On 14 February, earth scientists and ecologists received a Valentine's Day gift from the San Francisco, California-based company Planet, which launched 88 shoebox-sized satellites on a single Indian rocket. They joined dozens already in orbit, bringing the constellation of "Doves," as these tiny imaging satellites are known, to 144. Six months from now, once the Doves have settled into their prescribed orbits, the company says it will have reached its primary goal: being able to image every point on Earth's landmass at intervals of 24 hours or less, at resolutions as high as 3.7 meters—good enough to single out large trees. It's not the resolution that's so impressive, though. It's getting a whole Earth selfie every day.
The news has already sparked excitement in the business world, which is willing to pay a premium for daily updates of telltale industrial and agricultural data like shipping in the South China Sea and corn yields in Mexico. But scientists are realizing that they, too, can take advantage of the daily data—timescales that sparser observations from other satellites and aircraft could not provide.
"This is a game changer," says Douglas McCauley, an ecologist at the University of California, Santa Barbara, who wants to use Planet imagery to map coral bleaching events as they unfold. At present, coral researchers often rely on infrequent, costly reconnaissance airplane flights. "The previous state of the science was, for me, like taking a family photo album and shaking out all the photos on the floor and then being asked to haphazardly pick up three images and tell the story of the family."
McCauley is participating in Planet's Ambassadors Program, which provides free satellite imagery to researchers as it is collected, with no lag time, under an agreement that prohibits them from reselling the data. Joe Mascaro, a tropical ecologist who runs the program, says it was created in the fall of 2015 in response to queries from scientists yearning for access to the company's growing archive of data. Over the course of 2016, Planet approved the applications of about 160 researchers across a range of fields. "We anticipate there will be many new applications of our data that we didn't anticipate," Mascaro says. The company intends to expand the program in the months ahead, and says it is looking for projects that have social, humanitarian, and environmental impacts—and that have the potential for rapid publication in peer-reviewed journals.
Andreas Kääb, a geoscientist at the University of Oslo, applied to the program to obtain additional data for his work on glaciers, including an investigation into a massive glacial avalanche in Tibet last July that killed nine herders and hundreds of sheep and yaks. Kääb already had before-and-after imagery from Landsat and Sentinel-2, U.S. government and European Space Agency satellites that have, respectively, 30-meter and 10-meter resolution and revisit intervals of 16 and 10 days. But higher resolution Planet images provided Kääb with valuable, timely clues. The appearance of large crevasses before the avalanche indicated the glacier was "surging," although surges, typically somewhat slow, don't usually lead to avalanches. But Kääb also saw water pooling on the surface of the glacier—a sign of heavy rainfall or unusually high temperatures. That water might have seeped through the crevasses, soaking the sediments below the glacial bed and creating a lubricant that triggered the sudden slip. When he saw a second nearby glacier with similar patterns, "We warned Chinese authorities, but when our warning arrived the glacier had already collapsed," Kääb says. (No people, or yaks, were hurt.)
Kääb also used Planet images to study surface displacements along fault lines in New Zealand following the country's 7.8-magnitude earthquake last November. Though high-resolution GPS ground stations are typically used for this, not all faults have dense GPS networks monitoring them. He used Planet images to determine that two fault lines had slipped between 6 and 9 meters—showing that medium-resolution optical satellites can fill the gap.
Dave Petley, who studies landslides at the University of Sheffield in the United Kingdom, has not joined the Ambassadors Program yet, but says that access to the images would be "transformational" for his research. Orbital imagery has revealed some 80,000 landslides in the wake of the New Zealand earthquake. Aftershocks are likely responsible for many of them. But because available images can be weeks apart, "we just have to assume that everything happened in the main shock," Petley says. Daily images during the sequence of aftershocks would show how the landscape responds to different amounts of shaking, Petley says, and help with disaster response. "You want to know how many of your roads are damaged, how many valleys might be blocked."
Planet's images are also finding a niche among researchers who deal with human-caused calamities, like deforestation. Matt Finer, a researcher at the Amazon Conservation Association in Washington, D.C., gets weekly deforestation alerts based on Landsat images, but says they are too coarse to determine whether the damage is natural or human-caused. He now turns to Planet data to decide whether an event is concerning. He recalls one incident when his group spotted 11 hectares of forest loss in Peru, accompanied by extensive dredging—signs of an illegal gold mining operation. "The Peruvian government was on the ground within 24 to 48 hours, kicking the miners out," he says. In previous years, Finer says, hundreds of hectares might be lost before anyone acted.
Micah Farfour, a special adviser on remote sensing at Amnesty International in New York City, is using Planet images to monitor humanitarian crises as they unfold. Timely images can help her corroborate witness testimony or pinpoint emerging refugee crises. "It's a really, really amazing tool for narrowing down time frames," Farfour says. Still, images acquired from other private satellite companies, like DigitalGlobe, remain crucial to Amnesty's work, because they can offer the 30-centimeter resolution needed to, say, identify mass graves or count the buildings destroyed in a village that's been burned to the ground.
Another limitation of Planet's Doves is that they only have four spectral bands—red, green, blue, and near-infrared—compared with Landsat's 11 bands. "Planet's daily observation frequencies are incredibly useful," says David Roy, a remote sensing scientist at South Dakota State University in Brookings and co-leader of the Landsat science team. "But there are lots of things … that are probably not doable with Planet labs data." A major missing component, he says, are thermal bands in the far infrared, which enable Landsat to monitor the evaporation of water from plants. That's "quite important if you're looking at drought monitoring or water consumption, particularly in agriculture," Roy says. The Doves also lack a shortwave infrared band, which on Landsat can distinguish between different types of vegetation.
These concerns have not slowed the juggernaut of Planet. In early February, it made two major announcements: It had folded Landsat 8 and Sentinel-2 data into its archive and it had initiated a deal to acquire Google's Terra Bella satellite imaging division and its seven SkySats, which have the capability to image at 0.7 meters. However, a spokesperson for Planet declined to say whether scientists will have access to those higher resolution images once the deal is completed.
In the meantime, as more scientists publish their papers using Planet imagery, word is getting around. Mascaro says he was at a meeting of the American Geophysical Union in December 2016 when Kääb showed how Planet data were enabling the monitoring of glaciers. "Not surprisingly, I got a few Ambassadors applications from people who were in the room."