GRANADA, Spain — When a natural disaster hits, everything can change in moments. Regardless of whether it’s an earthquake, wildfire, flood, or tsunami, natural disasters have the potential to inexorably alter topography, often making it extremely difficult for first responders to know how to reach the most vulnerable.
For humanitarian workers responding to natural disasters, having instant access to up-to-date and accurate maps can mean the difference between life and death.
Increasingly, satellite maps and data are being tapped to help humanitarian workers better respond to natural disasters in a variety of ways, according to Juan Carlos Villagrán de León, program officer and head of the Bonn office at UN-SPIDER, or space-based information for disaster management and emergency response program of the United Nations’ Office for Outer Space Affairs.
“Earth observation can be a very useful tool in disaster response — you can get a sense of what areas are affected, and the degree of damage,” Villagrán de León said. “And its applications are only going to expand as technology evolves.”
Leveraging satellite data for disaster response
When the U.N. first began looking into the potential of satellite data to help with humanitarian disaster response more than a decade ago in 2007, it was clear that satellite data had real potential for humanitarian applications. However, there were a number of obstacles preventing its widespread adoption.
More from the Space Data for Development series:
“A decade ago, a pixel was the size of a football field, and that was actually considered really good,” said Adish Maudho, a geographic information system specialist who has been working with humanitarian applications of satellite data in various emergency scenarios — from the Ebola outbreaks in Liberia and Guinea, to mapping refugee settlements in South Sudan and coordinating flash flood response in Djibouti, to name only a few — since 2007.
“So if there was a flood, you could see there was water — but not which roads were blocked.”
There was also the issue of cost, as satellite images were often hundreds of dollars per image, and technical capacity, he said.
“When I first started using satellite data, making it usable involved a lot of coding,” he said. “You’d have to download these gigantic tiles of data, and then you’d have to stitch them together. It took a lot of time.”
Thanks to advances in technology, today’s satellite images are far more precise — with standard pixel definition of up to 30 cm. This allows for images of such high quality that analysts can easily differentiate between a car, truck, or cyclist — a level of specificity that would be unthinkable just a few years ago, according to Maudho.
There are also more technological tools available today that can help analysts more quickly process and stitch together satellite images.
According to Villagrán de León, satellite images give first responders a snapshot of the damage in the wake of a natural disaster, and provide critical information allowing humanitarians to better to plot out how to reach affected communities quickly.
For example, satellite images can be used to find what parts of a city have been flooded, or which roads are underwater and which can be used to deliver aid, to indicate whether helicopters or boats will be necessary. They can also be used to assess the damage caused by earthquakes, landslides, or fires.
“Most of the focus is on trying to forecast whether there is a specific event coming, and responding as quickly as possible. But there’s a lot of potential for using earth observation for early action.”— Juan Carlos Villagrán de León, program officer and head of the Bonn office, UN-SPIDER program
During the Australian forest fires earlier this year, for example, shortwave infrared imagery was used to identify the hottest part of the fire despite cloud cover, and to estimate the direction that fires were likely to spread based on which way the smoke was blowing, according to Rhiannan Price, the director of sustainable development practice at space and intelligence company Maxar.
And when a 7.8 magnitude earthquake hit Nepal, hundreds of digital humanitarians scattered around the globe were able to use pre- and post-disaster satellite images to crowdsource damage to critical infrastructure like roads and bridges, allowing first responders to plot out how to best reach victims.
Combining satellite data with local knowledge
In the absence of local, ground-level data, however, even the most detailed satellite images are of limited use, according to Nick McWilliam, a mapping technology specialist who volunteers with the humanitarian NGO MapAction.
“A satellite image will show you a building, but it won’t tell you if that building is a hospital or a residence,” he said — information that can be vital for responders who need to quickly get victims medical aid. Nor can a satellite image provide information about population density or age, which can give estimates of numbers of potential victims, and help assess vulnerability.
To improve its practical applications, satellite imagery must be integrated with other data sources — a practice that can be boosted with machine learning and artificial intelligence-
For example, Colorado-based Maxar has an entire corporate social responsibility unit dedicated to providing open-sourced earth observation data to humanitarian responders during national disasters.
To make the images more useful, they use AI algorithms to help pick out useful bits of data — for example, extracting buildings and roads from the image. They have also deliberately made all of the maps released under their disaster relief unit open-sourced, so that other organizations can use them as learning tools to prepare for future disasters.
“It’s really critical that communities around the world become more resilient and prepare for shocks and stressors,” Price said. “And for communities to build back smarter, it’s important that they can have access to that data for training purposes.”
Another organization, OpenStreetMap, combines satellite data with crowdsourced, Wikipedia-style datasets that allow contributors to overlay satellite maps with more specific ground-level data. Its Humanitarian OpenStreetMaps team focuses on mitigating or reducing the impact of natural disasters on communities by crowdsourcing comprehensive datasets on key lifeline infrastructure such as hospitals, roads, and bridges.
These datasets allow humanitarian responders instant access to multiple levels of data after a natural disaster, allowing them to respond quicker and with increased accuracy.
“Satellite data can be very useful when responding to [a] natural disaster, but local data is absolutely crucial,” McWilliam said. “You will always need that information gathered on the ground.”
Preparing for the next natural disaster
Looking into the future, UN-SPIDER’s Villagrán de León would like to see more of an emphasis on how to use satellite imagery to prepare for and mitigate the impact of natural disasters before they strike.
“Right now, most of the focus is on trying to forecast whether there is a specific event coming, and responding as quickly as possible,” he said, “But there’s a lot of potential for using earth observation for early action as well.”
For example, satellite imagery can be used to identify existing communities that may not show up on earlier maps, to make sure that vulnerable communities are taken into account and receive much-needed aid.
“The immediate aftermath of a disaster isn’t the best time to map villages, but it’s something that OpenStreetMaps does very well,” McWilliam said.
Satellite imagery can also provide useful baseline maps in regions known to be at risk for flooding, allowing governments to plot evacuation routes and build safe areas or shelters before disasters hit.
The data can also be used to create exposure and risk maps, which can inform governments on where it may be unsafe to build, and which communities are likely to be most at risk in future disasters.
“If you plan well enough in advance, you could even run simulations for the people who might be affected,” Villagrán de León said. “Using satellite data and combined on-the-ground data to prepare more comprehensively for disaster is really the next step.”
Update, Aug. 11, 2020: This article has been updated to clarify that shortwave infrared imagery was used to identify the hottest parts of the Australian wildfires, and that Maxar is a space and intelligence company.
Visit the Data for Development series for more coverage on practical ways that satellite data can be harnessed to support the work of development professionals and aid workers. You can join the conversation using the hashtag #DataForDev.