Forest fires that burned themselves out almost 40 years ago are still reshaping the terrain of the Takhini valley, west of Whitehorse. Over the years since the 1958 fires, the ground has subsided in places, creating little lakes about two metres deep.

"Around some of the lakes, you can see trees toppling inward toward the lake," says Chris Burn, Associate Professor of Geography at Carleton University, Ottawa. Burn has been monitoring soil and permafrost temperatures in the Takhini valley since 1982.

The small lakes in the Takhini valley are created by ground subsidence that results from melting of the permafrost beneath the ground surface. Much of the subsidence is a direct consequence of those long-ago fires.

Permafrost, permanently frozen soil, underlies much of the Yukon. In the Takhini valley, about 20 percent of the ground is underlain by permafrost. Above the permafrost is a layer of soil, called the active layer, that thaws in the summer and freezes again in the winter. It is this layer that supports vegetation. The thickness of both the active layer and the permafrost vary, depending on the location and local conditions.

In the Takhini valley, Burn says, the permafrost is normally about 16 metres thick and the active layer is normally one and a half metres thick. However, in parts of the burned area, the active layer is now more than three and a half metres thick.

Some melting of permafrost is common after a forest fire, says Burn. The surface layer of vegetation in a forest provides insulation that keeps the ground cold. Evaporation from vegetation also reduces the ground temperature. When the vegetation is burned off, more of the sun's warmth penetrates into the earth, raising the ground temperature.

Even a small change in ground temperature is enough to start melting the permafrost, Burn says, and the effects of the 1958 fire have caused the soil temperature to increase approximately three degrees.

After most forest fires, vegetation returns, the ground then cools, and the permafrost is re-established at its old level in a decade or so. In the Takhini valley, the vegetation succession is slow, partly because the area is very dry and partly because the 1958 fire was so intense.

The result has been surface subsidence, as the ice in the permafrost that supported the active layer melts. The ground slumps, creating a hollow in which water accumulates, forming small, shallow lakes. Trees may tilt toward the lakes as the ground subsides, and some of them are eventually swamped.

But most of the lakes aren't permanent. In 1982, when he began monitoring the area, many of the depressions contained water, Burn says. From 1982 to 1990 there was little change in water level, but since 1990 water levels in many lakes have dropped. Some are now dry.

"It's difficult, when you look at the climate records, to suggest this is due to changes in precipitation or temperature," Burn says. He suggests that slowly returning ground-level vegetation is the cause. Plants take up water through their roots and evaporate it through their leaves, reducing the amount of water in the active layer above the permafrost. That means there is less water available to recharge the lakes.

"That's one explanation, and to me it's the most satisfactory," says Burn.

While permafrost in the area of the Takhini burn is still melting, it won't disappear, the researcher says.

"It would take nearly 1300 years to melt 16 metres of permafrost under the present conditions."

The forest is gradually regenerating itself, restoring the insulating layer of vegetation. Long before the 1300 years are up, the soil temperature will likely drop again, the active layer will shrink, and permafrost will stabilize.

Meanwhile, the Takhini burn is providing useful information for researchers. The changes in ground temperature that affect permafrost are slow, and very few studies have recorded ground temperature data over a long enough period to reveal patterns in the temperature fluctuations.

"The Takhini record is one of the longest records of forest fire recovery in North America," says Burn. The information gathered in the Takhini valley will contribute to our understanding of the dynamics of permafrost across the north.

For more information about permafrost, contact Chris Burn at the Department of Geography, Carleton University, Ottawa, Ontario, K1S 5B6.