<<返回上一页

Fossils leaves reveal climate model errors

发布时间:2019-03-06 06:05:01来源:未知点击:

By Nicola Jones Analysis of fossils leaves have shown that the standard models used for climate prediction have huge errors when taken out of familiar conditions, say an international team of scientists. The team, involving researchers from the UK, Russia, Sweden and the Czech Republic, used the leaf fossils to calculate the temperature at various sites during the Late Cretaceous period, 95 million years ago. They then used some of the best climate models researchers have on offer, such as the Hadley Centre’s climate model, to calculate what temperatures at that time might have been like. To their surprise, they found their results differed greatly. Most striking were results in regions far away from the coast, where leaf-based results were much warmer. “We’re talking about an error on the order of 20 °C, so it’s not small – not by any means,” says modeller Paul Valdes from the University of Reading. Catherine Senior from the Hadley Centre says attempts to model the Cretaceous climate are useful. “Seeing these kinds of anomalies means we have to go back and see where things went wrong,” she says. The Cretaceous period is of particular interest because volcanoes at the time spewed huge amounts of carbon dioxide into the atmosphere, warming the planet up much in the same way as it is warming today. And if our models are seriously wrong for the past, Valdes says, then they could be equally wrong about the future. “It could be out easily by 5 °C,” says Valdes. The research team began their project by trying to match climate to leaf characteristics, including the jaggedness of the leaf edges, leaf size and shape. They found that sharp teeth, for example, were favoured in cooler, drier climates, because the surface area of the edge is important for water and gas exchange. By looking at 31 such characteristics of hundreds of modern day samples from around the world, they found they could calculate factors like the lowest temperature of the year with a surprisingly small error – just 5 °C. The team then collected thousands of leaf fossils from Czechoslovakia, Kazakhstan, Russia, and Alaska from the Late Cretaceous period, and calculated what the temperature would have been. In coastal regions, the answers they got from their model matched temperatures obtained from oxygen isotope measurements in ocean sediments. But in most of these sites, the temperatures did not match results from climate models. The largest difference was for the mid-continental site of the Vilui basin, where the leaves indicated that the coldest yearly temperature was a balmy 5 to 15 °C. The Hadley Centre’s climate model showed the Vilui basin plunged into freezing temperatures of 0 to minus 15 °C. “It’s a generic problem with all climate models,” says Valdes about the discrepancy. “No one has managed to get a warm winter in Russia in the Cretaceous.” He thinks it probably has to do with the models’ inability to represent cloud cover accurately. “If we had a lot more cloud at that time, it would solve the problem,” he says. Mark Eakin, chief of the NOAA Paleoclimatology program in Boulder Colorado, says that throwing drastically different conditions at these programs is one of “the best ways to test the limits of the models”. But he cautions that the conditions they put into the Hadley model, such as the amount of carbon dioxide in the air and the position of the continents, must have been very rough approximations. The models, he adds, are probably better suited to handling situations 100 years in the future,