May 6, 2019

Daily satellite imagery of the Earth now available to Stanford students and faculty

Planet's monthly mosaic, centered over Bangladesh, projected at full resolution at the Hive on the Stanford campus.
Stanford Libraries is one of the first academic institutions to add Planet’s dataset to its list of resources.


Stanford, CA--Stanford Libraries, with support from the Natural Capital Project and The Center on Food Security and the Environment (FSE), has entered into an agreement with Planet, a San Francisco based aerospace and data analytics company, to provide subscription access to their analysis-ready images. The Planet catalog of images join other large datasets, including CoreLogic and Schrodinger, available to the Stanford community courtesy of the Libraries.

With the world's largest constellation of Earth-imaging satellites in orbit, the new Planet subscription provides students, faculty and scholars with access to daily high-resolution images of Earth’s lands and oceans. Planet’s imagery can be used across several disciplines and has countless commercial and humanitarian applications—including agriculture, energy and infrastructure, the environment, and more.

“Planet offers an unprecedented combination of fine spatial resolution and temporal frequency,” said David Lobell, the Gloria and Richard Kushel Director of FSE and professor of Earth System Science at Stanford.

Lobell and Marshall Burke, assistant professor of Earth System Science, are interested in using Planet imagery to run predictive modeling around food security and sustainability, at continental scales.

“We're excited to test Planet data on some long-standing problems where coarser sensors have struggled, such as tracking economic activity in remote areas, or measuring cropping diversity in tropical agriculture,” said Lobell.

Combining machine learning and Planet imagery, the Stanford-based Natural Capital Project—a partnership between the Chinese Academy of Sciences, the University of Minnesota, the Stockholm Resilience Centre, The Nature Conservancy, and the World Wildlife Fund—is developing an algorithm for detecting and mapping dams globally. There are estimated to be over 16 million dams worldwide, with only a small fraction documented. 

“The addition of Planet data to the Libraries’ catalog of resources will offer faculty and students new tools to support the cutting-edge research being done across a variety of disciplines to address urgent global issues,” said Stace Maples, geospatial manager at Stanford Libraries. Planet’s archive, according to Maples, dates back to 2009, which also provides students and researchers with historical context to leverage.

Stanford is the second university to grant students and faculty access to the Planet satellite imagery. Faculty interest in Planet’s dataset is what spurred discussions that ultimately led to a signed agreement. “The Libraries negotiate large data contracts frequently. We were able to support faculty in securing access to the corpus of Planet images, as well as ensure the proper rights for data retention and internal redistribution were addressed,” said Maples, who manages the Stanford Libraries Geospatial Center.

Planet was started in 2010 by former NASA employees who began building satellites in their garage. Today, Planet designs, builds and operates more than 140 satellites in lower Earth orbit that constantly snap images of the planet. Planet’s images are used by decision makers in business, government, nongovernmental organizations, and now—academic institutions.

“At Planet, we are constantly experimenting with machine learning approaches to derive critical geospatial insights served at a global scale,” says Gopal Erinjippurath, Planet’s senior director of analytics engineering. “We are excited to see the students and faculty at Stanford access and explore our imagery. Stanford is a center of academic excellence, pioneering innovations in the broader realm of AI and machine learning. Planet intends to collaborate closely with research teams at Stanford to further new technologies in geospatial analytics.”

Students and faculty interested in using Planet imagery in their own research should submit a request through the Libraries' Geospatial Center. A topic guide has also been created to provide additional information on satellite imagery, including Planet.