Rather than searching for weird weather or enemy missiles, some satellites are helping researchers to track—and predict—the spread of deadly diseases.

With the pandemic spread of H1N1 swine flu and the continued advance of the H5N1 avian flu, scientists are anxious to better predict the spread of infectious diseases and are looking for new tools wherever they might be—even if that's hundreds of miles in the sky.

"Ideally we could predict conditions that would result in some of these major outbreaks of cholera, malaria, even avian flu," says Tim Ford of the University of New England in Biddeford, Maine. Ford and a group of experts have co-authored a perspective paper (pdf), due out next month in Emerging Infectious Diseases, that proposes making use of environmental data—tracked via satellite—to predict disease outbreaks.

"As climate changes, and even with many of our weather patterns, [it] directly affects the distribution of disease," Ford says. Hantavirus, the pulmonary disease spread by rodents, for example, has been linked to changes in precipitation. With more rainfall, vegetation increases, which then fuels rodent populations. And pinpointing an area as relevant conditions emerge—before an outbreak starts—buys precious time to spread public health messages.

Satellite imaging can also help warn of cholera outbreaks, which are predicted to worsen with climate change. The satellites provide information about water surface temperatures, which are key to the spread of this waterborne disease. One study found that giving people simple preventative instructions, such as filtering water through a sari cloth, reduced cholera-related deaths by an estimated 50 percent in some areas.

Remote data have already been used to map the avian flu in Asia. Xiangming Xiao, associate director of the University of Oklahoma's Center for Spatial Analysis in Norman, has been tracking likely outbreaks of this highly pathogenic flu by looking for key habitat and weather changes. The domestic duck—determined to be the main carrier of the disease—is a common inhabitant of Southeast Asia's rice paddies, and the movement of migratory birds—a secondary carrier—could be predicted based on temperatures. So using both land-use and temperature information from satellites, Xiao and his team could track the spread of the flu by estimating where the birds would be.

If visual data from satellites is combined with information from radar and LiDAR, (light detecting and ranging, which provides laser-measured data about 3-D contours), Xiao explains, researchers can really hone prediction of some diseases down to a tree line. "You can look at… the transition of pasture grassland to forests," he says, habitats which determine the range of deer. "And this has very important implications for tick-borne diseases, like Lyme disease."

Much of the satellite work, however, still relies on clear skies. And all of it has been dependent on quality information from willing providers, such as NASA and its Earth Observing System, the availability of which researchers hope will continue in the future.

Even with the clearest NASA images, though, current methods are far from perfect. They employ complex models and incomplete information, risking false alarms and missed outbreaks.

The satellite data are still just a portion of the equation. They allow researchers to start "standing back and looking at the picture from a distance," Ford says. He and others are heavily reliant on ground-based measurements and observations. Xiao notes that, "the in situ observations are still very, very important. So the key is to combine those together—that's a real challenge."

To make the predictions as precise as possible takes understanding the ecology not just of the place being studied, but also of the disease and the human population. "You see tremendous variations in different areas," says Ford of how diseases behave, and "in some sense, [that is due to] just difference in human behavior."

Judging the severity of avian flu's spread from satellite imaging, for instance, requires knowing how likely certain areas are to keep domestic chickens and ducks—a practice more common in countries that consume more poultry, Xiao explains. And getting precise poultry production statistics can be a real challenge, he notes, as record-keeping can vary greatly among countries and regions.

But Ford thinks that even with these limitations, "There's no reason at all we shouldn't be able to say, 'This summer is going to be a bad hantavirus year' or 'This season will likely have a high cholera risk.'"

Novel or long-dormant diseases present more challenges for remote prediction. "Whether we can predict emerging diseases is a whole other question," Ford says, especially as their vectors or risk factors might take time to assess. And some diseases that spread among people might turn out to be nearly impossible to predict using satellite and environmental data beyond what researchers already know about seasonal cycles, like that for the seasonal flu. And, the nonseasonal H1N1 flu, for example, "is probably going to be more to do with human patterns [and] rapid transport between countries" than environmental changes that can be mapped, Ford says.

Predicting infectious diseases is a crucial step in curbing them, Ford notes. "With all our medical advances and our advances in sanitation…we still have not been able to grapple with diseases," he says. But he is hopeful for the future of satellite-based prediction—even as it becomes a greater necessity in a changing climate and globalized world. "There's really nowhere on the globe that a pathogen can really remain isolated," he says.