03/04/20

Next COVID-19 outbreak ‘predicted via satellite’

Space spotlight bolivia epidemics - main
Earth observation satellites like Landsat could help predict the next COVID-19 outbreak. Copyright: Nasa (CC BY 2.0)

Speed read

  • Satellite data and machine learning could predict future COVID-19 outbreaks: Colwell
  • Satellites ‘provide mechanism for global surveillance and prediction of potential epidemics’
  • Everyday use of space technology no longer just science fiction

Send to a friend

The details you provide on this page will not be used to send unsolicited email, and will not be sold to a 3rd party. See privacy policy.

Satellite data could be used to predict future COVID-19 outbreaks, a leading environmental microbiologist says.

While satellites have been in orbit since the 1950s, it was only about 20 years ago that scientists began to harness their earth observation data to aid global public health.

In 2007, NASA was confident its satellites could “predict and prevent infectious disease outbreaks around the world”. Epidemiologists who lead in using space technology, however, say satellites could not have seen this coronavirus coming.

“We now have a mechanism for the global surveillance and prediction of potential epidemics.”

Rita R. Colwell, distinguished professor, University of Maryland, College Park

But now, Rita R. Colwell, a distinguished professor at the University of Maryland, College Park, has developed a predictive model for SARS-CoV-2, the scientific name for the virus that causes coronavirus disease – COVID-19.

SARS-CoV-2 is expected to become endemic — meaning it will join the rotation of viruses that affect human populations — and will likely reoccur in the future, virologists say.

Colwell, a molecular microbial ecologist, says a complex matrix of information could hold the answer to predicting when and where the next flare-up of COVID-19 cases occurs.

The ground-breaking research is expected to be published in GeoHealth, an American Geophysical Union journal.  

Colwell says her team has applied machine learning to data from China, Italy, Spain and the United States, to extract correlations with data gathered from satellites, as well as air temperatures and surface parameters of moisture, such as humidity and dew point.

“We’re in the process of testing; we think we may be able to predict for specific regions when the risk is highest,” she says.

Diseases and the environment

In the early 2000s, Colwell identified the environmental links to outbreaks of cholera, a bacterial infectious disease caused by consuming contaminated water or food.

Using data from the earth observation programme Landsat, scientists were able to show direct correlations between chlorophyll concentrations and phytoplankton blooms, and cholera.

“We were the first to develop a cholera predictive system,” says Colwell, speaking of her research partners.

“Satellites are really a very valuable public health tool. With the very sophisticated satellites up there now, we can draw on data from half a dozen satellites — that measure population movement, construction on the ground, to satellites that measure sea surface temperature, sea surface height, chlorophyll.

“We now have a mechanism for the global surveillance and prediction of potential epidemics.

Anticipating the next COVID-19 flare-up

During the deadly 2015 Ebola outbreak, Farhan M. Asrar, from the University of Toronto and France’s International Space University, found that people were increasingly understanding the benefits that outer space could provide disease management.

“Space assets are readily available and being used to benefit global health, including Ebola virus disease and other infectious diseases,” Asrar co-wrote in the Lancet.

“This benefit could be further enhanced by greater cooperation, investment, and partnership between the space sector and public health and humanitarian organisations.”

Timothy E. Ford, chair of the Department of Environmental Health Sciences at the University of Massachusetts Amherst, says that while it is difficult to foresee outbreaks of new viruses that are thought to transmit through wild meat or animal trading, “for pathogens transmitted through environmental routes or that have an environmental connection, satellites are powerful”.

“We’ll hopefully have retrospective data that could be linked with satellite imagery of major transport routes to see how the [SARS-CoV-2] virus spread.

“We may not have predicted it happening at the beginning, but we can map how to help in that process [of recovery].”

Future of satellites and outbreaks

While technology may as yet be unable to predict outbreaks of unknown diseases, analysts are keeping a watchful eye from space: “I guess I would call what we do computational epidemiology,” says Colwell.

“Satellites are not the be-all and end-all, but they can be a powerful predicter of potential future outbreaks,” Ford says.

He says satellites help direct mitigation, such as issuing warnings in areas where outbreaks are expected or diverting medical resources.

Colwell and her team created a sari cloth filtration method, rolled out in Bangladesh, to remove large amounts of cholera bacteria.

Simply folding a sari — a long piece of material wrapped around the body to form a dress — multiple times creates a water filter that can catch many contaminants. It led to a 50 per cent reduction in cholera cases, a feat Ford calls “an incredible success story”.And in 2018 and 2019, Colwell and co-researchers Antar Jutla, an engineering professor, and Anwar Huq, a cell biology and molecular genetics professor, made predictions that enabled medical and mitigation teams to plan ahead of Yemen’s cholera outbreaks.

The forecasting tool was able to predict with 92 per cent accuracy the high-risk geographic areas for outbreaks in Yemen.

The everyday use of space technology has ceased to be the stuff of science fiction.

“It’s not hypothetical any longer: it’s actual and it’s useful,” Colwell says.