This page has been archived on the Web

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please contact us to request a format other than those available.

Skip booklet index and go to page content

Threats to Water Availability in Canada

4. Floods

Alain Pietroniro,1 Robert Halliday,2 Nicholas Kouwen,3 Donald H. Burn,3 Charles Lin4 and Sal Figliuzzi5

1 Environment Canada, National Water Research Institute, Saskatoon, SK
2 R. Halliday & Associates, Saskatoon, SK
3 University of Waterloo, Department of Civil Engineering, Waterloo, ON
4 McGill University, Department of Atmospheric and Oceanic Sciences, Montreal, QC
5 Alberta Environment, Hydrology Branch, Edmonton, AB


Current Status

Flooding in Canada is primarily but not exclusively caused by hydro-meteorological conditions, either individually or in combination (Watt, 1989; Andrews, 1993), which can occur in the form of excess snowmelt-runoff, rain, rain on snow, ice-jams, or natural dams. Anthropogenic causes of flooding include changes in drainage patterns resulting from urbanization and flooding due to dam-breaks. Brooks et al. (2001) note that over the last century, damages in Canada have exceeded $2 billion with over 198 lives lost. The recent devastating floods on the Saguenay River during the summer of 1996 and the Red River in the spring of 1997 have brought the reality of natural hazards home to Canadians. The Saguenay flood occurred due to an unprecedented precipitation event over a 24-hour period and resulted in more than $1 billion in damages and loss of 10 lives (Commission scientifique et technique sur la gestion des barrages, 1997). The “flood of the century” in southern Manitoba was caused by a combination of hydro-meteorological factors, beginning with high antecedent soil moisture, heavy winter snowfall, and a rapid spring melt. Although there was no direct loss of life, damages (including flood-fighting costs) were estimated at $500 million in Canada and over $2 billion in total (IJC, 2000). Infrequent, large flood events can also cross the erosive thresholds along alluvial rivers and result in catastrophic erosion along valley bottoms. Such erosion represents another major risk from flooding, in addition to the inundation damage from floodwaters, and it can result in significant losses of property and infrastructure, even when these are situated above the flood level.

Public perception of flooding is often as a natural hazard that should be mitigated, and it is in this context that this chapter is written. It is important to understand, however, that there are beneficial ecological aspects of flooding that are often not considered, yet which are an important component of ecological sustainability. For example, changes in the flooding regime of the Peace-Athabasca Delta in northern Alberta have been attributed to a decrease in ice-jam frequency due to climatic changes and river regulation (Prowse and Conly, 1996). These changes, in turn, have resulted in a decline of fish and wildlife habitats and disruption of the entire ecosystem; hence, a change in flooding regime can have significant ecological impacts. Such impacts are not considered for this chapter, however, and the focus is confined to flooding as a natural hazard.

Adaptation and Mitigation

From the earliest settlements in Canada, people have chosen to live and work along rivers and lakes. Apart from their obvious value as a source of drinking water, these water bodies also supply a source of irrigation and power and a means of transportation, while the riverbank and floodplains provide aesthetically pleasing sites for housing and easy access for industrial discharge. As encroachment into the natural floodplain has increased with population growth, so has the damage caused by flooding. Structural measures such as dams, dykes and diversions have been engineered to mitigate flood risks, giving residents a sometimes false sense of security; however, these measures also have the potential to disrupt riparian ecosystems. Non-structural approaches to reducing flood damages, including floodplain regulation and flood forecasting, have gained favour therefore, and they provide a sound mitigation strategy for reducing damages. In both the structural and non-structural approaches to mitigation and/or adaptation, thorough hydrological analysis is required, typically including flood-frequency analysis, hydraulic and hydrological modelling studies, and robust engineering design.

Floodplain regulation requires defining the floodplain at an elevation that provides an acceptable level of risk to residents living above the defined zone, and suitable use (e.g., public parks, recreation areas) of the flood zone. Floodplain delineation is typically based on a derived regulatory river flow. These flow rates are often based on frequency analysis (statistical properties) of historic streamflows at a specified river location (Burn, 2002); however, in ungauged regions, they can also be based on regional analysis and comparison of gauged and ungauged regions, or through application of a hydrological model to a known design storm. In all cases, resulting flows are used in concert with a hydraulic model to determine water surface elevations for the river-reach or location of interest. The resulting elevation along the river banks becomes the basis for most floodplain zoning and planning. At the time of the termination of the Federal-Provincial Flood Damage Reductions Program (FDRP), whose primary purpose was to map urban flood-prone lands, approximately 700 out of 1100 Canadian flood-prone communities were mapped (Shrubsole et al., 2003). Although, historically, the FRDP analysis included most of the high-risk communities in Canada, many Canadians remain unaware of their exposure to flood hazards.

Structural approaches to flood mitigation usually include multi-purpose structures such as large dams. These structures are more often designed for preserving low flows, hydro-power production and irrigation; however, they can often provide some protection of downstream communities from extreme events. Design of these structures requires analysis of potential large flows for adequate spillage and dam safety requirements. The Probable Maximum Precipitation (PMP) and the Probable Maximum Flood (PMF) are criteria commonly used in design and safety analysis for all large dams in Canada (see discussion below). Any engineered flood control structure such as a dam, diversion or dyke must not only be designed appropriately, but must also be operated and maintained effectively.

Radarsat standard mode 6 image of the 1997 "Flood of the Century" near Winnipeg, Manitoba | Photo: T.J. Pultz

Radarsat standard mode 6 image of the 1997 "Flood of the Century" near Winnipeg, Manitoba, acquired on May 1, 1997. The flood extent is clearly evident as is the ring dike around the town of Morris and the floodway providing protection for the City of Winnipeg. Copyright 1997 Canadian Space Agency. Data processed and distributed by Radarsat International; data enhancement and interpretation by Canada Centre for Remote Sensing.

Flood Forecasting

Flood forecasting is a key non-structural approach to reducing flood losses and is highlighted here because of its potential. Forecasting is typically achieved through application of one or more modelling systems and there are many hydrological and hydraulic models available for this purpose. Hydrological models range in sophistication from a simple statistical rainfall-runoff relationship to more detailed, physically based algorithms describing the complete rainfall-runoff system. Hydraulic models are also used in flood forecasting to calculate travel time of the flood wave and its attenuation. These models use the standard equations of unsteady, non-uniform flow with various simplifications depending on the channel characteristics, available data, and accuracy requirements. Probabilistic models that take data uncertainties into account are also available. Probabilistic models apply a statistical distribution to input parameters such as precipitation forecasts and produce a large number of model runs that are statistically analyzed. The resulting forecast provides an entire distribution of future plausible conditions rather than a single outcome.

Dam Safety

While aspects of dam safety are described in Chapter 2, for the purposes of this discussion the failure of a flood control dam resulting from an extreme flood is explored--currently a very important issue for many provincial flood management agencies. Such high consequence dams are often designed to withstand the PMF, defined by the U.S. Federal Energy Regulatory Commission (2002) as "the flood that may be expected from the most severe combination of critical meteorological and hydrologic conditions that are reasonably possible in the drainage basin under study." Theoretically, the probability of exceeding the PMF is zero, and hence the dam can safely withstand all floods. The Canadian Dam Association’s Dam Safety Guidelines provide a similar definition (Canadian Dam Association, 1998).

As the estimation of the PMF uses historical data, it is re-estimated periodically as more data are collected. Occasionally, the revised PMF becomes significantly higher as new data are added. For example, Jarrett and Tomlinson (2000) provided an example where the revised PMF for the Olympus Dam in Colorado was almost four times larger than the original estimate. When this situation occurs, the dams may fail the safety check, leading to expensive spillway re-design and re-construction. There is, therefore, a considerable amount of concern about the validity and robustness of techniques used for PMF estimation.

The PMF is often calculated as the flood generated by the most severe precipitation possible at a site at a particular time of year, referred to as the PMP. The World Meteorological Organization (WMO) commissioned a manual to describe techniques to estimate the PMP (World Meteorological Organization, 1986). The manual explains that the method for estimating the PMP cannot be standardized and may need to be modified for a particular region (World Meteorological Organization, 1986). The techniques depend on the size and location of the basin of interest, the amount and quality of data available at the site, and meteorological conditions that produce severe precipitation events. These problems are particularly severe in high-relief areas, such as the Coast Mountains in Canada and the United States, where due to orographic effects, strong precipitation gradients exist. For example, with the use of paleo-hydrology records, it has been shown that PMP estimates in mountainous regimes are generally too high (e.g., Parrett and Jarrett, 2000). The WMO manual states that the PMP must be considered an estimate and that its accuracy cannot be assessed in an objective manner (World Meteorological Organization, 1986). Therefore, these standard PMP procedures may need to be modified for a particular region. Also, the use of deterministic rather than statistical approaches to PMP and PMF estimation could receive some consideration. Abbs (1999) recommended that increased effort be placed on numerical modelling of extreme rainfall events.

Fig. 1 Flood damages in Canada (after Brooks et al., 2001).

Fig. 1 Flood damages in Canada (after Brooks et al., 2001). Adjusted values based on Construction Price Index (Kulshreshtha, 2003).


Adaptation and Mitigation

Improvements in statistical methods along with technological advances in mapping have the potential to re-invigorate and renew mitigation strategies in many Canadian locations. The federal government currently has no role in floodplain management and some provincial governments have also terminated their programs, while others are applying the information obtained during the Flood Damage Reduction Program (FDRP) without modification. Of some concern is the fact that hydrological knowledge gained through improvements in understanding and longer observation periods in the last two decades has not been used to re-assess flood risks established under FDRP. If this situation continues, exposure to potential loss of life and property due to flooding may increase, as will the threat to federal resources in the form of federal disaster assistance. It would be prudent to examine flood risk in Canada in this context, as well as in light of new potential changes to hydrological regimes in Canada resulting from climate change.

In terms of improved flow analysis, there has been a recent tendency to incorporate regional analysis as opposed to single-site analysis for the estimation of flood frequency curves. As an example, the Flood Estimation Handbook (IH, 1999) recently developed in the U.K. incorporates a focussed regionalization approach (Burn, 1990) for estimating at-site design flood magnitudes. Regrettably, there is no equivalent set of guidelines for Canada nor are there currently federal standards or guidelines for regional analysis. As such, different approaches, of varying levels of sophistication, are employed in different provinces.

Contemporary remote-sensing techniques, such as Light Detection and Ranging (Lidar) for elevation mapping, and global position technologies (GPS) in combination with Geographic Information Systems (GIS) provide an opportunity to re-engineer flood mapping strategies. It is now feasible to create an accurate representation of a floodplain quickly and at relatively low cost.

Forecasting and Models

Combining weather forecasts with information on watershed conditions and a hydrological streamflow forecasting model can give advanced warning (up to 48 hours) of potential flash-flooding and can help save lives and reduce property damage. Precipitation measurements and forecasts are the most uncertain inputs in a flood forecasting system. Over the last couple of decades, knowledge of atmospheric processes has improved, computers have become more powerful, and it has become possible to model the atmosphere numerically. There are several Canadian atmospheric models produced by the Canadian Meteorological Centre that can be used for short-term weather modelling, including the Regional Finite Element (RFE) model (Mailhot et al., 1998), the Global Environmental Multiscale (GEM) model (Côté et al., 1998), and the Mesoscale Compressible Community (MC2) model (Benoît et al., 1997). These models can operate in forecast mode, where initial atmospheric conditions are specified and the model physics are used to predict future weather conditions (currently, the GEM model is used for weather forecasts in Canada).

Coupling hydrological forecasts to a well-calibrated, high-resolution Numerical Weather Prediction (NWP) model not only offers a valid substitute for precipitation data, but, more importantly, it provides a means to compute forecasted river flows with forecasted precipitation data. Research has recently been completed to develop further the optimal coupling of a high-resolution regional atmospheric model (for example the Canadian MC2) with a hydrological model (WATFLOOD) for flood forecasting. To achieve this, a conceptual framework for model development was initiated using different degrees of model coupling to obtain a complete two-way coupled model. The first level of coupling is referred to as model linking and requires calibrating and validating the high-resolution regional atmospheric model and the distributed hydrological model separately. This form of coupling then uses the simulated precipitation from the atmospheric model to drive the hydrological model to study flash flood events. A real-time experiment was initiated under the Mesoscale Alpine Project (MAP) in 1999 to forecast flash floods in the European Alps using 24-hour NWP forecasts for precipitation and temperature (Kouwen and Benoit, 2002). Under this modelling scenario, the two models use their own inherent land surface scheme and parameterizations. The hydrologic model forcing is derived directly from the atmospheric forecasts with the dominant input being precipitation. This is of great interest as precipitation is the single most uncertain variable for such hydrological studies.

In order to achieve truly coupled systems, the link between the models (MC2 and WATFLOOD) was established by implementing a common land surface scheme (the Canadian Land Surface Scheme [CLASS]) in each model. This model development took place under the NSERC-funded Simulation of Severe Precipitation for Flood Forecasting (SSPFF) research program. Results were promising and showed the ability to predict adequately precipitation and streamflow for the Saguenay event of 1996. Since that time there have been advances in developing an independent system for coupling atmospheric and hydrological models that do not require the common land-surface scheme. A modelling framework described by Pietroniro and Soulis (2003) as part of the Mackenzie GEWEX study (MAGS) has provided a conceptual context for further advancements of these coupled models. An example of such a coupled modelling study for the 1996 Saguenay flood is Lin et al. (2002). This and other developments have the potential to advance both hydrological and atmospheric research while maintaining operational linkages. There have been tremendous gains in applying distributed hydrological models for flood forecasting, and often the major sources of uncertainty are the meteorological forcing variables, particularly precipitation.

Land-Use Impacts

There is a widely held public perception that urbanization and land-use changes in the upper basin have an influence on large floods. This may or may not be the case. Contemporary urban design usually requires that stormwater management systems in new subdivision development not lead to an increased runoff. Often this is administered by local bylaw. In the absence of such regulations, runoff increases can take place. Current-day land-use changes in rural areas, such as destruction of wetlands or drainage, are unlikely to lead to significant or predictable changes in the frequency of large floods. This is because the changes tend to be modest in relation to the overall size of the basin. As well, during a large flood, the natural hydrometeorological conditions tend to overwhelm anthropogenic effects. Nevertheless, research into the effects of land-use change on smaller floods, particularly in sub-basins, would be useful.

Knowledge and Program Needs

Floodplain Mapping

The reality of floodplain management is that unless a flood actually occurs, the threat of flooding often falls out of the public consciousness. Improvements in statistical approaches, GIS and mapping technologies, and improved simulation models highlight the need for a review of past mitigation and adaptation strategies. There is a need to re-examine standard accepted engineering approaches to mitigation in the context of these advancements.

Potential improvements in determining regulatory floods include a number of approaches such as regional analysis that must be considered in addition to improving conventional single-station, flood-frequency analysis. Flood-frequency analysis based on the historic record of annual peak floods is also a fundamental tool in determining design discharge for floodplain zoning, flood protection infrastructure, and structures that span rivers. A basic assumption in frequency analysis is that climatic trends or cycles do not affect flood flows, but there is clear evidence that this is not the case (Gosnold et al., 2000), and that even modest changes in climate can result in large changes in flood magnitude (Knox, 1993). Climate change impacts in terms of precipitation should also be examined, including updated intensity-duration-frequency precipitation curves, which are vital to proper urban engineering design.

With the advent of remote-sensing techniques, in particular LIDAR and GPS, in combination with GIS, weather radar, and the improving ability of NWP models to forecast, now-cast and hind-cast precipitation, it is time to re-think flood mapping and forecasting mitigation strategies. Hydrological and hydraulic models applied in conjunction with atmospheric models in a probabilistic framework may provide a viable mechanism to examine potential future climate change scenarios in the context of floodplain mapping. These improved statistical and regional analyses of floods within a climate change context also should be examined.

Forecasting and Models

Mathematical models play an increasingly important role in flood mitigation and forecasting. There are two trends, however, that call for the application of significant research effort. First, the increasing availability of remotely sensed data, such as precipitation, snow water equivalent or evapotranspiration, requires modification of models to accept spatially distributed as well as more traditional point data. Improvements to the algorithms for transforming data to useful information, such as improved data assimilation algorithms using combinations of observed and modelled data, are vital to improvements in modelled precipitation--and consequently to hydrological forecasts. Clearly, continued improvements to operational weather models are also required.

The hydrological community needs to focus on hydrological modelling systems that are consistent with atmospheric modelling practice. As such, research into physically based distributed models that can be linked or coupled with atmospheric models should be highly encouraged. These models should be based on existing atmospheric land-surface schemes, and provide the link between the atmosphere and the land surface, simulating both the water and energy balance at the land surface. Physically based hydrological models allow for a more rigorous examination of discrete hydrological processes such as precipitation, interception, infiltration, interflow, and baseflow (Soulis et al., 2000). Such models could be used to examine, for example, anthropogenic impacts on a watershed. Issues such as effects of conversion of a land surface to agricultural purposes, e.g., drainage development or wetland destruction, on runoff volumes and peaks generate considerable public debate. These models should be developed by the hydrological community, and based on continued process and basin-scale experiments. Given the diversity of landscapes, and the heterogeneity of the land surface, continued improvements in hydrological models require an effective synergy between experimentalists and model developers in hydrology.

Data Needs

  • Accurate atmospheric forcing and hydrological information is essential for hydrological modelling and infrastructure design, particularly with the advent of climate change. A more consistent hydrometeorological approach to network design should be established. Streamflow gauges are integrators of atmospheric forcing, and the network in Canada should reflect this reality.
  • Techniques to make effective use of ground-based radar estimates of precipitation are essential for improved design and mitigation.
  • Cryospheric datasets of variables, in particular SWE and snow cover extent provide vital information for flood forecasters. There is no systematic system for estimating these quantities, and currently no established archive of these data.
  • Current technology allows for establishment of in situ soil moisture monitoring stations. These types of data could be established within the existing infrastructure and are vital input into data assimilation methods.
  • An ongoing research effort is required to develop reliable remotely sensed methods for monitoring soil moisture, SWE and other important state variables over different land-cover and terrain types in Canada.
  • Accurate, high resolution DEMs are needed for a variety of modelling applications.
  • Potential users need to be aware of NWP output products that can benefit the operational flood-forecasting community (e.g., distributed forecasts of local runoff or soil moisture), and could also be archived and made available to the engineering-design community. Data-assimilated variables from the NWP provide our best estimate of the state of the atmosphere and the land surface, and represent, where possible, a blending of model and observed data.

Research Needs

  • Currently the technology used in operational forecasting models lags the research models by 10 to 20 years, although some have been upgraded to include new data sources. There is a need to test newer and more comprehensive models in operational settings. The major difference between these applications is in the implementation and use of spatial data, both from a meteorological forcing aspect (this includes data-assimilated meteorological inputs in real time, remote sensing state variables) and from a physiographic input perspective (e.g., digital elevation models, satellite-derived land-cover information).
  • Flood frequency analysis based on the historic record of annual peak floods is also a fundamental tool in determining the design discharge for floodplain zoning, flood protection infrastructure, and structures that span rivers. Frequency analysis requires an assumption of stationarity so that climatic trends or cycles do not affect flood flows, but there is clear evidence that this is not the case (Gosnold et al., 2000), and that even modest changes in climate can result in large changes in flood magnitude (Knox, 1993). A research challenge is to determine how aspects of climate change can be incorporated into flood frequency analysis for planning purposes.
  • The use of weather models coupled with hydrological models may provide one of the only avenues to explore future scenarios for adaptation and mitigation. Furthermore, the basic assumptions of homogeneity and independence of any time series of flood peaks can easily be called into question, particularly when evaluating the relatively short Canadian climate and hydrometric records (Booy and Morgan, 1985; Klemes, 1987; Watt, 1989). Potential improvements in the determination of regulatory floods include a number of approaches, such as regional analysis, that must be considered in addition to improving conventional single-station, flood-frequency analysis.
  • The method of calculating the PMF or other design storms should be examined, particularly under the pretext of climate change. Design storms derived from historic data (e.g., Intensity-Duration-Frequency curves, 100-year flow estimates) may have changed substantially for certain regions of the country. There is an important need to re-assess these design criteria from a deterministic modelling and statistical approach.
  • Erosive thresholds are most likely to be crossed during severe floods along rivers where the channel planform is close to the meandering-braid transition. A better understanding of the erosive threshold would allow river reaches that are susceptible to large-scale erosion, during extreme floods, to be recognized, and assessments to be made of the vulnerability of valley bottom development and infrastructure to large-scale erosion.


Abbs, D.J. 1999. A numerical modeling study to investigate the assumptions used in the calculation of probable maximum precipitation. Water Resour. Res. 35(3): 785-796.

Andrews, J. (ed.). 1993. Flooding: Canada water year book. Ministry of Supply and Services, Ottawa.

Benoît, R., M. Desgagné, P. Pellerin, S. Pellerin, Y. Chartier and S. Desjardins. 1997. The Canadian MC2: a semi-Lagrangian, semi-implicit wideband atmospheric model suited for finescale process studies and simulation. Monthly Weather Review 125: 2382-2415.

Booy, C. and D.R. Morgan. 1985. The effect of clustering of flood peaks on a flood risk analysis for the Red River. Can. J. Civil Eng. 12: 150-165.

Brooks, G.R., S.G. Evans and J.J. Clague. 2001. Flooding, p. 101-143. In G.R. Brooks (ed.), A synthesis of natural geological hazards in Canada. Geological Survey of Canada Bulletin 548. Ottawa.

Burn, D.H. 1990. Evaluation of regional flood frequency analysis with a region of influence approach. Water Resour. Res. 26: 2257-2265.

Burn, D.H. 2002. Hydrological extremes, p. 1017-1021. In El-Shaarawi, A. and W. Piegorsch (ed.), Encyclopedia of environmetrics, Volume 2. John Wiley & Sons, Ltd., Chichester.

Canadian Dam Association. 1998. Dam safety guidelines. Edmonton, Alberta.

Commission scientifique et technique sur la gestion des barrages. 1997. Report, The scientific and technical committee on the management of dams in Quebec. Canadian Dam Safety Association, Quebec.

Côté, J., S. Gravel, A. Méthot, A. Patoine, M. Roch and A. Staniforth. 1998. The operational CMC-MRB global environmental multiscale (GEM) model. Part I: design considerations and formulation and Part II: results. Monthly Weather Review 126: 1373-1418.

Federal Energy Regulatory Commission. 2002. Engineering guidelines for the evaluation of hydropower projects. Washington, D.C. 644 p. Internet. Cited 2 October 2003.

Gosnold, W.D., J.A. LeFever, P.E. Todhunter and L.F. Osbourne Jr. 2000. Rethinking flood prediction: does the traditional approach need to change? Geotimes 45: 20-23.

IH (Institute of Hydrology). 1999. Flood Estimation Handbook. Centre for Ecology and Hydrology Wallingford. Wallingford, U.K.

IJC. 2000. Living with the red. A Report to the Governments of Canada and the United States on Reducing Flood Impacts in the Red River Basin. International Joint Commission, Ottawa and Washington.

Jarrett, R.D. and E.M. Tomlinson. 2000. Regional interdisciplinary paleoflood approach to assess extreme flood potential. Water Resour. Res. 36(10): 2957-2984.

Klemes, V. 1987. Hydrological and engineering relevance of flood frequency analysis. In V.P. Singh (ed.), Hydrologic frequency modelling. Kluwer Academic Publishers.

Knox, J.C. 1993. Large increases in flood magnitude in response to modest changes in climate. Nature 361: 430-432.

Kouwen, N. and R. Benoît. 2002. Regional forecasting of river flows using a high resolution numerical weather model coupled to a hydrological model. In Proceedings, International conference on flood estimation, Bern, Switzerland, March 6-8, 2002.

Lin, C.A., L. Wen, M. Béland and D. Chaumont. 2002. A coupled atmospheric-hydrological modeling study of the 1996 Ha! Ha! River basin flash flood in Quebec, Canada. Geophys. Res. Lett. 29(2): 13/1-13/4.

Mailhot, J., S. Bélair, R. Benoît, B. Bilodeau, Y. Delage, L. Fillion, L. Garand, C. Girard and A. Tremblay. 1998. Scientific description of RPN physics library, Version 3.6. Recherche en Prévision Numérique, Atmospheric Environment Service, Dorval, Québec. Internet. Cited 2 October 2003.

Parrett, C. and R.D. Jarrett. 2000. Flood hydrology for Dry Creek, Lake County, Northwestern Montana, p. 4.13.1-4.13.10. In Hotchkiss, R.H. and M. Glade (ed.), Proceedings, building partnerships: joint conference on water resources engineering and water resources planning & management. ASCE.

Pietroniro, A. and E.D. Soulis. 2003. A hydrology modelling framework for the Mackenzie GEWEX Program. Hydrolog. Process. 17: 673-676.

Pietroniro, A., E.D. Soulis, K. Snelgrove and N. Kouwen. 2001. A framework for coupling atmospheric and hydrological models. In Dolman, A.J., A.J. Hall, M.L. Kavvas, T. Oki and J.W. Pomeroy (ed.), Soil-vegetation-atmosphere transfer schemes and large scale hydrological models. IAHS Publication No. 270.

Prowse, T. and M. Conly. 1996. Impacts of flow regulation on the aquatic ecosystem of the Peace and Slave rivers. Northern River Basins Study Synthesis Report No. 1.

Shrubsole, D., G. Brooks, R. Halliday, E. Haque, A. Kumar, J. Lacroix, H. Rasid, J. Rousselle and S.P. Simonovic. 2003. An assessment of flood risk management in Canada. ICLR Research Paper Series - No. 28. Institute for Catastrophic loss Reduction. University of Western Ontario

Soulis, E.D., K.R. Snelgrove, N. Kouwen, F. Seglenieks and D.L. Verseghy. 2000. Towards closing the vertical water balance in Canadian atmospheric models: coupling of the land surface scheme CLASS with the distributed hydrological model WATFLOOD. Atmosphere-Ocean 38(1): 251-269.

Kulshreshtha, S. 2003. Personal communication.

Watt, W.E. (ed.). 1989. Hydrology of floods in Canada: a guide to planning and design. Associate Committee on Hydrology, National Research Council, Ottawa.

World Meteorological Organization. 1986. Manual for estimation of probable maximum precipitation. Operational Hydrology Report No. 1, WMO No. 332, Geneva, Switzerland.

Date modified: