Everyone knows how hard it is to predict what the weather will be like from one day to the next. So how does one go about figuring out what the weather was like tens of thousands of years ago?

In April a team of scientists from the University of New Brunswick will spend time searching in what might seem a most unlikely place; the mud in the bottom of lakes in the Beaver Creek area.

They hope that the midges, diatoms and plant pollen they find there will tell them what the climate was like during the last Ice Age in the unglaciated land known as Beringia. And figuring that out might also shed some light on why Beringian mammals such as mastodons, mammoths and long-horned bison started to go extinct after the end of the Ice Age.

"You have to ask what the environment was like for that sort of odd community to have ever existed in the first place," says Les Cwynar, a biology professor at the University of New Brunswick who is the team leader for this project.

Cwynar and other researchers want to figure out exactly when the climate warmed after the Pleistocene Ice Age, and how quickly that warming took place. "We know that the climate changed, but what was the magnitude of the change. We have not been able to put any numbers on that."

This research might also throw light on a more contemporary issue. Climate models predict that Arctic and boreal regions will be affected the most by global climate change, so establishing how climate change proceeded in the past could increase our understanding of what is happening now.

Fortunately for the researchers, midges and diatoms did not go extinct when the climate warmed the last time around. Instead these organisms adapted to changes in temperature, increasing and decreasing their numbers as the climate changed.

Scientists have mucked about in lake bottoms for decades, looking for clues to climate change, but Cwynar's team has a major advantage over previous efforts. Ian Walker, one of the team members, has developed a mathematical model that relates the abundance of midges and diatoms to water temperatures.

"We establish the modern relationship and then we make the assumption that this relationship has been constant through time," says Cwynar. "Then we can plug the number of midges or diatoms into the equation and figure out what past temperatures were."

Once they have the modern relationship settled, then they will go after the midges that lived at the same time as the mastodons. They will use a hand-cranked drill to push a tube through four- to ten- metres of lake-bottom sediments, pulling out a core full of mud and gunk that could be as much as 30,000 years old.

Cwynar and his team will investigate lakes in the northern Yukon and Alaska as well as around Beaver Creek. They hope eventually to produce temperature curves for Alaska and the Yukon for the period from 10,000 to 30,000 years ago.

The Canadian researchers will also collaborate with a group of American scientists who are working on changes in lake levels.

"They are interested in finding out how the moisture balance has changed and that is very difficult to answer," says Cwynar.

If the amount of precipitation was the only factor, it would be easy enough to find the link between changing water levels and climate. But air temperature also plays a role as it affects the amount of water that evaporates from the lake.

If Cwynar's team can establish what the temperatures were, then it will be easier to tease out whether the climate was wetter or drier at a particular water level.

Cwynar has received a million-dollar grant for his research, which will be spread over five years. For more information he can be contacted at (506) 452-6197 or by e-mail at cwynar@unb.ca.