig (Waterloo), G. Vanderkamp (National Water Research Institute).

Runoff process studies and model development

Detailed process studies have improved the understanding and ability to predict runoff in cold regions. The following summarises key advancements of this project.

a) Mass flow studies: Most runoff from organic-covered, permafrost hillslopes occurs as subsurface, saturated flow above an impermeable frost table [11,16,18,19,20,21]. The saturated hydraulic conductivity (Ks) decreases by 2-3 orders of magnitude over a depth range of ~0.4 m [12], and the frost table (i.e. zero degree isotherm) is relatively impermeable to water [16]. Therefore the degree of soil thaw determines the rate of lateral subsurface drainage. The physical basis for the large decrease in Ks with depth was examined from image analysis combined with flow measurements. For example, the mean pore diameter, D and a coefficient C of the relation between friction factor, f and Reynolds Number, NR, (i.e., f=C/NR) were found to control hydraulic conductivity [9]. Since C=2D2/k, where k is the soil permeability, the relation between C and d was incorporated into the Cold Regions Hydrological Model (CRHM) to approximate the variation in permeability with depth [25]. Subsequent digital image analysis of resin-impregnated peat samples of known Ks [31] showed that Ks is controlled by the pore hydraulic radius R. The strong dependence of Ks on R implies that peat soils subjected to similar degrees of decomposition and compaction have similar Ks values. Similarly, the Ks profiles of widely-occurring organic-covered permafrost terrain types (i.e. tundra, taiga, boreal peatland) were all found to exhibit a uniformly high and low Ks in the upper and lower regions of the profile respectively, separated by a transition zone in which Ks decreases abruptly with depth. Therefore, a new algorithm has been incorporated into CRHM that describes Ks as a continuous function of depth below the ground surface.

b) Heat flow studies: Three methods (gradient, calorimetric, and flux plate) were used to estimate the ground heat flux, all of which provided consistent results, indicating that theoretical equations give good results in peat [24]. A related study [33] evaluated the performance of 5 algorithms at 4 cold regions site types, and identified appropriate model approaches for a wide variety of cold regions conditions. Spatial variation in frost table depth was shown to be controlled mainly by the spatial pattern of soil moisture [24, 35] and tree canopy density [41, 43].

c) Hillslope runoff modelling: Mass and thermal algorithms were developed and coupled [25], to form a physically-based numerical model for estimating the volume and timing of subsurface runoff from organic-covered permafrost hillslopes. The resulting model has been tested on a variety of cold regions terrains [15, 16]. Another algorithm SFASH (Simple Fill And Spill Hydrology) was created to simulate the affects of differential soil thaw on spatial and temporal variations of subsurface drainage patterns [35]. Current studies are focussed on further development and testing of coupled mass-energy flow algorithms within CRHM and the distributed hydrological model GEOtop [51].

d) Flow path studies: Geochemical studies [16, 19, 21] were used to delineate hillslope and basin flowpaths, and to define their relative importance over time and space. They also produced new knowledge on the transformations of major ion concentrations and loads of snowmelt and rainfall input draining toward to the basin outlet, and on the cumulative influence of runoff processes and pathways on the fate of chemicals deposited to snow and ground surfaces.

Collaborators include Drs. S.K. Carey (McMaster), S. Endrizzi (Zurich), M. Hayashi (Calgary), J.W. Pomeroy (Saskatchewan), E.D. Soulis (Waterloo), and P. Marsh (NWRI).

New concepts of runoff generation in cold regions

While traditional theories of runoff generation may apply to flat, homogeneous tundra, any degree of topographic complexity introduces stark variations in radiation and aerodynamic energy, which in turn affects the accumulation and melt of snow, active layer thaw, soil moisture, evapotranspiration, and therefore, the volume and timing of runoff. This close coupling of soil thaw and drainage led to a new, energy-based concept of runoff generation from organic-covered permafrost that emphasises the frost table topography [34]. Given the depth-dependancy of Ks, this topography defines the spatial distributions of Ks and hydraulic gradients, and reveals preferential pathways and local drainage directions [34]. The ground thaw depth is strongly correlated with cumulative ground heat flux [20], and thus the spatial pattern of thaw mirrors the pattern of snowcover removal [38]. Since organic-covered permafrost terrains support similar peat-forming species, the hydraulic [23, 31] and thermal [24] properties of their soils are similar, which adds to the transferability of the new concept. The soil thaw rate over hillslopes can be accurately simulated from the areal depletion of the snowcover [38]. By simulating snowmelt and soil thaw spatially on a >25 000 m2 hillslope at Granger Creek using the TONE model [38], information unobtainable at the point and plot scales, such as the spatial and temporal patterns of flow-zone Ks, tortuosity of flow pathways and slope-integrated drainage rates was obtained [38]. The transit times of major runoff pathways and their seasonal variations were also defined [21]. Current research is focussed on developing and testing new coupled energy and mass flow algorithms within the GEOtop distributed hydrological modelling framework [51], with field studies at Trail Valley Creek and Wolf Creek.

Collaborators include, S.K. Carey (McMaster), J. Craig (Waterloo), Drs. S. Endrizzi (Zurich), P. Marsh (NWRI), E.D. Soulis (Waterloo).

Laboratory studies on organic soil properties

In classical soil physics, key microscopic soil properties have long been assumed to be un-measurable by direct means and are, therefore, indirectly measured or derived by theoretical formulae containing a coefficient accounting for the affect of pore shape on flow. 3D images from CT scans were used to directly measure pore geometric properties known to control the theoretical shape factor [37, 39], and to demonstrate how such properties vary with changes in soil water pressure. Shape factors derived from direct measurement of these properties were used to compute unsaturated hydraulic conductivity (Ku) from a variety of classical equations. The computed Ku values closely matched values measured with a permeameter for different moisture contents [37]. The 3D analyses also showed that the pore size distribution is dominated by a single large pore-space, whose volume and surface area is 3-orders of magnitude larger than the next largest pore, and >99% of the total inter-particle pore volume [36]. Since the total porosity of peat is approximately 2-3 times larger than in mineral soils, the single large pore implies a more effective flowpath due to a greater degree of hydraulic connectivity among different soil regions, lower tortuosity of connected pores, and lower resistance to flow due to larger pore spaces. The single large pore was a more effective flowpath at low pressure heads compared to smaller pores [39].

To investigate key hydrological and thermal processes related to permafrost thaw under a controlled environment, four large (0.6 m diameter, 0.75 m depth) undisturbed cores of peat with surface vegetation (0.25 m high) were sampled from Scotty Creek in August 2007 and transported to the Biotron facility at the University of Western Ontario. There they were instrumented with soil moisture and temperature sensors, water table monitoring wells, and various above- and below-ground energy flux sensors. These experiments include simulations of soil thawing, freezing and moisture redistribution processes [48, 53]. On-going experiments are focussed on evaluating the sensitivity of these and other processes to changes in soil temperature, snow-cover thickness, ground heat flux, tree- and shrub-canopy densities, wind regimes and other site factors.

Collaborators include Drs. S.K. Carey (McMaster), M. Hayashi (U. Calgary), J.S. Price (Waterloo) and R. Schincariol (University of Western Ontario).