Scientists at the National Ecological Observatory Network, or NEON, study how ecosystems change over time. But their field instruments and low-altitude aircraft capture only short windows of data, and NEON does not have a standing satellite imagery archive of the same sites.
For his Information School Capstone, MLIS student John Adler is helping NEON address these limitations. His project, “Archiving a Large Satellite Ecological Dataset,” will create a large, offline archive of satellite imagery to help researchers develop more complete pictures of ecological change. The archive will include images from a 5-kilometer radius around selected NEON sites, extending back six years.
Hosted on a University of Washington Information School server, the archive could help NEON researchers study ecological signals such as peak greenness, when plants are most productive, and normalized difference vegetation index, or NDVI, which uses reflected light to measure vegetation health. Those measures can help scientists understand when and where to collect the most useful environmental data.
Adler’s background in the Navy, as well as his previous work with NOAA and NEON, made the project “a natural fit,” he said. “I'm always trying to improve and tweak things, and this is a way to continue science and get information out to people.”
“I was an old-school navigator in the Navy,” Adler recalled, “meaning I literally would use a sextant and look up at the stars and the sun, and it was really important because at the time, the GPS constellation was not in full operation. And in Antarctica, everything you normally think about in navigation doesn't apply.”
That background in mapping and scientific data shaped the way Adler approached librarianship. He also brought a longstanding interest in libraries, having served for several years on the board of the Nederland Community Library in Colorado. When his wife’s work brought them to Seattle, the residential Master of Library and Information Science program offered a way to connect those interests: science, information access, mapping and service.
The technical work of the Capstone project begins with identifying the coordinates for each NEON tower site, then using programming to select the best cloud-free, non-winter satellite images from Planet, a private satellite company. Adler will also build an archived data structure large enough to hold an estimated two terabytes of that imagery.
That work could help make NEON’s data collection more efficient and its datasets more robust. “Given their high spatial resolution and temporal coverage, these data are ideal for a range of ecological applications,” said John Musinsky, a research scientist with NEON. The final archive is expected to include eight-band imagery that extends beyond ordinary visible-light photographs into bands such as near infrared, which can help researchers evaluate plant health and productivity. NEON researchers will be able to test whether the satellite archive can supplement existing field and aircraft measurements, especially in off years when a site is not visited by NEON aircraft. “Results so far have been very promising,” Musinsky said.
The project is also shaped by a broader question: Who should control and preserve digital information? By housing the stand-alone archive at the Information School, Adler is helping to create a locally managed repository so researchers have a robust backup copy. “The fact that you can set your servers up with your local IT department and create your own repository is, I think, pretty powerful,” said Adler.
Adler credits his MLIS coursework with helping him see how the pieces fit together. Classes in government information, data science, digital preservation, and the history and future of libraries all contributed to the project’s shape. He also credits his network of professors with connecting him to the right people and tools.
For Adler, that has been one of the most meaningful parts of the MLIS program: seeing how librarianship applies to emerging forms of digital information. His Capstone reflects an expansive view of what librarians can do, including, for example, building systems that help environmental scientists find, secure and apply massive digital datasets.
Adler imagines a future in which libraries can be like “escrow agents” for an increasingly cloud-based culture where books, music and other intellectual property are often licensed through digital platforms rather than owned as physical objects.
“It might sound crazy to give the digital rights and a copy to the Library of Congress,” he said, “but this is the transition period we’re in, and I’m thinking on a 500-year scale. Thinking about traditional books and manuscripts, 500 years is not a lot of time.”