Via Satellite

Space Scientist Highlights Commercial Remote Sensing Advancements

Remote sensing technologies have come a long way since scientists and researchers began taking photos of the Earth from the skies more than 100 years ago. Over decades, rudimentary satellite imaging systems have evolved into multi-faceted solutions capable of producing detailed photographs that tell a bigger story.

Today, these images can be leveraged for multiple purposes, such as urban geography, plant-health monitoring, military intelligence, and even coral reef and mangrove stress research.

But we’re only beginning to tap into the potential of what’s possible on Earth and other planets, such as Mars, said Tanya Harrison, director of research at Arizona State University NewSpace Initiative, who led a special session on Wednesday afternoon at SATELLITE 2019.

Throughout the 30-minute presentation, Harrison showed attendees — through a detailed slideshow presentation — exactly what commercial remote sensing is capable of producing today. The session’s timing, given real-world natural disasters, was serendipitous: The frequency of environmental challenges such as landslides, mudslides, tsunamis, hurricanes, and other emergencies has opened up a huge market need for Earth Observation (EO) technologies.

“You can actually see where entire crusts of earth have been displaced,” Harrison explained as she flipped through slide images of Earth, produced by a commercial remote sensing system. “Rather than looking at what NASA or the European Space Agency (ESA) is collecting … and hoping for the best, you get higher resolution imaging and cadence imaging.”

Admittedly, the photographic comparisons between satellite images produced by government bodies like NASA and the ESA appeared blurry, and far less detailed than those produced by commercial counterparts like Planet.

For these reasons, it seemed to baffle Harrison that many individuals in the science community have yet to take advantage of commercial remote sensing applications. In one slide, Harrison, showed an informal Twitter poll she conducted earlier this year, for which she asked her 30,000 plus followers, "Q for scientists who use satellite images in their research: do you utilize commercial image datasets?” A majority, 53 percent, responded “no.”

After following up with some of the scientists who participated in the poll, Harrison learned that multiple factors — from the perception that commercial remote sensing technologies cost too much to the assumption that the government agencies have the best technology available — drove their decisions. “They didn’t know how to access this data,” she explained. “All they know how to use is NASA data and ESA data.”

Harrison expressed hope that these data will help the University derive new information for multiple applications in the future. In March, Arizona State University announced its partnership with Planet, a San Francisco-based Earth-imaging company which operates one of the largest constellations of satellites currently in orbit. Through the partnership, ASU students and researchers will have access to the growing catalog of imagery from the company’s Dove and RapidEye 3-5m satellites, according to a press release.

One of Harrison’s biggest areas of interest is land-erosion patterns and landslides, through which the information derived from images can potentially make gigantic difference in outcomes.

For example, in the event of the Fagraskogarfjall landslide of 2018 — one of the largest landslides ever recorded — images taken by the ESA showed the size and scope of the landslide, but commercial images provided a more granular level of data. When you see a view of the landslide from the ESA Sentinel versus Planet SkySat, “you can see a huge amount of detail being formed,” she said.

These pictorial insights don’t just look cool — they highlight critical trends that can help with environmental investigations.

“We have to look at the morphology and shape to try and backtrack and figure out, ‘How did this happen in the first place? Will more happen because of the conditions, such as heavy rain, or other issues?’” Harrison said. “We can see data that might indicate another landslide can happen in the future.”

High cadence and resolution imaging already aides first responders who are trying to mitigate natural-disasters and provide much-needed relief during emergencies. During the 2017 hurricane season, for example, EO technologies were a critical aspect of response and recovery efforts. “This is a huge application for that data because [it facilitates] rapid response,” said Harrison. Data can help first responders answer basic questions such as “where’s the best place to land a helicopter?” or even detect roads that are blocked or flooded areas.

Commercial remote sensing applications can also provide insights for Mars and other planets, when coupled with other ML technologies. “If we have people working in AI, Machine Learning (ML), and big data, they may be able to think of 100 ways to process data to get more information out,” said Harrison. VS