Aerial War against Disease

Satellite tracking of epidemics is soaring

Imagine this: It’s the height of summer a decade from now, and stifling heat blankets New York City. A hidden foe has lurked out of sight for many months, and authorities are on high alert. The streets are deserted, Central Park is sealed, helicopters circle overhead, and residents remain inside with their windows firmly closed. Dousing the city, the surrounding suburbs, and perhaps wider areas with protective chemicals is the only defense.

SATELLITE FEVER. INTREPID’s preliminary risk map for West Nile fever. Land-surface temperatures (color variation) and locations of birds harboring West Nile virus (black dots) help researchers predict where outbreaks are likely to occur. TALA Research Group, University of Oxford.
SNAPPING SNAILS. Satellite image of the mountainous Anning River Valley in Sichuan, China, where schistosomiasis is endemic. Ditches that line agricultural fields (shown green in the image) often contain potential snail habitat. E. Seto
POORLY POSSUM. Inquisitive deer and cattle appear to contract the disease via contact with sick possums. J. McKenzie

In this fictional scenario, the city isn’t under terrorist attack. The concealed enemy is the deadly virus that causes West Nile fever, which first emerged in New York in 1999. Attempts at vaccines (SN: 3/16/02, p. 164: Clever Combo: Hybrid vaccine prevents West Nile virus in mice.) haven’t borne fruit, and the heat is creating the ideal conditions for drawing virus-carrying mosquitoes out of their winter refuges.

However, officials of the future are prepared to fight the disease, courtesy of an unlikely medical tool: the National Oceanic and Atmospheric Administration’s (NOAA) series of polar-orbiting satellites about 850 kilometers above Earth.

The NOAA satellites aren’t sending pictures of crows–the most common avian reservoir of West Nile virus in the United States–but they’re producing maps of vegetation and weather affecting the mosquitoes that carry the disease to people.

In 2002, satellites already enable people to forecast weather, spy on enemy armies, broadcast television signals, and phone a friend in the Amazon jungle.

A growing group of researchers has also been harnessing satellites in the battle against infectious diseases (SN: 8/2/97, p. 72; 2/2/02, p. 78: An El Niño link with a tropical disease?).

The amount of data created by environmental satellites is increasing at an exponential rate, and researchers worldwide are catching on to the value of this information. The February Photogrammetric Engineering and Remote Sensing details several research projects that use satellite data to reveal patterns of diseases including schistosomiasis, tuberculosis, Ebola, Rift Valley fever, and West Nile fever.

Follow the flora

The rationale behind the use of satellites to fight vector-borne disease lies in the field of landscape epidemiology, which began in Russia in the 1930s.

Theory holds that features of the landscape, such as temperature, humidity, rainfall, and plant growth, can help public health officials pinpoint where deadly microbes–or the animals that carry them–dwell. Ideally, that indicates where people are at greatest risk and so where governments should target disease-control measures and medicines.

Researchers first discussed adding satellites to the tools of landscape epidemiology in the early 1970s, with the launch of NASA’s Landsat-1 satellite, says Robert Venezia, program manager for public health applications at NASA’s headquarters. In recent years, NASA has promoted the use of satellite data for health purposes. The agency is currently working with an international team of universities and U.S.-government bodies under the banner of the International Research Partnership for Infectious Diseases, or INTREPID. The program’s researchers are today in the early stages of creating a satellite-borne early-warning system for West Nile virus in the United States, says epidemiologist Simon Hay at the University of Oxford in England.

With it, workers will be better prepared, as in the fictional example above.

The scientists are focusing on areas with warm land surfaces, high humidity, and vegetation types likely to harbor mosquitoes. Data on these variables will be combined with other information, such as bird-migration routes, to create risk maps for West Nile fever during each summer.

Two other deadly diseases–Rift Valley fever and Ebola–are also the focus of remote-sensing initiatives. These viral infections, which cause swelling and fever, strike in intermittent outbreaks in Africa. Rift Valley fever was first documented in the 1930s in the area of Kenya from which it takes its name. It kills few people, but it’s a major problem in livestock, where it causes spontaneous abortions and many deaths.

Ebola is more deadly to people and more enigmatic. Named after a river in Sudan, it was reported appearing for the first time in 1976. Reappearing periodically, it spreads rapidly and kills up to 88 percent of the people it infects.

Despite the disease’s deadliness, scientific understanding of Ebola is limited. Infections first show up in people who have been in tropical forests, but researchers still have no idea which animal species act as vectors or reservoirs for the virus.

Scientists at the NASA-Goddard Space Flight Center in Greenbelt, Md., are approaching the challenges of Ebola and Rift Valley fever in a different way from the strategy employed to study U.S. West Nile virus. They’re not only trying to predict future outbreaks of the African diseases but also attempting to better understand their causes.

The only documented Ebola outbreaks have occurred within two limited periods. The disease struck in Congo and Sudan during the 1970s, before satellite data were being collected consistently in Africa, and resurfaced for a further sporadic round of infection in Gabon, Ivory Coast, and Congo during the 1990s.

Compton J. Tucker of Goddard and his colleagues suspect that transitions between wet and dry periods in tropical forests may lie at the heart of Ebola outbreaks. Using images taken by the NOAA and Landsat satellites from 1994 to 1996, the team has studied vegetation changes as a surrogate for rainfall patterns.

That’s relatively easy to do from space. Chlorophyll–the green pigment in leaves–reflects light in a distinctive way, so satellite images clearly reveal vegetation. The researchers found that images taken during three separate Ebola outbreaks indicated a transition from below-average to above-average vegetation cover within equatorial Africa’s forests. The finding suggests that the disease “may follow when a rare tropical dry period is brought to an abrupt end with a change to very wet conditions,” says Tucker.

He notes that it will take additional research to confirm that a wet-to-dry transition is the trigger and to explain why such an event spreads the virus.

“It’s fortunate for those affected by Ebola that we have so few outbreaks to study,” Tucker says, “but it makes our job more difficult.”

Creature features

The researchers have had more luck nailing down the conditions that promoted Rift Valley fever during the 1980s and 1990s.

From the satellite data, Rift Valley fever appears linked to above-average rainfall in savanna ecosystems. The wet conditions lead to booms in mosquito populations that spread Rift Valley fever to people.

The accuracy with which the researchers can use satellite data to predict the disease’s course has already led the team to create a disease-surveillance and early-warning system. Computers analyze daily satellite data for the telltale predictors of mosquito outbreaks in areas prone to the disease, says Tucker.

The researchers pass along 2-week summaries of the data to the U.S. Army and the World Health Organization. These groups can warn their personnel and others in threatened areas to take precautions against infection, for instance, by stepping up immunizations. Satellite surveillance programs can now “monitor diseases in both space and time,” Hay says.

So far, satellite epidemiology has focused mostly on insect-borne diseases because correlations among weather, plant cover, and insect infestations are clearly established. But scientists are starting to use NASA’s Landsat satellites–700 km above Earth–to track other disease carriers.

In China, researchers are using satellites to map areas at highest risk from the debilitating disease schistosomiasis. Caused by a parasitic flatworm that spends part of its life in amphibious snails, the illness afflicts 200 million people worldwide. The parasite, Schistosoma japonica, burrows into the skin of people who venture into contaminated water. Inside a victim, it migrates to internal organs, such as the lungs or liver, causing serious long-term illness.

In the mid-1990s, Edmund Seto of the University of California, Berkeley was approached by colleagues from China to help find high-risk areas and monitor the spread of the disease as the environment changes during the country’s rapid development.

Seto’s team of researchers turned to data from the Landsat TM satellite to identify areas most likely to harbor snails. In China, the snail host is divided into two subspecies, one that lives in certain types of mountainous irrigated farmland and the other that favors marshy floodplains of the lower Yangtze River. Both snails are susceptible to environmental change and tightly linked to their favored habitats.

The researchers selected sites in the two habitats and then compared painstakingly collected field data with satellite images of the regions taken with seven visible-light and infrared wavelengths. Different combinations are useful for classifying different types of vegetation and land cover. The team noted whether each 30-meter-by-30-meter pixel in the images matched snail habitat, as determined by the field data.

As the data accumulates, computer programs may be able to recognize snail habitat–and thus, areas of high disease risk–in future images of the same or other areas of China, so health officials can pinpoint risky areas without sending researchers into the field.

As China undergoes major environmental changes, the technology will enable scientists to monitor potential snail habitat. Seto cites the example of the enormous Three Gorges Dam being constructed on the Yangtze River. To be completed in 2003, the dam will create a reservoir covering 395 square miles, alter the course of rivers and streams, and perhaps create large new tracts of snail habitat. “The impacts are huge,” says Seto.

Possum possibilities

In New Zealand, veterinary researchers are using remote sensing to spy on a disease culprit that’s larger and fluffier than snails or mosquitoes. The brushtail possum carries the bacterium responsible for bovine tuberculosis (TB).

Mycobacterium bovis, a close relative of the germ responsible for tuberculosis in people, preferentially afflicts cattle. It can, however, infect a person who drinks tainted milk, as it commonly did before pasteurization.

Bovine TB still strikes dairy and beef herds, causes major reductions in milk yields, and makes livestock unfit for export to many countries. The disease also continues to pose a threat to farm and slaughterhouse workers.

Eradicating the disease in cattle wouldn’t be a problem if possums didn’t complicate matters, says Joanna McKenzie, of Massey University in Palmerston North, New Zealand. These marsupials are also susceptible to M. bovis.

The brushtail possum is a relentless marauder that invaded New Zealand from Australia. Listed among the world’s 100 most-invasive species, it is one of New Zealand’s worst pests. It costs the government there $13 million a year in pest-control expenses.

The otherwise nocturnal mammal becomes disorientated during its dying throes from bovine TB, and it then wanders into pastures during daylight, says McKenzie. Inquisitive deer and cattle often come to investigate, sniffing and licking the afflicted animals and, in the process, getting a strong draught of TB bacteria.

The traditional approach to fighting bovine TB in possums is extensive culling of both the healthy and ill animals in wide areas of infection. McKenzie’s group hopes to use remote sensing to focus efforts, which would reduce costs and bring hope of eradicating the disease from the country.

The researchers studied a 60-square-kilometer area of New Zealand’s North Island. The location has rolling hills, thousands of cattle, and numerous pockets of bovine-TB-infected possums. McKenzie and her colleagues used images from the French SPOT3 satellite to map out vegetation types for each 20-square-meter piece of the overall area. Vegetation type determines food sources and suitable den sites for the possums, says McKenzie.

Because the animals prefer to make dens in large trees or steep slopes, the researchers also took into account topographic data indicating the slope of each pixel. Overlaid with a map of farms within the area, the data enabled McKenzie’s team to predict infected possum hot spots.

To test the accuracy of the approach, the researchers and agricultural authorities focused annual possum-eradication campaigns on 30 farms that satellite data had identified as possum hot spots. From 1997 through 1999, the researchers directed intensive trapping and killing programs on those areas, and during the seasons of high risk, farmers kept their grazing cattle out of pastures where they would be likely to encounter sick possums.

After the 3-year trial, the incidence of TB in cattle in the test area was significantly lower than among livestock on a similar group of farms receiving traditional, less-targeted efforts to control possums.

High-tech intelligence

Researchers’ interest in satellite mapping of infectious diseases has been building steadily, says Hay. “The new generation of satellite sensors is also continually opening up new opportunities,” he adds.

One recent advance is the development of radar satellites. Traditionally, satellites rely on passive detection of reflected energy, such as visible light or infrared radiation. Radar satellites provide similar information but generate their own signals and then receive the reflections. So, they can pick up images night and day. The wavelengths that the radar satellites emit also enable them to see through clouds.

Researchers have been finding it tough to keep abreast of technological advances in recent years, with new satellites improving the scope and quality of data. At any given time, NASA has in store about 1 petabyte of data from all its Earth-observing satellites, says Venezia. This breathtaking quantity–one quadrillion bytes–is roughly equivalent to the amount of data stored in a mile-high stack of CD-ROMS.

Developing the human-health applications of all those data is more important today than at any other time in the last 30 years, says Durland Fish of the Yale University School of Medicine. Many “new vector-borne diseases are emerging . . . and we have very few people trained in how to handle them,” he says.

“Thirty years ago we thought we were on top of infectious disease,” says Hay. However, with pests’ growing resistance to drugs and insecticides, many of the diseases have resurged. Increasingly mobile populations are escalating the problem, and now there is the additional threat of deliberate introduction of disease by terrorists. The numerous satellites encircling Earth may provide powerful medical intelligence.

John Pickrell is a freelance writer based in Sydney and the author of Flames of Extinction: The Race to Save Australia’s Threatened Wildlife.

More Stories from Science News on Health & Medicine