Add space satellites and supercomputers to the list of birdwatching tools.

Scientists at Oak Ridge National Laboratory and Cornell University's Lab of Ornithology are combining those high-tech tools with a database of bird sightings contributed by birdwatchers to learn how climate change is affecting bird movement in the United States.

"The approach we're taking here is we're trying to bring together as much environmental data as we can to try to understand what influences the bird migration," said Bob Cook, a distinguished research scientist at ORNL involved with the effort. "We're trying to address a really important question with regard to climate change: How might climate change influence the migration patterns of birds?"

That includes information about rainfall, temperature and snow cover, as well as the start of spring greening and the composition of land cover—forested, urbanized or grassland, for example.

The land cover information is drawn from a NASA satellite sensor, MODIS—that's short for "Moderate Resolution Imaging Spectroradiometer". Bird sightings are taken from an online database run by Cornell and the National Audubon Society. Launched in 2003, E-bird allows "citizen scientists" to submit detailed reports via an Internet checklist.

Combining and analyzing all that data will require computing might provided by TeraGrid, a National Science Foundation-administered network of supercomputers.

Steve Kelling, director of information science at Cornell's ornithology lab, said the new project will allow scientists to link bird sightings to climate conditions.

Via the seven-year-old E-bird database—which accepts observations recorded a century ago, as well as present-day bird sightings—"we have really good information on the location where observations were made," Kelling said. "We can link those with other kinds of environmental observations, like land cover, type of climate, temperature, elevation and human demographic information."

Adding in the MODIS satellite data provides information about when spring greening begins and when fall starts, he said, two things that seem to be important environmental cues for bird migration.

Potential for a fatal mismatch

Eventually, the scientists would like to develop models that can forecast how future climate shifts might affect bird populations.

"We'd like to be able to shift the greening index to occur two weeks earlier or two weeks later and see if that influences the model's predictions of when birds will arrive at certain latitudes," Kelling said.

Climate change could produce a mismatch between a bird species' cue to migrate or nest and the availability of food, he noted, a phenomenon that's been observed with some species in Europe. For example, if the American Robin miscalculates spring and arrives before the insects it eats are ready, the birds could starve.

Several recent reports—including two by the Interior Department and one from the National Audubon Society—have found evidence that climate change is already altering bird habitat and migration patterns in the United States. Kelling said the advantage of the new project is harnessing the power of E-bird.

It's the only dataset that gives information about patterns of bird movement throughout the year, he said, noting that many other studies have relied on data collected through the Audubon's annual Christmas Bird Count or similar events.

In contrast, birders submitted 11 million individual bird sightings to E-bird between January 1 and July 31 of this year.

"It's just an immense amount of information," said Cook, whose work on the bird project is a proof of concept for a larger effort he's helping to direct, the Data Observation Network for Earth.

The five-year, NSF-funded program aims to help research scientists find new ways to visualize and explore large amounts of information.