The ability to observe and monitor changes to the Earth’s surface via satellite — known as “remote sensing” — has become an indispensable tool for scientists and policymakers, vastly expanding our understanding of climate, forests, farming and more.
Too often, however, our high-tech vision has been obscured by something so quotidian: clouds.
That’s no longer a problem, thanks to new technology that for the first time is at the fingertips of scientists around the world.
An open-source data platform recently launched by the European Space Agency (ESA) is enabling scientists to lift the lid on clouds, allowing for faster and more precise monitoring and mapping of changes to land use in tropical places where clouds are a persistent feature.
Oddly enough, this leap has been made possible through new use of an old technology: radar.
In 2014 the ESA launched Sentinel-1A, a satellite that captures pictures of the Earth with radar, using pulses of microwave energy to penetrate all weather conditions and create images based on the returning signals.
Thanks to the new platform, these high-resolution images and data are freely available to anyone — a game-changing development for scientists such as Jenny Hewson, director of habitat monitoring in Conservation International’s (CI) Moore Center for Science, along with her colleagues who have depended mostly on optical satellite imagery until now (e.g., what you see in Google Earth) for creating maps and monitoring ecosystems.
“For us, this represents a very exciting time and a paradigm shift as far as what we can do and what we have access to,” Hewson said. “Including radar as another tool in our belt, especially in areas with persistent cloud cover where optical data just isn’t helpful, poses a huge opportunity in the way we monitor and map ecosystems.”
A main goal of scientists like Hewson: Understanding land-cover change. For example, being able to understand when and where forests are cleared for farms could enable policymakers to understand how and why it happens, in order to better manage it. Land-cover change can result in biodiversity loss and increased carbon emissions, so being able to see through clouds will enable up-to-date monitoring that can more accurately pinpoint which ecosystems face the biggest threats.
Merging two technologies
Hewson and colleagues recently employed the satellite to produce three case studies: mapping rice production in Vietnam; mapping land cover and land-use types in Sumatra, Indonesia; and mapping land-use types in San Martín, Peru.
Extensive data mining and a series of trial-and-error tests showed the researchers that in some applications, like mapping areas under rice production during the cloudy growing season, radar is a huge advantage. For others, however, the radar data alone does not produce the best mapping and monitoring products; instead, a combination of optical and radar data was the most beneficial.
For example, when trying to differentiate between mature and immature plantations and mangrove extent in Sumatra, Hewson said the radar imagery alone helped, but not entirely.
“I could see the different ages of plantations, which is helpful in terms of plantation rotation and separating plantations from natural forest, which often look consistently the same in optical data,” she explained. “When you put optical and radar together, though, you get the benefit of both of these technologies working together, yielding more land cover and land-use information, including various plantation ages and mangrove forest extents.”
A map of land cover in Sumatra, Indonesia created by combining optical and radar satellite data. (Sources: Copernicus Sentinel data (2015) and U.S. Geological Survey)
An array of applications
Monitoring mangroves is of special interest to scientists and policymakers throughout the tropical world, as these coastal forests simultaneously sequester carbon while serving as a natural barrier from storms — significant given that about half the global population lives within 60 kilometers (37 miles) of a coast. (In the Philippine village of Silonay, for example, mangroves are credited with reducing damage inflicted by Typhoon Haiyan in 2013.)
Knowing, then, where mangroves are thriving — and where they’re being degraded and cleared — has significant implications for where scientists at CI and elsewhere focus their attention.
That’s just one of the ways researchers can apply the new mapping methods made possible by the free radar data. According to Karyn Tabor, CI’s director of early warning systems (including Firecast, a remote-sensing tool created by CI), radar is also well-suited to map changes in rice paddies — critical in a country like Vietnam, where 15 million people depend on rice for food security and livelihoods.
“With the more frequent imagery available due to the cloud-free observation of the radar data, we’re able to easily separate wet and dry areas and map where the rice fields are and separate them from other agriculture,” Tabor explained. “This helps us determine how much production there is per hectare and could also help predict future needs of the population versus what can be grown.”
Another potential application: Identifying crops of shade-grown coffee, which Hewson said is difficult to map with optical imagery given the surrounding forest canopy.
“When you look at an optical image, it’s the tops of the tallest trees that you’re seeing, and it can be really tricky to see what’s under the canopy,” she said. “It may be with radar that we can actually see what’s below the canopy, such as shade-grown coffee, differently.”
“We don’t know,” she continued. “Maybe we’ll get stuck in the canopy, but maybe the radar will penetrate all the way to the ground. What’s exciting, though, is that it’s a question we haven’t been able to ask in the past, and now it’s something we can look into.”
Pushing the envelope
Tabor is enthusiastic about the findings that these images can make possible.
“Setting the standards for forest monitoring with remote sensing has always been CI’s stronghold,” she said. “So it’s only natural that we are looking at this new tool to go beyond forest monitoring. We’re always pushing the envelope on how conservation can use remote-sensing technologies.”
New technologies, of course, do come with limitations.
“These file sizes are massive,” Hewson said. “A reliable Internet connection is a must to download them, which isn’t always the case in all of our offices around the world. And radar systems are much more complex than optical data, requiring different skill sets and knowledge.”
For Hewson and Tabor, the benefits of incorporating radar outweigh the obstacles. Swift technology advances are already alleviating some challenges, with increases in processing power, cheaper data storage and more sophisticated web tools.
“We have this incredible archive of data now, and it’s amazing what new studies are coming out, just from making the data free and accessible,” Tabor said. “Everybody’s in this wonderful data exploration frenzy.”
Cassandra Kane is a staff writer for Conservation International.