Permafrost in the Northwest Territories

Thawing permafrost, one result of rapid warming in Canada’s Arctic, is causing widespread impacts to northern infrastructure. Using a climate risk analysis protocol (the PIEVC Protocol), a risk assessment was undertaken to better understand the threat from permafrost degradation in the Northwest Territories.

Key Points

  • Climate change induced permafrost thaw will have major socioeconomic impacts on the built environment in Canada’s north.
  • Assessing risks from thawing permafrost at the local scale is difficult due to limited data availability and complex interactions between the different variables that influence permafrost thaw.
  • Design criteria should follow a risk-based approach that incorporates degrees of uncertainty and ranges of data, rather than the traditional engineering approach of adhering to fixed numerical thresholds.

According to a 2019 Environment and Climate Change Canada report, Canada is warming at twice the average global rate, while the Arctic is warming at three times that rate. In Canada’s north, many buildings and other infrastructure are built atop permanently frozen ground that isn’t permanently frozen anymore. A 2017 analysis by the Northwest Territories Association of Communities put the annual cost associated with infrastructural impacts of thawing permafrost at $51 million, in a region with few financial resources compared to southern Canada. In the Yukon, thawing permafrost has compromised the structural performance of Dawson City’s rec centre so severely that it forced the temporary closure of the centre in late 2017, and has created the need for an entirely new building in the near future. Many houses in the north (particularly in Nunavut, where permafrost is essentially continuous throughout the territory) have been built on the assumption that the permafrost underneath them will remain solid – an assumption that can no longer be made.

In response to this threat, in 2019, the Northwest Territories’ Ministry of Municipal and Community Affairs tasked the professional services firm WSP with performing a screening-level risk assessment1 of the hazards posed by a changing climate (including thawing permafrost) across the territory, using the Public Infrastructure Engineering Vulnerability Committee (PIEVC) climate-risk analysis Protocol.

The team of climate scientists and engineers immediately ran into issues accessing the relevant data.

“A common challenge regarding all the permafrost work that we did on buildings in the north, was how it was hard to get good aggregated data on actual permafrost conditions.”

-Elise Paré, WSP

While geoscientists have amassed a great deal of permafrost data, including location-specific data from geotechnical investigations for specific projects, the information has not yet been aggregated at the territorial level to update previous maps of permafrost conditions. These data gaps may be remedied in the coming years by the NSERC PermafrostNet partnership, a network of permafrost researchers currently working to improve understanding of Canadian permafrost and permafrost thaw.

In the meantime, the WSP consultants utilized the available permafrost maps to consider climate change impacts on permafrost on a broad scale rather than at a granular level where they could say, for example, that one particular community might be at higher risk than another 30 km away. This high-level analysis was intended to be an important first step in assessing the risks posed by permafrost thaw, and will allow the territory to effectively focus future efforts. The hope is that with greater availability of localized permafrost data and climate projections, their estimates can be refined at a later date.

The PIEVC Protocol requires identification of thresholds above which there is risk of a negative impact from climate change. First, the climate threshold for permafrost thaw had to be identified. The consultants began with their best available map of permafrost in Canada and then overlaid a second map describing temperature. Roughly speaking, the -1ºC isotherm (a contour line connecting locations with the same temperature) marks an upper threshold for the maintenance of permafrost. If future-projected average annual temperature changes show an increase above that -1ºC threshold, some degree of permafrost degradation can be expected. Data available at can be used to illustrate how both average and minimum temperatures are projected to change across the Northwest Territories, and do indeed show a northward shift of temperatures near the -1ºC threshold (Figure 1).

Figure 1. Mean daily minimum temperature for northern Canada

Mean daily minimum temperature for northern Canada, centred on the Northwest Territories, in the 1990’s (1981-2010).
Projected mean daily minimum temperature for the same region in the 2060’s (2051-2080) under a high emissions scenario.

Daily minimum temperatures near -1ºC (green shades), are not seen in this region in the 1990s, but are projected to be achieved by the 2060s, indicating some degree of permafrost thaw can be expected in these regions. (Note that daily minimum temperatures of -1ºC indicate even warmer daily mean temperatures). Maps of mean and minimum temperatures can be obtained from

While important, considering changing temperature was only the first step of the analysis, since permafrost also responds to ground moisture. Wet and warm conditions, occurring together, can accelerate the thermal breakdown of permafrost. To examine this effect, the WSP consultants seized on work by permafrost scientist Steve Kokelj of the Northwest Territories Geological Survey, that indicated that permafrost slump could occur when daily rainfall exceeded 40 mm.

To apply this rainfall threshold up to the current time horizon, WSP used historical rainfall intensity-duration-frequency (IDF) curves produced by Environment and Climate Change Canada (ECCC). These datasets describe the historical frequency of extreme rainfall of short duration (from five-minutes up to 24 hours) at a particular location. IDF curves are typically used by engineers or water resource managers for flood forecasting and urban drainage design. Historical IDF data are readily available via

The challenge with using IDF curves for this application though, stress the WSP consultants, is that they are based on rate-of-rainfall data in the past, and cannot on their own be used to estimate extreme rainfall in a changing climate. Moreover, IDF curves are available only at selected locations across Canada. WSP therefore made use of the IDF_CC Tool developed at Western University to determine the probability of future rain events surpassing the threshold. This Tool provides IDF curve projections that consider climate change across the whole Canadian territory. “There are limitations associated with the use of this climate change data at such a significant spatial scale, so the Tool outputs should be paired with a sensitivity analysis specific to the area of the project, before being relied upon for design purposes,” Paré adds. Despite these limitations, however, they are a good starting point. “This is what we had to use for projecting sub-hourly extreme storm events, which are used in building design criteria,” explains her colleague, Jean-Phillipe Martin.

Ultimately, the consultants concluded that while increasing temperature is the primary driver of permafrost thaw, increasing precipitation and precipitation extremes could accelerate degradation, particularly where heat is transmitted to ice-rich permafrost via rain, causing thawing through a process called thermo-erosion. It was further concluded that greater permafrost thaw hazard is due to higher incidence of ice-rich permafrost and fine grained sediments. Thawing of ice-rich permafrost can cause abrupt and dramatic changes to the landscape, such as slumping, sinkholes, landslides, and the sudden expansion or drainage of water bodies.

For this reason, WSP concluded that permafrost thaw was a major concern for buildings in the northern Sahtu and Beaufort-Delta regions of the territory, where the permafrost is ice-rich. This is of greatest importance where buildings contain critical services for vulnerable populations such as hospitals, schools, and community centres, and the team identified those as most at risk.

Experienced in applying climate science, the WSP consultants also understood the importance of viewing climate projections as ranges of possible outcomes, due to the inherent uncertainty in future socioeconomic development and greenhouse gas emissions pathways. Dealing head-on with the uncertainty in future climate projections is something that both Paré and Martin say needs to be more widely adopted across the engineering and planning community with respect to climate risk.

“The most important thing about any of these indicators is that they allow us to initiate a discussion about changing our paradigm in designing for future climate,” says Paré.

“Engineers tend to want a specific number, such as those provided in standards and codes, where the risk of uncertainty has been managed collectively by others,” she explains. When integrating climate projections into design criteria, a risk management approach should be taken. The changing climate has necessitated working with degrees of uncertainty and ranges of data rather than with precise data that are assumed to be certain. In this respect, changes in the approach and language of experts are necessary. The products developed by can help to hasten this evolution.

“You should not just have just one number. You should actually be thinking: ‘What are the consequences of failure? How badly will it go if I get it wrong? This is way more important than being able to say whether it will be a 20% or 22% increase.” This kind of thinking, Paré concludes, is how we can start to embrace a risk-based approach to designing buildings in the rapidly changing North.


  1. WSP Canada (2021). Government of Northwest Territories Assessment of Climate Change Impacts on Infrastructure in All NWT Communities. Online:

Related Variables

Explore variables to learn about how data was used to impact climate related decisions in specific contexts.

Mean temperature describes the average temperature for the 24-hour day.

The average temperature is an environmental indicator with many applications in agriculture, engineering, health, energy management, recreation, and more.

Technical description:

The average of the daily maximum temperature (Tmax) and the daily minimum temperature (Tmin). Use the Variable menu option to view annual, monthly or seasonal values for this variable.

Intensity Duration Frequency (IDF) curves relate short-duration rainfall intensity with its frequency of occurrence and are often used for flood forecasting and urban drainage design.

As most people have experienced, a very intense storm can bring lots of rain over a very short period of time, overwhelming storm drains, flooding basements, and washing out bridges and roads. Therefore, when engineers and hydrologists are designing safe and reliable infrastructure, they need to know how often these damaging rainstorms occur.

One way to determine this is to use Historical Rainfall Intensity-Duration-Frequency (IDF) Curves. Historical IDF Curves show the frequency (probability of occurrence) of extreme rainfall rates persisting for short periods of time (5, 10, 15, 30, 60 min; 2, 6, 12, and 24 h) based on historical rainfall data recorded at a specific weather station. They do not reflect rainfall amounts or rates over wider areas such as entire cities or river basins. The curves are produced to support the development of infrastructure capable of handling extreme rainfall events thereby reducing the risk of flooding and the associated damage to property, and risk to people. IDF Curves are most often used by professional groups such as: engineers, water resource managers, urban and regional planners.

Use with caution: Historical IDF Curves alone cannot be used to assess future extreme rainfall. Since IDF Curves are based on the analysis of historical rate-of-rainfall data, they do not explicitly incorporate any projected future trends due to a changing climate. 

Due to the sparsity of sub-daily hourly or minute-by-minute rainfall observations and the high levels of uncertainty associated with future projections of extreme rainfall at a specific location and over short time periods high spatial and temporal resolution, Environment and Climate Change Canada does not produce future IDF curves.

For further technical information on how IDF Curves are produced, please refer to Environment and Climate Change Canada’s Engineering Climate Datasets page or contact the Engineering Climate Services Unit at [email protected].