The Processes, Patterns and Impacts of Low Flows Across Canada

The Processes, Patterns and Impacts of Low Flows Across Canada Donald H. Burn, James M. Buttle, Daniel Caissie, Greg MacCulloch, Chris Spence and Kerstin Stahl Abstract: his paper provides an overview of low flow characteristics for six regions of Canada: the Arctic; the Mountains; the Prairies; southern Ontario; the Canadian Shield and the Atlantic. Processes that influence low flows are contrasted between the six regions examined. Data from a common analysis period for 51 gauging stations are used to evaluate flow duration curves and to explore the relationship between low flows and drainage area. he results reveal a diversity of processes influencing low flows and illustrate important regional differences in low flow characteristics and the impacts associated with low flows. Résumé : La présente communication offre un survol des caractéristiques de basses eaux pour six régions du Canada : l’Arctique; les Rocheuses; les Prairies; le sud de l’Ontario; le Bouclier canadien et la région de l’Atlantique. Les processus qui influent sur les étiages sont comparés entre les six régions étudiées. Les données obtenues à partir d’une période d’analyse commune pour 51 stations hydrométriques sont utilisées afin d’évaluer les courbes des valeurs classées des débits et d’analyser la relation entre les basses eaux et les bassins hydrographiques. Les résultats révèlent une diversité de processus qui influent sur les basses eaux et illustrent les importantes différences régionales dans les caractéristiques des basses eaux et dans les impacts qui y sont associés. Donald H. Burn1, James M. Buttle2, Daniel Caissie3, Greg MacCulloch4, Chris Spence5 and Kerstin Stahl6 Department of Civil and Environmental Engineering, University of Waterloo, Waterloo, ON N2L 3G1 Department of Geography, Trent University, Peterborough, ON K9J 7B8 3 Fisheries and Oceans, Moncton, NB E1C 9B6 4 Water Survey of Canada, Calgary, AB T2G 4X3 5 Environment Canada, Saskatoon, SK S7N 3H5 6 Department of Geography, University of British Columbia, Vancouver, BC V6T 1Z2 1 2 Submitted June 2007; accepted October 2007. Written comments on this paper will be accepted until December 2008. Canadian Water Resources Journal Revue canadienne des ressources hydriques Vol. 33(2): 107–124 (2008) © 2008 Canadian Water Resources Association 108 Canadian Water Resources Journal/Revue canadienne des ressources hydriques Introduction Low streamflow can have important socio-economic impacts for water supply, reservoir operation, waste load allocation as well as implications for the health of aquatic ecosystems. As an example, the recreational fishery in Atlantic Canada and Québec has a significant economic value, exceeding $200 million annually (Loftus et al., 1995). Many communities depend on the well-being of fisheries resources for sustenance and financial survival. Water use and low flow can have a significant impact on fisheries and other aquatic resources, particularly during the summer period when such events are associated with high water temperatures. It is thus important to understand the processes that affect low flows to manage properly the available water resources and to mitigate the impacts of low streamflow values. he considerable diversity in climate, landscape and land use in Canada results in geographic regions of the country exhibiting distinct differences in low flow characteristics. he intent of this paper is to compare and contrast critical low flow characteristics for major regions of Canada to improve comprehension of the processes that result in low flows and the impacts of low flows on Canada’s economy, culture and environment. Smakhtin (2001), in an excellent review of low flow hydrology, devotes considerable discussion to defining what is meant by low flow. Smakhtin distinguishes between drought events and low flows, the latter representing a regular occurrence when runoff input to a water body is reduced. In this paper, a similar interpretation of low flows is adopted such that low flows represent hydrological conditions expected to occur once or more during a year. During low flow events, streamflow contributions are typically maintained with catchment storage from groundwater, lakes, wetlands, or glaciers. he nature of low flows across Canada and the relative importance of these contributions to low flow over space, time and scale are discussed below. Streamflow data from 51 Water Survey of Canada (WSC) gauging stations are used, with each station representing natural (unregulated) flow conditions. he gauging stations are drawn from six regions of Canada: the Arctic; the Mountains; the Prairies; southern Ontario; the Canadian Shield; and the Atlantic. Stations were selected to be representative of the regions but were not intended to encompass all available stations in a region. he goal was to have a similar number of stations for each of the regions; station availability for the Arctic region was thus the limiting factor. Figure 1 shows the location of the gauging stations. Figure 2 shows plots of runoff depth versus day of the year for a representative river from each region. he plots depict daily hydrographs for a common year (1991) and illustrate the timing of low flows for each region. he next sections describe unique low flow features of each region. his is followed by an analysis and comparison of low flows for the regions. Arctic Region Description of Region Canada’s subarctic and arctic regions are vast, stretching from the Yukon in the west to northern Québec in the east and from Hudson Bay in the south to Ellesmere Island in the north. he physiography is diverse. Wetlands are predominant across the Hudson Bay Lowlands and subarctic plains of the Mackenzie River Valley. hese two areas of low relief are characterized by flat peat bogs, palsa bogs, and ribbed or channel fens. he Canadian Shield is the largest physiographic region in northern Canada and is characterized by shallow soils with numerous outcrops of Precambrian bedrock. Bedrock outcrops are common across the Arctic because the Cryosolic soils that have developed on sparse colluvial and morainal deposits tend to be thin. Glaciers and ice caps predominate in the mountainous regions on the eastern rim of the Archipelago. Precipitation tends to decrease with distance northward while the eastern arctic tends to be wetter than the west at similar latitudes. Mean annual precipitation ranges from less than 100 to 600 mm. he Arctic region experiences a mean annual air temperature less than 0°C and discontinuous permafrost in the southern Arctic that grades to continuous permafrost further north (Ecological Stratification Working Group, 1996). Unique Features of Region from Low Flow Perspective he pervasive cold conditions control the nature of low flows in the region. Sub-freezing temperatures cause precipitation to fall as snow, nullifying landscape © 2008 Canadian Water Resources Association Burn, Buttle, Caissie, MacCulloch, Spence and Stahl 109 Figure 1. Location map of stations used in the analysis. Dashed lines indicate the approximate limits of each region. runoff to streams for up to nine months of the year. he low storage capacity of the thin permafrost active layer allows for fast delivery of water from hillslopes to the stream (Woo, 1983) leaving little water on the landscape to maintain streamflow between rainfall events. Storage within the drainage network increases with watershed size and flows are maintained longer between rainfall events in larger watersheds. he lowest levels of precipitable water in Canada are in the north (Hay, 1971), a condition exaggerated at inland locations. In winter, areas of upland or mountainous terrain with adjacent open water tend to accumulate the most precipitation (Rouse, 1993). hese highlands also intercept atmospheric moisture in the summer, reducing precipitable water in the atmosphere over much of the northern interior (Walsh et al., 1994) and often creating dry conditions. he biophysical differences between the palsa bogs and fens of the subarctic plains and Hudson Bay Lowlands result in different hydrological functions. he presence of permafrost in the bogs and peat plateaus makes these areas natural sources for runoff © 2008 Canadian Water Resources Association 110 Canadian Water Resources Journal/Revue canadienne des ressources hydriques 2a 2b Figure 2. Runoff hydrographs for a representative river from each region showing the timing of the annual low flow event for the year 1991. For graphing purposes, zero flows are plotted as 0.0001 m3/s. © 2008 Canadian Water Resources Association Burn, Buttle, Caissie, MacCulloch, Spence and Stahl to the permafrost free fens. Early in the warm season, runoff is readily generated with ample snowmelt input relative to the small soil storage capacity created by thin active layers. As the summer progresses and the active layer grows, small precipitation amounts cannot offset the storage capacity of the high porosity peat. When it manages to do so, water is conducted much less readily because of the lower hydraulic conductivities encountered at depth (Quinton and Hayashi, 2005) minimizing the contributing area to the stream and significantly reducing flow. he resulting disconnected state of the drainage network remains until precipitation or snowmelt is large enough to fill storage capacity and reconnect the system (Quinton et al., 2003). he impact is that low flow conditions can be extended for some time after precipitation inputs return to normal (Stewart et al., 2002). he cold conditions of the region can create stores of frozen water that can partly alleviate low flows at small scales. In the high arctic, dense late lying snowpacks can provide continuous runoff throughout the summer, maintaining wetland storage and hydrologic connectivity (Woo, 1993). In the subarctic, melting aufeis plays a similar role (Reedyk et al., 1995). Timing of Low Flows In this region, low flows almost always occur during winter (see Figure 2). hey can be at the end of winter, when the streams have been denied landscape runoff for several months, mid-winter when the streams freeze completely, or early winter during freeze up, when increases in hydraulic resistance constrict water moving downstream (Prowse and Carter, 2002). Summer period low flows are caused by low precipitation relative to evapotranspiration, which reduces storage, contributing areas and streamflow. his is particularly prevalent in the dry interior of the subarctic plains. Socio-Economic Impacts of Low Flows Specific to this Region he small population in northern Canada relative to the generally ample water supplies keeps socioeconomic impacts of low flows to a minimum. Some communities experience drinking water quality issues when the levels of lakes used for supply become low. 111 Road networks in the Canadian north are not very extensive. In the summer, the most cost effective way to transport goods to communities along the Mackenzie River is often by barge. Low water levels in certain reaches can be problematic if not forecast well in advance (Kerr, 1997). Mountain Region Description of Region he Pacific and western mountain region of Canada (the “Canadian Cordillera”) stretches from the Pacific Ocean in the west to the Rocky Mountains in the east and encompasses much of the Province of British Columbia and southwestern Alberta. he region is characterized by several mountain chains with elevations over 3000 m.a.s.l., vast plateau areas and deep valleys. Geology and geomorphology are highly variable. he mountains expose volcanic and sedimentary bedrock. Pleistocene glacial, fluvioglacial and glaciolacustrine deposits are found throughout the region; considerable Holocene alluvial deposits exist in the large river valleys. he complex relief of the Canadian Cordillera creates strong gradients and differences in temperature and precipitation. Mean annual precipitation ranges from 250 to over 7000 mm. During the winter months frontal weather systems formed over the Pacific Ocean bring moisture into the region. Orographic effects result in high amounts of precipitation on the western slopes of the mountain ranges while the eastern slopes and the valleys in the lee of the mountains are very dry. Except at low elevations along the coast, winter precipitation falls as snow. Summers are relatively dry throughout the region and considerable climatic moisture deficits develop, particularly in the southern interior where summer temperatures are highest. As a result of the diverse physiographic and climatic characteristics, all large river basins encompass many different biogeoclimatic zones (Krajina, 1969) and even small basins commonly span a large range of elevations and include a diverse landscape of exposed rock, old and young forest, and possibly alpine meadows with small wetlands and lakes. Almost half of the currently gauged watersheds in British Columbia contain a glacier. © 2008 Canadian Water Resources Association 112 Canadian Water Resources Journal/Revue canadienne des ressources hydriques Unique Features of Region from Low Flow Perspective and Timing of Low Flows Climatic influences on low flow timing and influences of basin storage properties on streamflow recession and magnitude are modified by elevation and basin hypsography. Hydrological regimes in the Canadian Cordillera are predominantly nival with winter low flows (see Figure 2). he winter low flow season begins when temperatures drop below freezing and ends when temperatures rise above zero. As winter temperatures generally decrease with elevation, with distance from the coast, and with latitude, the timing of winter low flows is controlled by the location as well as the elevation of the basin. Exceptions are low-elevation pluvial regimes along the coast, where summer is the primary low flow season. During summer, after a prolonged period with a climatic moisture deficit, a secondary low flow period typically develops in basins where snowmelt occurred early in spring and/or over a short period of time (i.e., basins at low elevations or with small elevation differences). In glaciated basins, increased ice melt contributions during the warm season maintain high streamflow levels beyond the snowmelt period. Otherwise, summer streamflow depends on availability and release of surface (lakes and wetlands) and subsurface (groundwater) storage. As mountain soils tend to be shallow and aquifers have limited storage capacity, during a dry year it is not unusual for small catchments in the southern part of the Cordillera to become dry in August or September. In the large valleys across the region, extensive alluvial aquifers can provide late summer baseflow to rivers after being recharged from the river during the spring freshet (Scibek et al., 2007). Environmental and Socio-Economic Impacts of Low Flows While dilution of effluents from industry and mining can be a problem during both winter and summer low flows, environmental and socio-economic impacts of reduced flows and degraded water quality tend to be greatest during the warm and dry season. he salmon fishery is an important industry in British Columbia and an economic and cultural pillar for many First Nations. Low water levels and high stream temperatures threaten the habitat for cold-water species such as salmonids. In dammed rivers, reservoir operation has to be adjusted during low flow periods to meet instream flow targets (mainly for fish but also for other aquatic species), which affects hydropower production during the summer. In the wetter parts of the Cordillera, municipal water supplies as well as irrigation facilities rely on sustained inflow during the dry season. In dry years many communities need to impose water restrictions. Prairie Region Description of Region he Canadian Prairies refer to the interior continental plains of North America stretching from the Rocky Mountains on the west to Lake Winnipegosis on the east. Prairie landscapes are generally considered to be grasslands; however, this study also includes the aspen parkland belt that buffers the true prairie grassland. he combined grasslands and aspen parkland landscapes of the interior plains are almost exclusively within the Nelson River Basin, which drains the land from the Rocky Mountains to Hudson Bay. he combined grasslands and parkland area of the Canadian Prairies corresponds to the most populated areas of Alberta, Saskatchewan and Manitoba due to agricultural opportunities in the area. he extensive agricultural development makes it difficult to distinguish a dividing line between the prairie and parkland regions; however, they are distinct in many ways. Native vegetation in the grasslands is low lying and drought resistant. he Prairies are generally too arid for shrubs and trees (Spalding, 1980); however, trees can be found in protected coulees that incise the plain and offer protection from the harsh winds as well as being collection points for winter snow drift, thus increasing moisture availability. In contrast, the aspen parkland supports the growth of groves of trees interspersed with fescue grass prairie. he geomorphology for both landscapes is predominantly glacial. Most of the Prairie region is covered with clay-rich glacial deposits with a thickness of 100 m or more. he region is dotted with pothole lakes left behind by the retreating continental ice sheet. hese ponds and sloughs create opportunities for a variety of vegetation and wildlife not common © 2008 Canadian Water Resources Association Burn, Buttle, Caissie, MacCulloch, Spence and Stahl on the Prairie. he natural interception and storage potential due to the hummocky terrain, pothole lakes and sloughs creates opportunities for reduced rate of contribution to the main stem rivers. he Prairies accumulate only about 15% of the amount of snow that amasses on the eastern slopes of the Rocky Mountains. Potential evapotranspiration ranges from 600 to 900 mm per year and tends to exceed the average annual precipitation on the Prairies. he Prairies represent one of the driest parts of the country; total precipitation is generally less than 500 mm/year with significant portions of the prairie showing average accumulation of less than 300 mm/ year (Fisheries and Environment Canada, 1978). Unique Aspects of the Region in the Context of Low Flows he Prairies are impacted by intensive water management schemes in support of agriculture. hroughout the region all major streams exhibit some degree of regulation. Discussing low flow streams that rise in the prairie grasslands of Canada presents an interesting paradox in that they exhibit “extreme low flows” on a regular basis. hat is, frequently, natural prairie streams are unable to maintain flow with catchment storage from groundwater, lakes and wetlands. Many streams demonstrate periods of zero flow annually and some for significant portions of time during any one year. Timing of Low Flows For grassland sites, flows generally peak during the last week of March to mid-April and steadily decline until the end of October. Due to the lack of record for the winter months from November to February for many grassland sites, the conclusion that the lowest flows occur during that time is anecdotal. Nevertheless, indications are that there is very little surface water available during these times. For parkland sites, peaks generally occur during the first two weeks of April. As in the case of the grasslands, parkland streams experience steady decline through to the end of December and reach their lowest flows in early January (see Figure 2). 113 Socio-Economic Aspects of Low Flow Conditions he limited availability of water on the grasslands has necessitated the development of water management and delivery infrastructure. Water is captured and stored in reservoirs for distribution to farms and rural municipalities to serve agricultural, municipal and recreation needs of many communities. Good soil, ample sunshine and the availability of water throughout the growing season make irrigated agriculture possible in some areas, mostly in southern Alberta, where under natural conditions surface water resources are extremely limited by late July. he limited availability of surface water resources was well known during the building of the trans-continental railway in the late 1800s. To accommodate settlement, the development of irrigation projects began at the same time, where water was available. Low flows also limit the effluent release opportunities from community and industrial facilities. Southern Ontario Region Description of Region he southern Ontario region is bounded by the Great Lakes on the south and west and the Canadian Shield to the north and is part of the Great LakesSt. Lawrence Lowlands (Bird, 1980). Bedrock for much of this area consists of Paleozoic sedimentary rocks (sandstones, limestones, dolomites and shales) that overlie the Precambrian Shield basement. hese sedimentary formations dip to the southwest, thus creating a series of plains and escarpments that are overlain with surficial materials deposited by numerous glaciations (Bird, 1980). he retreating ice resulted in large plains that have been modified by melt waters and lakes in some areas (Bird, 1980). Drumlins formed by the glaciations are notable in the Peterborough area and south of the Bruce Peninsula as are extensive end moraines, many of which are important groundwater recharge areas. he surficial hydrology is determined by a combination of physiography and land use. Areas with widespread deposits of sand and gravel tend to be well drained with little permanent water while areas with gentle slopes and shallow soils are often poorly drained and contain extensive wetlands. © 2008 Canadian Water Resources Association 114 Canadian Water Resources Journal/Revue canadienne des ressources hydriques Southern Ontario possesses a continental climate that is modified by the presence of the Great Lakes. Annual precipitation varies from 800 mm in the eastern and western portions of the region to 1000 mm in the lee of the Great Lakes, where lake effect snow results in greater accumulation of precipitation. Evaporation increases from north to south and runoff varies from approximately 200 mm in the south to close to 400 mm in the north, with the impact of lake effect snow again very noticeable (Fisheries and Environment Canada, 1978). Unique Aspects of the Region in the Context of Low Flows he southern Ontario region is characterized by a high level of development with the largest population density in Canada. Urban areas, major industries, along with important agricultural areas, create high demands for water implying that low flows are of great concern in this region. here are extensive aquifers in the region and a large portion of the population derives potable water supply from groundwater systems (Kelly et al., 2004). he unique geographic position of this region results in there being few large watersheds and because of its highly developed state even fewer watersheds with natural flow conditions. he Grand River Basin is the largest watershed in the region with a drainage area of 6800 km2, although this watershed is highly developed with only the headwater tributaries representing natural flow conditions. Surficial geology conditions exert a strong control on spatial variations of low flows both between and within drainage basins. hus, basins draining major aquifer complexes, such as the Oak Ridges Moraine, generally produce substantial quantities of baseflow and sustained and relatively large low flows (Hinton, 2005). Conversely, the drainage networks of basins located on till-covered plains that do not receive regional groundwater inputs can often become spatially disconnected due to water losses to evaporation and infiltration through the channel bed during summer droughts, leading to the cessation of streamflow (Todd et al., 2006). Timing of Low Flows he occurrence of low flows in the southern Ontario region is extremely regular. he annual minimum daily low flow tends to occur in late summer or early fall (see Figure 2). For the ten sites examined (see Figure 1) the median date of occurrence of the annual minimum flow, based on a common 11-year analysis period, ranges from August 27 to September 10, a window of only 14 days. he annual minimum daily low flow generally results from low precipitation and high evapotranspiration losses characteristic of late summer and early fall in this region. Annual minimum daily low flows occasionally occur in the winter; however, frequent melt events and rain-on-snow events during the winter tend to maintain baseflow at a higher level than that experienced towards the end of the hot dry summers. Socio-Economic Aspects of Low Flow Conditions he large population of this region (close to 12 million people) results in a large demand for potable water supply. Industrial activity and agriculture are also large users of water resulting in considerable concern with regard to low flow conditions and substantial economic impacts when water shortages occur. As a result of the large level of development in the region, ecological systems are under considerable stress, which can be exacerbated by low flow conditions. he Ontario government recently introduced the Ontario Low Water Response, a plan to manage low water conditions resulting from low precipitation and high temperatures. Canadian Shield Region Description of Region he Canadian Shield consists largely of gneissic Precambrian rocks with small enclaves of metamorphosed volcanic and sedimentary rock (Bird, 1980). he exposed surface of the Shield occupies much of eastern and central Canada, although this discussion is confined to the area south of 60°N since the more-northerly portions of the Shield are © 2008 Canadian Water Resources Association Burn, Buttle, Caissie, MacCulloch, Spence and Stahl 115 discussed in the section on Arctic low flows. he largescale morphology of the Shield consists of basins (e.g., Hudson Bay), domes (e.g., Algonquin dome in southern Ontario) and arches (e.g., the Frontenac axis in southeastern Ontario). Flat uplands occupy the crests of the domes and arches. he outer margins of the Shield range from low-relief portions where the unfaulted Shield dips below younger sedimentary rock (e.g., central Manitoba and south-central Ontario) to deeply dissected and mountainous faulted sections (e.g., the north shore of the estuary and Gulf of St. Lawrence) (Bird, 1980). Glacial removal of preglacial, deep-weathered mantles has left extensive areas of exposed bedrock (Bird, 1980). he surficial geology is relatively thin (<1 m thick) till blankets and till veneers with extensive areas of lacustrine silts and clays, and major glacial and glaciofluvial deposits. hese may be several meters thick and can act as major local aquifer systems. hin, well-drained stony soils predominate, interspersed with pockets of poorly drained but deeper silt/clay and organic soils. Luvisols have developed on clay-rich tills and lacustrine deposits (e.g., the Clay Belt in northeastern Ontario and north-western Quebec). Northern parts of the Shield in Manitoba and Quebec consist largely of rockland and cryosols (Clayton et al., 1977). he Shield possesses a continental climate, with long, cold winters and short warm summers. Annual precipitation (P) decreases from southeast (>1200 mm below Quebec City) to northwest (∼400 mm in northern Saskatchewan) and mean annual temperatures increase from north to south (Fisheries and Environment Canada, 1978). Mean annual potential evaporation decreases from south (∼700 mm in southern Ontario) to north (<200 mm in Ungava) resulting in an increase in P – PET (potential evapotranspiration) in a roughly west to east direction across the Shield; almost all the Shield south of 60°N has an annual water surplus available for streamflow. lakes) in the basins. In the south, water availability to sustain low flows is controlled by the balance between precipitation and evapotranspiration. hese controls on water availability are superimposed on a landscape that is relatively restricted in its capability to deliver groundwater to streams and thus sustain baseflow. Groundwater in the Shield’s bedrock areas is largely confined to faults and fractures (Fisheries and Environment Canada, 1978) and bedrock porosity is low even when it is fractured (Steedman et al., 2004). Groundwater systems are generally shallow and localized, and substantial groundwater inputs to Shield surface waters usually occur only in areas with significant surficial deposits of sand and gravel (Steedman et al., 2004). he influence of a basin’s morphological properties on its low flows depends on basin scale. Steedman et al. (2004) suggest that perennial streams develop on the Shield where the drainage area exceeds ∼25 ha, although records from the Dorset research basins of south-central Ontario show that streams in basins >50 ha in size may periodically dry up during summer. Areas of thin or no soil cover shed water quickly (Allan and Roulet, 1994; Buttle et al., 2004) and basins dominated by this surface type experience several days of zero flow during the summer each year. his issue is exacerbated by “fill-and-spill” processes in stream networks (Spence and Woo, 2003) where network connectivity in small basins can become disrupted during summer drought as groundwater levels along the stream channel drop below the elevation of downvalley bedrock sills (Buttle et al., 2004). Conversely, basins that have appreciable till cover (>1 m in depth) can store considerable quantities of spring snowmelt and rainfall and deliver it to the stream network throughout the summer (Devito et al., 1996; Buttle et al., 2004). Unique Aspects of the Shield in the Context of Low Flows In northern parts of the Shield, low flows generally occur in late winter/early spring in advance of spring snowmelt (see Figure 2). his reflects the long period of sub-zero temperatures and the lack of liquid water inputs to the basins, such that streamflow is only sustained by groundwater inputs and the draining of surface water reservoirs. In southern parts of the Shield, periodic water inputs during winter (e.g., rain-on- In the north, prolonged cold periods during winter mean that the precipitation that falls as snow does not immediately contribute to streamflow, such that low flows in northern Shield basins are dictated by the rate of drawdown of storage reservoirs (groundwater, Timing of Low Flows on the Shield © 2008 Canadian Water Resources Association 116 Canadian Water Resources Journal/Revue canadienne des ressources hydriques snow) that can sustain some winter flows coupled with summer evaporation losses from basins mean that low flows generally occur in late summer. hus, the timing of low flows in Shield basins exhibits a transition from a unimodal distribution of late summer low flows in the southern portion of the Shield, to a bimodal distribution (both late winter and late summer flow minima) in central parts of the Shield, to a unimodal distribution of late winter low flows in the northern Shield. Socio-Economic Aspects of Low Flow Conditions on the Shield he Canadian Shield supports over two million people (Urquizo et al., 1998), and although the region’s economy is dominated by forestry and mining, tourism and recreation are of increasing importance. Steedman et al. (2004) estimate that forestry directly affects 60% of the Boreal Shield surface area, agriculture affects <1%, while urbanization and mining both affect <0.1%. hus, the influence on low flows of such activities as water extractions for irrigation of agricultural lands is relatively insignificant on the Shield. Many Shield rivers have been impounded for hydroelectric power production. Reservoir operations generally increase downstream low flows relative to pre-impoundment conditions (Prowse et al., 2004). Changes in stream low flows as a result of reservoir operations and flow diversions can affect communities on the Shield in a variety of ways. hese include the capability of receiving streams to dilute inputs of municipal and industrial wastewater to acceptable concentrations, and the impacts of modified low flow quantity and quality on fish communities that may support recreation and tourism activities. Atlantic Region Description of the Region he Atlantic region comprises the four Atlantic Provinces; New Brunswick, Nova Scotia, Prince Edward Island and Newfoundland. his region is characterized by an extensive forest cover and an extensive coastline. he mountains within Atlantic Canada are the old and weathered northern extension of the Appalachian Range with the highest peaks just over 800 m.a.s.l. Drainage basins in Atlantic Canada are composed mostly of granite, metamorphic and sedimentary rock. Two main types of forest are observed in Atlantic Canada, namely the Boreal and Acadian Forest. he Boreal Forest is observed mainly in Newfoundland, Cape Breton and northern New Brunswick whereas the Acadian Forest occupies much of the southern part of the Maritime Provinces. Precipitation is somewhat homogenous throughout New Brunswick and Prince Edward Island, at approximately 1100 mm annually; however, the southern part of New Brunswick along the Bay of Fundy experiences slightly higher precipitation as well as cooler temperatures. Higher precipitation is also observed in Nova Scotia and Newfoundland with values typically in the range of 1200 to 1500 mm. Newfoundland shows a gradient in precipitation from north (≈ 1200 mm) to south (≈ 1500 mm). he region generally experiences a Maritime and cool climate, especially along the coast; however, hot summer temperatures (>30°C) can be experienced particularly in central New Brunswick. Mean annual air temperature varies between 2°C (northern Newfoundland) and 3°C (northern New Brunswick) to the north and up to 7°C in the southwestern portion of Nova Scotia. Unique Features of the Region – Low Flow Perspective here are regions of Atlantic Canada that have large groundwater contributions, such as Prince Edward Island, where baseflow conditions are maintained throughout the low flow periods. As such, these areas do not normally experience severe low flows. As observed in other regions in Canada, the storage within drainage networks increases with basin size, such that larger rivers experience less severe low flows. Although small catchments generally experience more severe low flow conditions, only a few of the smallest drainage basins (i.e., less than 300 to 500 ha) experience zero flow during the most severe dry periods. Timing of Low Flows he Maritime climate of Atlantic Canada results in highly variable occurrence of low flow conditions © 2008 Canadian Water Resources Association Burn, Buttle, Caissie, MacCulloch, Spence and Stahl throughout the region; this is somewhat related to the mean annual air temperature. he colder regions of northern New Brunswick and Newfoundland experience more winter low flow occurrences whereas the warmer regions of Atlantic Canada almost exclusively experience summer low flows (see Figure 2). For example, both northern New Brunswick and Newfoundland can experience over 70 to 90% of the annual low flows during winter. In contrast, Nova Scotia generally experiences less than 20% of low flows in winter while the warmer region of Nova Scotia (southwestern area) experiences only summer low flow periods. Socio-Economic Impacts of Low Flows Specific to this Region Low flows in Atlantic Canada have a significant socioeconomic impact on a variety of instream (e.g., fish habitat related) and off-stream (e.g., municipal water supplies, small irrigation projects and aquaculture) water uses. Some areas are more problematic than others as a result of higher demand for off-stream use, particularly for agriculture. For instance, Prince Edward Island has a significant water demand for irrigation as does the northwestern part of New Brunswick and the Annapolis Valley in Nova Scotia. Municipal water supplies also exert a significant demand on water resources throughout Atlantic Canada and probably represent one of the most significant water demands in Newfoundland, as agriculture is not as intensive in this province. Comparison of Low Flow Characteristics he diversity of processes that affect low flows in Canada is clear from the summary listed in Table 1. Lack of precipitation and moisture is a fundamental cause of low flows and is therefore common to all regions. Low flows in all regions except southern Ontario and the immediate west coast are affected by sub-freezing temperatures that prevent runoff and therefore result in winter low flow conditions. In many of the regions, low flow conditions are moderated by the slow release of moisture from storage in lakes and wetlands. In a similar manner, moisture from the soil layer can sustain flow conditions during dry periods if the soil layer is 117 thick enough and capable of transmitting flow. Basin size affects low flow conditions in some regions with larger basins generally experiencing less severe low flow conditions because of increased storage. Contributions from aquifers also tend to mitigate low flow conditions in areas where aquifer storage exists and is hydraulically connected to the river. Finally, contributions from the melting of the snowpack and glacier melt impact low flows in the Mountain region. It should be noted that the relative contribution of the different processes to low flows varies from region to region. Several analyses were conducted for the selected gauging stations (see Figure 1) for each region. For a common period of record (1990 to 2000) a flow duration curve (FDC) was derived for each gauging station. he 1990 to 2000 period was selected to maximize the data availability for the stations during a common period of record. A common period of record is important to minimize, as much as possible, the role of climate in producing variations in the FDCs. With a common period of record it can be assumed that all regions experienced similar continental-scale climatic variations. he common period selected was driven by the data limitations for the Arctic region. he FDCs are summarized in Figure 3. For each region, the median FDC and the upper and lower bound for the collection of curves for each region were calculated, although only the median FDCs are displayed in Figure 3 to avoid excess clutter on the graph. Figure 3 plots discharge standardized by catchment drainage area to facilitate comparison of the results. Apparent from Figure 3 is the extreme dryness of the Arctic and Prairie regions, where some catchments experience intermittent streamflow. he FDCs have a steeper slope (i.e., faster decreasing flow with increasing percentage, Figure 3) for these two regions, which generally indicate relatively low baseflow contributions from groundwater and/or lakes or larger reductions in contributing areas during dry periods compared to other regions. he Mountain region experiences the greatest range in FDCs with the lower bound similar to the wettest Prairie conditions and the upper bound representing, by a considerable margin, the wettest conditions of all stations. he southern Ontario, Shield and Atlantic regions exhibit similar magnitude and range of FDCs. hese results suggest similar low flow processes are observed within these regions and these similarities in processes are also reflected in Table 1. he FDCs of the Shield region have a noticeably milder slope (i.e., smaller decreasing © 2008 Canadian Water Resources Association 118 Canadian Water Resources Journal/Revue canadienne des ressources hydriques Figure 3. Flow duration curves for each region. Shown are the median curves derived from the stations for each region. The values of Q50, Q75 and Q90 are highlighted. Table 1. Summary of the dominant processes influencing low flows in regions of Canada. Region Process Low precipitation or moisture availability Sub-freezing temperatures Lake and wetland storage Soil water storage and release Basin size Aquifer storage and hydraulic connection Snowpack, snowmelt and glacier melt Arctic Mountain Prairie S. Ontario Shield Atlantic • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • © 2008 Canadian Water Resources Association Burn, Buttle, Caissie, MacCulloch, Spence and Stahl 119 flow with increasing percentage) than all other regions, including the southern Ontario and Atlantic regions, which suggests that this region experiences more stable flows and corresponding baseflow contributions. Median flows (Q 50) for most regions are very close to each other at approximately 0.02 m3/s/km2 with the exception of the Arctic and Prairie regions which showed much lower Q 50. he low 0.0007 m3/s/km2 Q 50 value for the Arctic region and the 0.0003 m3/s/km2 on the Prairies reflects the dry climatic conditions in these two regions. Flows at a higher percentage of time equalled or exceeded were also very consistent among all regions (e.g., at Q 80 and Q 90) except the Arctic and Prairie regions. Systematically dry conditions often cause intermittent flows in smaller Prairie streams and reduce the Q 80 and Q 90 values. he extremely cold conditions experienced by Arctic catchments exacerbates this pattern of common zero flows in the already dry environment of northern latitudes. At the higher percentages, flows for the Arctic region were generally lower than those of the Prairie region. he characteristics of the FDCs were further explored using a low flow index (LFI) derived based on values from the FDCs resulting in (1) where Q90, Q75 and Q50 represent the flow value exceeded 90, 75 and 50% of the time, respectively. Clausen and Biggs (1997; 2000) suggest the use of the LFI and indicate that the LFI has been found to be correlated with other low flow indices, such as the mean annual minimum flow divided by Q50. Large values of LFI are associated with FDCs with a gradual slope while small values of LFI result from comparatively steep FDCs. Values of LFI versus drainage area are displayed in Figure 4. he intermittent nature of streamflow in the Arctic and, to a lesser extent, the Prairies, is apparent from the predominance of low LFI values from these regions (Figure 4). he Shield catchments tend to have comparatively large values for LFI, presumably as a result of the predominance of lake storage available to maintain streamflow during dry periods. he low flow index of the Arctic and Shield streams converges at larger scales (see Figure 4) perhaps because the larger Arctic streams (i.e., Back, helon) travel through the Canadian Shield. he wide landscape variability commonly experienced at smaller scales is reflected in a general tendency for a greater range of LFI values for smaller catchments than for larger catchments. Figure 5 is a log-log plot of average annual minimum daily flow for each station (calculated for the reference period) versus drainage area. Bars on the graph show the range of values for each station. For graphing purposes, zero annual minimum flows are plotted in Figure 5 as 0.0001 m3/s. Hamilton and Saso (2007) suggest that a plot of minimum annual flow versus drainage area for a collection of sites can be useful for predicting low flows at ungauged sites. Yue and Wang (2004) argue that, based on the scaling behaviour of low flows, drainage area can be used to describe most of the variability in low flows. However, other authors (Kroll et al., 2004; Laaha and Blöschl, 2007) have developed multivariate regression relationships to estimate low flows. Since the focus in this paper is more on understanding regional differences in low flow processes, and less on developing a low flow estimation model for Canadian conditions, a univariate regression model is explored. It should be noted that some of the annual minimum values may be from summer and others from winter (this will be more prevalent in some regions than in others). Since the processes causing low flows differ between seasons, the relationship with basin characteristics may vary as well. his implies that estimation of low flows at ungauged basins is a challenging undertaking (Smakhtin, 2001). Superimposed on the graph in Figure 5 is the best fit line relating annual minimum flow to drainage area of the form (2) where Qmin is the annual minimum daily flow (m3/s), A is the drainage area (km2) and α and β are parameters estimated by regression. he adjusted R2 for the data fit is 0.355 and the equation is shown on Figure 5. here is clearly considerable scatter around this line both in terms of the annual minimum flows for a given site and in terms of the overall position of the collection of flows for individual sites. Several sites exhibit a variation in annual minimum flow, over the analysis period, of several orders of magnitude and a range of annual values that encompasses at least one order of magnitude is common. here is evidently a difference between the majority of the Arctic and Prairie sites and the sites from the remaining regions, with the Arctic and Prairie sites falling well below the regression line for all stations. As a result, two subsequent relationships © 2008 Canadian Water Resources Association 120 Canadian Water Resources Journal/Revue canadienne des ressources hydriques Figure 4. Low Flow Index (LFI) for the stations analyzed versus drainage area for the different regions. were fit, one for the Arctic and Prairie stations and a second for the remaining stations. hese are also shown on Figure 5 as dashed lines, with the resulting equations provided. he line fit to the Arctic and Prairie stations results in an adjusted R2 of 0.548 while the line fit to the stations from the remaining regions has an adjusted R2 of 0.893. While there is still considerable scatter around the Arctic/Prairie relationship, it is apparent that the behaviour of these stations is different from the behaviour of the data for the remaining regions. he slopes of the two new relationships are similar, but the intercepts are not. his may reflect a combination of lower moisture availability for the Arctic and Prairie regions as well as limited contributions from groundwater. he new relationship for the remaining regions now provides a much better fit. he southern Ontario stations generally fall somewhat below the best fit line, although the deviation from the line is not excessive. In addition to the best fit lines, there may also be interest in a lower envelop curve indicating the minimum low flow observed for basins of a given size. his information could be useful for determining the availability of streamflow for dilution of effluent from sewage treatment plants or for deciding on flow abstractions for irrigation or other uses. In these cases, knowing the minimum low flow may be more useful than knowing the average annual low flow for a given basin area. Based on the data in Figure 5, a prudent approach may be to develop lower envelop curves on a regional basis. Conclusions Low flows have a variety of impacts in different regions of Canada, including fisheries concerns, assimilative capacity of receiving water bodies, availability of potable water supply, navigation, and recreation and tourism. Diverse processes have been shown to be influential in controlling low flows in various regions of © 2008 Canadian Water Resources Association Burn, Buttle, Caissie, MacCulloch, Spence and Stahl 121 Figure 5. Annual minimum daily flow versus drainage area. Each station is plotted as a point that corresponds to the mean annual daily minimum flow while the bars define the upper and lower bounds for the annual low flow values. The solid line depicts the best fit relationship fit to all data. The two dashed lines show separate fits to the Arctic/Prairie stations and the remaining stations. Canada. Although there are some common controlling processes, such as low moisture availability, there are also region specific processes, such as snowmelt and glacier melt. Each region was demonstrated to have distinct features that exerted a control on the low flow regime for that region. Flow duration curves dramatically illustrated the relative dryness of the Prairie and Arctic regions and the similarity in the low flow response for the remaining regions. he Prairie and Arctic regions were also shown to respond differently in terms of the relationship between minimum annual daily flow and drainage area. he scatter in this relationship indicates that estimating low flows at ungauged catchments remains a challenge. Future research should focus on this topic. Acknowledgements he research described in this paper was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC) and the Sustainable Forest Management Network. he authors acknowledge the advice and assistance provided by Stuart Hamilton, William Quinton, Ming-ko Woo and Kathy Young. he authors appreciate the helpful comments provided by three anonymous reviewers. © 2008 Canadian Water Resources Association 122 Canadian Water Resources Journal/Revue canadienne des ressources hydriques References Allan, C.J. and N.T. Roulet. 1994. “Runoff Generation in Zero-Order Precambrian Shield Catchments: he Stormflow Response of a Heterogeneous Landscape.” Hydrological Processes, 8: 369-388. Bird, J.B. 1980. he Natural Landscapes of Canada. 2nd Edition. John Wiley & Sons, Toronto, ON. Buttle, J.M., P.J. Dillon and G.R. Eerkes. 2004. “Hydrologic Coupling of Slopes, Riparian Zones and Streams: An Example from the Canadian Shield.” Journal of Hydrology, 287: 161-177. Clausen, B. and B.J.F. Biggs. 1997. “Relationships Between Benthic Biota and Hydrological Indices in New Zealand Streams.” Freshwater Biology, 38: 327342. Clausen, B. and B.J.F. Biggs. 2000. “Flow Indices for Ecological Studies in Temperate Streams: Groupings Based on Covariance.” Journal of Hydrology, 237: 184197. Clayton, J.S., W.A. Ehrlich, D.B. Cann, J.H. Day and I.B. Marshall. 1977. Soils of Canada, Vols. I and II, Agriculture Canada, Ottawa, ON. Devito, K.J., A.R. Hill and N.T. Roulet. 1996. “Groundwater-Surface Water Interactions in Headwater Forested Wetlands of the Canadian Shield.” Journal of Hydrology, 181: 127-147. Ecological Stratification Working Group. 1996. “A National Ecological Framework for Canada.” Agriculture and Agri-Food Canada, Research Branch, Centre for Land and Biological Resources Research and Environment Canada, State of Environment Directorate, Ottawa/Hull, ON. Fisheries and Environment Canada. 1978. Hydrological Atlas of Canada. Supply and Services Canada, Ottawa, ON. Hamilton, S. and P. Saso. 2007. “River-Ice Effects on the Estimation of the Annual Low-Flow in Ungauged Cold-Region Basins.” Presented at the 14th Workshop on Hydraulics of Ice Covered Rivers, Quebec City, QC. Hay, J.E. 1971. “Precipitable Water over Canada: Distribution.” Atmosphere Ocean, 9: 101-111. Hinton, M.J. 2005. “Methodology for Measuring the Spatial Distribution of Low Streamflow in Watersheds.” Geological Survey of Canada, Open File 4891, 62p. Kelly, R.I., P.J. Barnett and A.F. Bajc. 2004. “he Quaternary Geology of Southern Ontario: An Overview and Implications for Groundwater Resources.” Presented at the 49th Annual Meeting of the Geological Association and the Mineralogical Association of Canada. May 12-14, 2004. Brock University, St. Catharines, ON. Kerr, J.A. 1997. “Future Water Levels and Flows for Great Slave and Great Bear Lakes, Mackenzie River and Mackenzie Delta.” In: Mackenzie Basin Impact Study Final Report, Cohen, S.J. (Ed.) Environment Canada, Downsview, ON: 73-91. Krajina, V.J. 1969. “Ecology of Forest Trees in British Columbia.” Ecology of Western North America, 2(1): 1146. Kroll, C.N., J.G. Luz, T.B. Allen and R.M. Vogel. 2004. “Developing a Watershed Characteristics Database to Improve Low Streamflow Prediction.” Journal of Hydrologic Engineering, ASCE, 9(2): 116-125. Laaha, G. and G. Blöschl. 2007. “A National Low Flow Estimation for Austria.” Hydrological Sciences Journal, 52: 625-644. Loftus, K.K., L.A. Greig, T.A. Pinfold, M. Kilfoil and J.D. Meisner. 1995. “Comprehensive Development Strategy for the Anadromous and Inland Recreational Fisheries of New Brunswick.” Canadian Manuscript Report of Fisheries and Aquatic Sciences, 2216: 23p. © 2008 Canadian Water Resources Association Burn, Buttle, Caissie, MacCulloch, Spence and Stahl Prowse, T.D. and T. Carter. 2002. “Significance of Ice Induced Storage to Spring Runoff: A Case Study of the Mackenzie River.” Hydrological Processes, 16: 779788. Prowse, T.D., F.J. Wrona and G. Power. 2004. “Dams, Reservoirs and Flow Regulation.” In: hreats to Water Availability in Canada. NWRI Scientific Assessment Report Series No. 3 and ACSD Science Assessment Series No. 1, National Water Research Institute, Meteorological Service of Canada, Environment Canada, Burlington, ON: 9-18. Quinton, W.L. and M. Hayashi. 2005. “he Flow and Storage of Water in the Wetland-Dominated Central Mackenzie River Basin: Recent Advances and Future Directions.” In: Prediction in Ungauged Basins: Approaches for Canada’s Cold Regions. Spence, C., J.W. Pomeroy and A. Pietroniro (Eds.), Canadian Water Resources Association, Cambridge, ON: 45-66. Quinton, W.L., M. Hayashi and A. Pietroniro. 2003. “Connectivity and Storage Functions of Channel Fens and Flat Bogs in Northern Basins.” Hydrological Processes, 17: 3665-3684. Reedyk, S., M.K. Woo and T.D. Prowse. 1995. “Contribution of Icing Ablation to Streamflow in a Discontinuous Permafrost Area.” Canadian Journal of Earth Sciences, 32: 13-20. 123 Spence, C. and M.K. Woo. 2003. “Hydrology of Subarctic Canadian Shield: Soil-Filled Valleys.” Journal of Hydrology, 279: 151-166. Steedman, R.J., C.J. Allan, R.L. France and R.S. Kushneriuk. 2004. “Land, Water, and Human Activity on Boreal Watersheds.” In: Gunn, J.M., R.J. Steedman and R.A. Ryder (Eds.), Boreal Shield Watersheds: Lake Trout Ecosystems in a Changing Environment, Lewis Publishers, Boca Raton, FL: 59-85. Stewart, R.E., N. Bussieres, Z. Cao, H.R. Cho, D.R. Hudak, B. Kochtubajda, H. Leighton, P.Y.T. Louie, M.D. Mackay, P. Marsh, G. Strong, K.K. Szeto and J.E. Buford. 2002. “Hydrometeorological Features of the Mackenzie Basin Climate System During the 1994/95 Water Year: A Period of Record Low Discharge.” Atmosphere Ocean, 40: 257-278. Todd, A.K., J.M. Buttle and C.H. Taylor. 2006. “Hydrologic Dynamics and Linkages in a WetlandDominated Basin.” Journal of Hydrology, 319: 15-35. Urquizo, N., T. Brydges and H. Shear. 1998. Ecological Assessment of the Boreal Shield Ecozone. Environment Canada, Ottawa, ON. Walsh, J.E., X. Zhou, D. Portis and M.C. Serreze. 1994. “Atmospheric Contributions to Hydrologic Variations in the Arctic.” Atmosphere Ocean, 32: 733-755. Rouse, W.R. 1993. “Northern Climates.” In: Canada’s Cold Environments, French, H.M. and O. Slaymaker (Eds.) McGill-Queen’s University Press: 65-92. Woo, M.K. 1983. “Hydrology of a Drainage Basin in the Canadian High Arctic.” Annals of the Association of American Geographers, 73: 577-596. Scibek, J., D.M. Allen, A.J. Cannon and P.H. Whitfield. 2007. “Groundwater–Surface Water Interaction Under Scenarios of Climate Change Using a High-Resolution Transient Groundwater Model.” Journal of Hydrology, 333: 165– 181. Woo, M.K. 1993. “Northern Hydrology.” In: Canada’s Cold Environments, French, H.M. and O. Slaymaker (Eds.) McGill-Queen’s University Press: 117-142. Smakhtin, V.U. 2001. “Low Flow Hydrology: A Review.” Journal of Hydrology, 240: 147-186. Yue, S. and C.Y. Wang. 2004. “Scaling of Canadian Low Flows.” Stochastic Environmental Research and Risk Assessment, 18: 291-304. Spalding, D.A.E. 1980. A Nature Guide to Alberta. Hurtig, Edmonton, AB. © 2008 Canadian Water Resources Association