This page has been archived on the Web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

Audit and Evaluation Annual Report 2009–2010

3 Findings and Recommendations – Network Sustainability and Configuration

3 Findings and Recommendations – Network Sustainability and Configuration

3.1 Comparison of Hydrometric Programs

A detailed comparison of the effectiveness of the hydrometric programs in different countries would require a level of study that was beyond the scope of this assessment. However, it is possible to use a few measures that, although far from being perfect, give some comparative information about the hydrometric program in selected countries.

The first simple measure used is the geographical hydrometric station density. The World Meteorological Organization recommendations for hydrometric station density range from a minimum of 1 station per 1000 km2 to 3.3 stations per 1000 km2 (World Meteorological Organization 1981). This comparison measure is simplistic and does not take into account other parameters. For example, countries with spatially variable climate, complex terrain and varying land use would require a higher station density than a country with more uniform climate, topography and land use.

A second simple measure used is the station density per volume of water in a given country. This measure recognizes that the more water a country has, the more hydrometric stations are needed. To assess the volume of water, the audit team used the Total Actual Renewable Water Resource (ARWR), which “gives the maximum theoretical amount of water annually available for each country in cubic kilometres” (World Resources Institute). However, again, this measure does not take into consideration other parameters such as those mentioned above.

A third measure used is related to the capacity of various countries to fund their hydrometric network. For this measure, the audit team used the Gross National Income (GNI) adjusted by Purchasing Power Parity (PPP), which is the sum of value added by all resident producers plus any product taxes (less subsidies) not included in the valuation of output, plus net receipts of primary income (compensation of employees and property income) from abroad. In other words, GNI measures the total income of all people who are citizens of a particular country.

Table 4 provides a comparison of those three measures for seven countries, including Canada.

The analysis of Table 4 indicates that four of the seven countries surveyed have station densities in the range (1 to 1.33) recommended by the World Meteorological Organization. It is notable that these four countries are the smallest in terms of land and have the highest population densities. The three largest countries in terms of land (United States, Canada and Australia) have low station densities combined with the lowest population densities. The reasons for this pattern are likely to be the following:

  • the high costs of maintaining stations in unpopulated areas;
  • the lower population density, which provides a lower tax revenue per unit area to support government inventory programs.

Table 4: Comparison of hydrometric network

The station densities for provinces and territories in Canada follow a similar pattern. Only Prince Edward Island, the smallest province, has a station density within the range recommended by the World Meteorological Organization.

As well, Table 4 shows that Canada has the highest total ARWR to cover, by a significant difference margin, and the lowest station density per volume of water (total ARWR), which means that the Canadian National Hydrometric Program is small compared to its responsibilities.

Finally, the station density per GNI indicates that Canada and Australia are two countries investing more in their hydrometric network when the capacity to fund programs is taken into consideration. Of course, from a hydrologic perspective, GNI is irrelevant when the objective is to properly characterize water resources over a large and hydrologically complex area such as Canada.

Furthermore, the overriding issue with the size of the hydrometric network is that the importance of our water monitoring programs is undervalued. It has been estimated (Environment Canada 2004) that water’s measurable contribution to the Canadian economy ranges from $7.5–$23 billion annually and the amount invested in water monitoring should reflect this economic value.

3.2 Network Size and Configuration

Based on documentation reviewed and interview results, the size and structure of the hydrometric network is considered insufficient for the overall characterization of water resources in Canada. The needs of specific clients are met where a gauge is established for a particular purpose. It is axiomatic that where data are collected for a specific purpose, those data needs are likely to be met. The needs of other clients for water resources planning, environmental assessment, project approvals, climate change analysis and other scientific requirements are not generally met by the current network configuration, particularly in northern Canada. It was noted that it is not just a question of the total number of stations, as some stations are located for specific needs and are not always in the best locations for research and hydrological analysis. Particular concern was expressed regarding the loss of key long-term stations in the Hydrometric Basin Network. The loss of stations in the 1990s due to budget cuts was a loss to hydrologic records in Canada.

The network was originally developed for water resources engineering purposes (flood plain management, hydropower), and new issues of greatest concern are not well addressed by the current network configuration. Climate change also results in data becoming outdated where there are significant shifts in the hydrologic signal. All these needs require consideration of climate variability and change, and it is this issue that underpins the most pressing need for improved data collection in Canada.

The changing climate in Canada can be observed through reductions in glacier mass and trends in temperatures in the Arctic. The influence of the changing climate on water resources is much less evident. The annual, seasonal and daily variability observed at water monitoring stations is much greater than any underlying long-term trends in the data, and therefore the long-term trends are difficult to detect. This is made even more complex by the existence of short-term trends, some lasting several decades that reflect climate oscillations resulting from, among other things, periodic changes in ocean currents. Thus, a trend in stream flow may be detected over the past 30 years but it does not necessarily mean that the trend will continue, as it may be caused by a climate oscillation. The complexity is further compounded by land-use changes and possible water use upstream of a flow gauge that would affect the flow records.

Detection of climate change is difficult given that the time aspect of trends is not consistent because of climate oscillations, land use changes and water use. Furthermore, the spatial distribution of trends is also not consistent. Recent research by Ehsanzadeh and Adamowski (2007) found that water monitoring stations in northern Canada have experienced an upward significant trend in seven-day low flows while a significant downward trend dominated the Atlantic provinces and southern British Columbia. In other parts of the country, no significant trends were found. The shifts in the annual timing of seven-day low flows also varied across the country.

Predictions of the impacts of climate change on water resources have primarily been carried out with the use of atmospheric and ocean Global Coupled Models (GCMs), to provide future climate scenarios in order to drive hydrologic models that predict changes in rivers, lakes and streams. However, GCMs are not effective for predictions of extremes (both wet and dry), which are the most important issues in water resources planning. The models are considered reasonable for prediction of average climate conditions, but averages do not tell the complete story. For example, at a specific location, the average rainfall may be projected to increase with climate change, but high intensity rainfalls may decrease and periodic drought conditions may be more prevalent.

Given the current inadequacy of prediction tools for making specific forecasts of the impacts of climate change on water resources, it is imperative that robust water monitoring networks be maintained and enhanced. However, even with the most comprehensive water monitoring system in place, it is still extremely difficult to detect whether the predictions of the impacts of climate change are valid. The projected changes are gradual and are masked by the “noise” of the natural variability and climate oscillations. Rigorous statistical and scientific tools have to be applied to a relatively long data set to determine with any confidence whether a trend exists. In addition, hydrologic models can be improved with comprehensive spatial data in pristine areas, collected over a shorter time period. Therefore, a monitoring strategy should focus on maintaining long-term stations in pristine areas not affected by land use changes and water use, and on establishing new stations in pristine areas not currently represented. Stations located in small  and medium-sized basins provide these kinds of opportunities. Larger basins are more likely to be influenced by storage, abstractions and land use changes.

In the absence of definitive trends in water resources data, engineers and planners are continuing to primarily rely on analysis of historical data for decision making. At some point, engineers, water resource planners and policy makers who depend on professional advice will have to adopt climate change trends as part of project design. The magnitude and nature of those trends can only be definitively determined from analysis of data from a robust water monitoring network. Alternative approaches could be used, such as adding arbitrary safety factors to account for climate change. However, these arbitrary factors are not scientifically defensible and could lead to over-designed structures and other costly decisions. At a specific location in Canada, it is not known with certainty whether the magnitude of floods will increase or decrease or whether droughts will be more or less severe.

Collection and organization of data are the absolute basis of the scientific method. Without data, Mariotte would never have discovered in 1684 that rainfall was the origin of flow in the River Seine. Prior to his findings it was thought that river flow originated in underground springs. It could be argued that much of our current understanding of the effects of climate change is as primitive as the idea that all flow in the River Seine originates in springs. We require extensive data networks to support rigorous scientific analysis, so that we will be able to make well-informed decisions regarding the impacts of climate variability and change on water resources in Canada.

A business review of the hydrometric network in British Columbia (Azar et al. 2004) investigated the economic benefits of the hydrometric network. It was found that sectors such as water supply, agriculture and sewage disposal are reasonably well-served by the hydrometric network. The major economic sectors of forestry, transportation, small hydro, mining, and oil and gas are the least well-served. These sectors require short- and long-term regional data, often on small streams throughout the province.

The benefits of the hydrometric network for all sectors primarily relate to cost savings in design and construction and reduced operating costs. Where data are inadequate, there is increased uncertainty in the design process, and there are increased risks of project failure and/or environmental impacts. Sometimes, conservative decisions are made to compensate for the risk, which leads to increased costs and can affect project feasibility. These decisions are made not only by project designers and operators but also by regulators. This sometimes leads to less water being allocated for the project than is available; in other instances, approvals may be delayed or, in the extreme, not provided due to this uncertainty.

In the case of the small hydro and mining sectors, investors require low risk regarding available water supplies for power generation, mill operation and waste disposal. Stream flow records are key to demonstrating project feasibility, and the absence of adequate data can lead to reduced investment.

The estimated benefit/cost ratio of the current hydrometric network in British Columbia was estimated to be 19.1 (Azar et al. 2004). Every dollar spent continuing to support the present network returns more than nineteen dollars in benefits. It was also concluded that expansion of the network, including integrating other data in addition to the data recorded under the National Hydrometric Program, is in the best economic interests of the province, and would promote provincial goals of economic growth and sustainable resource development.


6. As part of the strategic planning referred to in recommendation #1, the Assistant Deputy Minister, Meteorological Service of Canada, with the collaboration of the NAT and NHPCC, assess the current network risks and vulnerabilities, evaluate the demands and establish priorities to ensure the network provides the largest benefits for the financial resources available and that resources are optimized to address the areas of greatest concerns.

7. The Assistant Deputy Minister, Meteorological Service of Canada, with the collaboration of the NAT and NHPCC, consider the integration of the hydrometric network with the climate data network, both for network design and data reporting, to improve the scientific value of the hydrometric and climate networks.

3.3 Sustainability of a Coordinated Monitoring Program

The Water Survey of Canada is a well-respected organization among water resources professionals in Canada. A nationally coordinated program is the best method of hydrometric data collection, particularly for data archiving and access and for establishing national standards. It was noted that signed Agreements on Hydrometric Monitoring with the Parties assist the continuity of the program.

There are concerns about the decrease in the number of Water Survey of Canada stations since 1980, and the apparent lack of commitment to overall long-term funding for the network. Specifically, the national hydrometric network is inadequate for small  and medium-sized basins. This deficiency is partly a result of the historical development of the network, and partly because it is more challenging and costly to monitor flows on small basins and maintain the same data quality.

The national hydrometric network in northern Canada is also inadequate. Part of the problem is that stations are costly to maintain in remote locations. While the network in the north is generally regarded as the weakest part of the network in Canada, it is also under the most threat for a number of reasons:

  • INAC has been reducing its contribution: only 10 percent of stations are now funded by INAC compared with 40 percent previously.
  • When budgets are cut, there is pressure to close stations that are the highest cost to maintain.
  • A number of stations are tied to specific research projects. When those projects end, there is often no funding to continue with monitoring.
  • While the network in the north is sparse, the needs are also sparse.

The National Hydrometric Program has also not been totally able to effectively respond to the increasing needs for hydrologic data. As a result, many organizations collect their own data, which is thereby “lost” for the overall benefit of Canada. Pressures exist to provide increased services to meet the needs of additional Parties and to respond to other data collection requirements. It is important to meet both site-specific needs and broader scientific requirements.

All this poses a risk to the sustainability of a coordinated monitoring program.

Historically, the funding for the National Hydrometric Program was primarily from the federal government. That proportion has gradually declined and the federal government is now contributing less than 50 percent to the program. At the same time, the Parties and clients would like more influence on the overall program management, which is, to an extent, justified by their funding contributions.

In addition to the risks mentioned above, there is a risk caused by the funding mechanism of the program. As per the Agreement on Hydrometric Monitoring, either Party has the authority to modify or terminate their agreements on March 31 of any year, when a 12-month written notice is provided. A reduction in funding might cause the National Hydrometric Program to reduce its workforce.

A number of different ideas were proposed by respondents to make the partnerships more effective and provide more opportunities to tailor the program to clients needs. These ideas include the following:

  • Incorporating other data in the national database: The National Hydrometric Program may consider taking the lead in integrating other data into the national database so that the national database becomes a “network of networks.” The risk of not enabling this is an increased “balkanization” of hydrologic data in Canada, and a less relevant national database. While the other data may not meet the same quality standards as the program’s data, it would still be of considerable value to hydrologists. If all hydrometric stations operating in Canada (operated by municipalities, environmental agencies, consultants, private developers and hydropower utilities) were included, it is estimated that the network could increase in size between 10–30 percent.

  • Integrating the hydrometric network with the climate data network: This integration could also be implemented both for network design and data reporting. As a simple example: for improving hydrological models and understanding the relationship between climate and water, climate stations could be maintained in watersheds where there is a stream gauge.

  • Providing value added services online: The National Hydrometric Program may consider providing value added services online, such as basic data analyses and station information--for instance, rating curves and survey benchmarks. Environment Canada developed the Consolidated Frequency Analysis Program to analyze Water Survey of Canada data for flood frequency. This program is widely used in Canada but needs to be updated with a Windows interface.

  • Reducing the cost of operating stations in remote locations: The National Hydrometric Program may consider examining some of its operations, to reduce costs by lowering data standards for a number of stations in remote locations. Seasonal operation of gauges would be an example where the winter flows would not be measured. Furthermore, in northern Canada, there are many potential stream gauging locations in bedrock sections where the rating curve would be stable and fewer visits to the gauge would be required. An improved operational method of program delivery should also review operational procedures to respond to different types of data needs while also focusing on cost reductions.

  • Continuing to explore modelling: Modelling is currently not accurate enough to replace data collection, and requires good data for model calibration. With changes in watersheds over time, particularly land cover changes, the collection of continuous data will always be required. However, models can be useful as a complement to a data collection program. For example, they can be used to fill missing gaps in data and generating flow estimates on a catchment nearby to a stream gauge. Nevertheless, it is also possible that scientific advances could result in model improvements such that, combined with remote sensing, flow estimates from models in ungauged areas could be improved even without nearby data. It could, however, take many years to obtain that level of modelling capability.

It is clear that the best method for collecting and archiving hydrometric data, and establishing national standards, is via a nationally coordinated program. The National Hydrometric Program has been making great strides in the past several years in improving service delivery to clients. Hydrometric data are available for download with an effective web interface that is efficient for the user community. The number of real-time stations, which are of particular value for flood forecasting and water management, has been increased. This has been achieved in a context of resources restraints and a decrease in the overall number of hydrometric stations. As a result, it is necessary for the program to continue looking for service and technological improvements through innovative, cost effective solutions.


8. The Assistant Deputy Minister, Meteorological Service of Canada, with the collaboration of the NAT and NHPCC, continue looking for service and technological improvements through innovative, cost effective solutions, especially in remote locations, through the NAT strategic planning exercise.

Date modified: