How maps packed with data help scientists fight malaria
JUDY WOODRUFF: The speed and impact of the Ebola epidemic highlighted the need for better ways to quickly predict potential outbreaks. Researchers believe data can help in their fight other diseases like malaria.
Tomorrow is World Malaria Day, making it a good time to look at the potential.
NewsHour special correspondent Spencer Michels reports.
SPENCER MICHELS: Maps are nothing new. In one form or another, they have been around for centuries. These days, we use them in our cars, we use them to illustrate the news. Now scientists have found a powerful new way to use maps to attack disease.
Epidemiologist Hugh Sturrock is trying to stamp out malaria in parts of Africa, and from his campus cubicle at the University of California San Francisco, he is trying to make high-tech maps of the risk of outbreaks of malaria, maps that will be crucial to effectively fighting the disease, but will be easy to use in the field.
HUGH STURROCK, University of California, San Francisco: If we can understand and predict where diseases are most likely to occur, then we can target those high-risk areas. We were motivated to try to build a platform that would allow non-experts to generate risk maps themselves, essentially at the click of a button.
SPENCER MICHELS: Worldwide, between 600,000 and a million people, mostly young children, die each year from malaria. The disease is spread by female mosquitoes seeking human blood. Health workers need accurate maps showing on-the-ground conditions to know where to spray insecticide and where to stock clinics.
Sturrock's maps for Swaziland in Southern Africa show where malaria cases have occurred, plus water conditions, temperatures and elevations. Until now, those facts have not been easy to analyze, even though the data has been collected.
HUGH STURROCK: There are more large-scale rainfall patterns and temperature variations and are really only available using sort of satellite information. We want to sort of bring all of that data to the hands of those people in the village.
SPENCER MICHELS: Sturrock's maps rely on data, much of it photos, that have been, and still are, collected by NASA satellites circling the globe. But that information, 40 years' worth, has languished in government vaults in South Dakota.
Now Google Earth Engine has acquired it, for free, and is working with the university and many others to put it to work. For several years, Google has been storing data, trillions of measurements, on thousands of computers that it owns. But, until recently, and, in fact, even now, using that data, making sense of it has been very difficult.
Sturrock, with the power of thousands of Google's computers at his fingertips, is combining the satellite pictures with on-the-ground information, using algorithms.
REBECCA MOORE, Manager, Google Earth Engine: An algorithm is nothing more than a recipe.
SPENCER MICHELS: Computer scientist Rebecca Moore manages Google Earth Engine.
REBECCA MOORE: These scientists are saying, I will look at this kind of satellite imagery, and then I'm going to overlay where there have been outbreaks of malaria in the past, and where there have been mosquitoes in the past, and so on, and they mix that all together into a numerical recipe, and out comes a prediction.
SPENCER MICHELS: Those predictions and the maps that produce them point to where there's a need for insecticide-treated bed nets to keep out mosquitoes. That's the goal of Nothing But Nets, part of the U.N. Foundation.
Elizabeth Ivanovich, its global health officer, says that accurate on-the-ground information is a vital component of any risk map, and in parts of Africa, collecting that data has yet to occur.
ELIZABETH IVANOVICH, United Nations Foundation: A lot of work has gone on in Swaziland to get those data systems up to speed, so that you actually when a case occurs and when a death occurs and exactly the location that that is happening in.
And that's just not the case in many countries in sub-Saharan Africa, especially countries with a much higher burden of disease.
SPENCER MICHELS: But it's not just fighting malaria that benefits from satellite data. Today, such information has become a hot commodity. Satellite pictures can provide evidence of environmental problems and clues to solving them.
Satellites record ships at sea, and the images, plus other data sent by ships, can point to overfishing and where it is happening. There's dramatic satellite imagery of the growth of urban sprawl in Las Vegas, and the shrinking of Lake Mead, its water source, that could be used for planning.
The applications so far may be just the start of a host of uses for the mined data. In the public health field, the model of the malaria project could, Sturrock says, be used with Ebola and animals that are possibly spreading the disease.
HUGH STURROCK: It is possible to map cases of Ebola and to relate those to variables that are linked to the distribution of fruit bats.
There's no reason why we can't use a lot of that — those techniques and those models and that data.
REBECCA MOORE: It's billions of megabytes of satellite imagery data. And never before has there been a technology platform that could allow scientists and government agencies to kind of mine that archive, right, and turn that into knowledge.
SPENCER MICHELS: For the PBS NewsHour, I'm Spencer Michels in Mountain View, California.