Duane Froese of Canada’s University of Alberta and colleagues looked at ice wedges – structures that indicate the presence of premafrost – in the central Yukon Territory. The wedges were a few metres under the surface, beneath a layer of volcanic ash that the team dated at roughly 700,000 years old.

"Some of the recent modeling work on the future of permafrost and some earlier research had suggested that 'old' permafrost was unlikely to exist," Froese told environmentalresearchweb. "It seemed like a challenge to see how far back it might go. I suspect in this region, we’re near that limit."

The news offers a faint glimmer of hope for the future fate of permafrost. The great age of the ice means it survived interglacial periods around 120,000 and 400,000 years ago when the climate is thought to have been warmer than today. Scientists had believed that permafrost melted completely from the interior of Yukon and Alaska about 120,000 years ago.

"Our results don’t change the fact that future warming will lead to release of greenhouse gases from thawing permafrost," said Froese, "but based on the incredible antiquity of the ice wedges we documented, we suggest that permafrost that is more than several metres below the surface is more resilient to climate warming than previously thought."

Froese and colleagues from the University of Toronto and Geological Survey of Canada did their survey within the discontinuous permafrost zone, where permafrost is relatively warm – within a few degrees of 0°C – and only a few to tens of metres thick.

Ice wedges form if the ground cracks in winter and water fills the space the following spring. Over a period of years, this results in distinctively-shaped flat-topped ice wedges beneath the depth of summer melt.

"There is, of course, strong interest in the carbon stores of permafrost across the Arctic which this region, and these deposits represent a significant component which we show is probably stable over the coming decades," said Froese.

The loss of permafrost across the Arctic region also has implications for infrastructure, ecosystems and economic development.

Writing in Science, the researchers say their findings suggest that numberical models of permafrost degradation still do not adequately consider controls on permafrost thickness and distribution such as surface cover, deeper ground ice and excess ice.

Now the researchers plan to keep working on the interglacial record to get a better perspective on what impact past climate had on Arctic and Subarctic landscapes and ecosystems. "One of our major foci at present is the last interglaciation, about 120,000 years ago, when climate was a few degrees warmer than present, sealevel was higher, and likely Arctic sea ice was more restricted," said Froese. "We would like to know more, in a regional sense what happened to permafrost, and place stronger constraints on the magnitude of melt. Stay tuned."