There is an underground hazard on the highway outside Whitehorse.
Unless drastic measures are taken, it is inevitable, reprehensible and inevitable.
If and when it reaches the Alaska Highway, many of the northern communities of Yukon will be cut off from Whitehorse, as would that city’s only connection to neighboring Alaska.
That’s why Fabrice Calmels, chair of research in permafrost and geoscience at the University of Yukon, has marked the area around it with sensors. He hopes to be able to give governments an early warning of danger long before the ground under his feet slips.
The danger is called permafrost recession. It occurs when the permanently frozen layer of soil beneath large areas of Canada north of the 60th parallel begins to melt.
When that layer contains a lot of water in the form of ice—in Canada, this is most often the case—and that ice melts, the ground often can’t support the load on top of it.
When this happens, it can cause roads to break or crumble. It can also cause a sinkhole to open in the middle of a road.
And, when deceleration occurs, there is often a chain reaction – more permafrost is exposed to air, accelerating thawing, and deceleration becomes a runaway process.
To be fair, the danger, located about 34 kilometers west of Whitehorse Airport by road, is greater than that of a turtle.
It has gone 69 meters in the last five years, and is now 37 meters off the road. But recently it has gained momentum. Last summer – fueled partly by two weeks of weather above 30 °C – it has risen to 18 metres.
If it continues at its current average speed – which is not certain – it will cross the path along the highway in two or three years. Yukon’s Department of Highways and Public Works is already looking at ways to mitigate the problem.
They’re looking at everything from circling the highway to finding ways to keep the permafrost cool, to digging to replace and stabilize the ground beneath the highway.
The problem, says Idris Muhammad, manager of design and construction for the department, is that all those options are expensive, part of the cost of trying to build on permafrost.
Calmeles’ sensors are the first phase of an early-warning system for permafrost recession, which he is developing in collaboration with his mentor, Michel Allard, professor emeritus at Laval University, who has access to three airports in remote Nunavik communities in the North. Similar sensors are installed. Quebec for the same purpose.
“Permafrost touches everything,” says Calmeles. “Building on permafrost has always been hard.”
Much of what is built on permafrost today includes measures designed to reduce the effects of that layer’s destabilization. Houses built on permafrost are built above the ground so that air can pass under them, so that the permafrost remains frozen. Roads on permafrost cost five times more to build and maintain than on non-permafrost land because they must compensate for possible ground movement due to changes in the permafrost layer.
But much of that technology is based on a relatively stable climate, Calmels says, one where there is a thermal equilibrium between infrastructure, air temperature, and permafrost temperature.
“But if our (global) temperatures are rising, it becomes even more difficult.”
When the warning system is rolled out, scientists hope they will be able to automatically analyze the data coming from those sensors, predicting whether thawing permafrost could cause a meltdown that threatens those sites. and automatically sends alarms to those who need to know about such things.
Those warnings will come on two levels, says Calmels. The first, detecting small movements and temperature changes in permafrost, will warn officials that a recession is likely in the near future. The second level, a mass movement or loss of sensors, would indicate that a meltdown – or a sinkhole – had just occurred, giving officials the chance to close any affected roads.
Calmels’ sensors, some deployed in boreholes drilled into the permafrost, measure ground temperature at various depths, air temperature, soil moisture and precipitation each hour.
These also include inclinometers, which can be used to generate information about whether the ground is moving, how much and how fast it is moving and about its speed relative to other layers.
His team also conducts deceleration surveys from the air using GPS markers and a drone to track its movement.
All data goes to a data logger, which then sends the information to a nearby substation called a gateway. In the case of an Alaska Highway meltdown, that entrance is at a farmer’s place about a kilometer away.
From the gateway, the information is uploaded via the Internet to a central computer at the University of Laval in Quebec City. There, an algorithm – now in its final stages of development – analyzes the raw data and determines whether there is a risk of turmoil in the near future.
There is evidence that this system works. Over the past year, Calmels placed sensors directly in front of an Alaska Highway meltdown, and using the data he collected, was able to predict the next part of the downturn six days before it actually happened.
Eventually, the entire system will be automated, and if the algorithm determines that an alarm should be raised, it will be sent automatically via an email distribution list that needs to know.
The project, funded by both the federal and regional governments, is being undertaken in collaboration with the University of Laval.
In Quebec, Laval’s Allard has installed similar sensors near airports in three Nunavik communities – Salluit, Tasyuzak and Inukjuaq – to monitor whether any permafrost changes could cause runways to jolt.
Here, Allard is also concerned about melting permafrost causing landslides, a process similar to the meltdown in the Yukon, but occurring on a greater slope.
When the permafrost layer thaws, the ground layer above it, the active layer that freezes perennially and thaws again, can loosen and begin to slide. Scientists call that an “active layer detachment failure.”
“The idea is this: how can we assess the risk of landslides occurring in the future?” He says. “How can we create a warning system that will tell, for example, the mayor, civil defense organizations that, if the weather continues, the risk of landslides will begin.”
Such warnings are extremely important in the far north of Quebec. There are 14 Inuit communities in Nunavik, all of whom rely entirely on their airports for supplies when the weather turns and ice shuts off shipping access.
Once Allard and Calmels’ warning systems are up and running, all three airports will have little advance notice of changes in permafrost that could affect the airport’s runways. Allard says he is already talking with the government about expanding that warning system to another 11 communities.
And it’s not a stretch to believe that the early warning system they developed could be adapted to other, non-permafrost regions.
For example, given adequate study, networks of sensors and automated warning systems could be implemented on slopes in BC to warn officials when a mudslide or rock slide is imminent, and when someone’s Have to send a sweeping alarm. It’s a warning that could be useful during last week’s floods and landslides in southern BC, which killed at least one person and trapped hundreds of people on its highways.
“It’s not that complicated, you know,” Allard says. “The main systems are data loggers and (local) communications. And those systems are widely available, and they can be deployed anywhere in the world.
“Therefore, if some landslide specialist wants to send a signal, the system can be installed on their equipment and put into use.”
Allard expects the automatic warning system to be operational by next summer.