Reliance on just one model no better than flipping a coin!
A new study published in the journal Global Ecology and Biogeography represents the first real test of the performance of models used to forecast how species will change their geographic ranges in response to the Earth's changing climate
Despite the weight of scientific evidence that the Earth is warming and that this is already affecting wildlife, many people - and a few scientists - still refuse to believe it is actually happening. These climate change skeptics usually justify their position by insisting that scientists' forecasts are just too inaccurate. Of course, we can never really know what the future will bring, but in a fascinating new study published this week in the journal Global Ecology and Biogeography a group of Oxford Scientists have tested the ability of environmental science to predict the future… by going back to the past.
Dr Miguel Araújo and his colleagues from Oxford University's Biodiversity Research Group imagined they were back in the 70's and were trying to predict the geographic ranges of British birds in 1991 using 16 commonly used climate-envelope models and the real data on how the climate had changed during this period.
Climate envelope model forecasts typically involve a three-step process: First, for each species, mathematical models are developed to link the species to its present climate envelope (actual environmental conditions where the species is found). Second, a climate change scenario for some point in the future, typically 2020 or 2050, is applied to generate a new potential range distribution for the species. Third, this new projected distribution is compared to the present distribution, allowing the scientists to forecast whether the species distribution, will grow, or shrink, or even become extinct.
Unlike previous studies that have provided untestable forecasts of range changes in response to future climate change, the Oxford study was able to directly compare the predicted range changes with what actually occurred. Surprisingly, the ability of any single model to accurately predict the 1991 distribution was very poor. The results of models applied to particular species were spectacularly variable. For 90% of species the models could not agree whether their geographic range would expand or contract. In the small minority of cases (10%) where all the models agreed about the direction of change, they only had a 50% chance of getting that direction right. "It would be just as accurate and a lot less hassle just to toss a coin" says one of the co-authors, Dr Richard Ladle.
So, will we ever be able to predict accurately how climate change will affect the distributions of animals and plants? The Oxford Group may have found a solution. "The accuracy of the predictions can be drastically increased if a set of alternative models are compared and used together to create a 'consensus' projection" Says Dr Araújo. Using the same data set for British birds, the consensus prediction was shown to be vastly superior to any single model and could predict bird range expansion or contraction with an accuracy of over 75%.
To avoid further accusations of crystal ball gazing, environmentalists and scientists now need to find further ways of improving the accuracy of models to provide more meaningful inputs into environmental policy making. "If we don't improve our forecasting soon then not only will the climate skeptics find it easy to criticize climate change research, but we will be left making decisions about the future of the planet based on guesswork" says Dr Ladle.
Source: Eurekalert & othersLast reviewed: By John M. Grohol, Psy.D. on 21 Feb 2009
Published on PsychCentral.com. All rights reserved.
A Freudian slip when you say one thing mean your mother.
-- Author unknown