Every day hundreds of satellites peer down at the Earth from their orbits in the outer layers of the atmosphere. Over the decades the sensors aboard these space-borne observers have collected petabytes of imagery cataloguing the planet’s changing landscape and humanity’s impact on it. It represents a vast quantity of data largely untapped by newsrooms.
This month, more than 70 journalists, scientists and other experts met to explore this new frontier in data journalism at Google’s San Francisco offices. The daylong event – co-hosted by The Center for Investigative Reporting and Google on Sept. 8 – was the 20th in CIR’s TechRaking series.
The series explores the intersection of journalism and technology. The goal of the latest installment was to learn more about remote sensing and to create a lasting resource for other reporters working with imagery data – The Space Journalism Handbook.
Remote sensing typically uses sensors to measure radiation from the sun reflected by the Earth, or emitted by active sensors like radar or LIDAR. The data is collected from different bands along the electromagnetic spectrum, which gives a broader visibility than the visible spectrum alone. Google’s Chris Herwig defined remote sensing as gathering information about something you can’t physically touch.
Here’s the beauty of working with imagery: Because the information is stored as pixels, which are actually just numbers, you can then do math with those numbers resulting in new views of the data. This allows you to do things like peek through cloud cover, or clearly distinguish burn scars on the ground from active wildfires. Other applications include mapping, land cover classification, deforestation and climate monitoring.
— Willie Shubert (@WillieShubert) September 8, 2017
The day was structured around presentations from data and tool providers, punctuated by lightning talks from journalists giving insights into their projects. Among the journalists:
- Martha Mendoza, a national writer for the Associated Press, presented her remote-sensing project Seafood from Slaves, a Pulitzer Prize-winning investigation that resulted in enslaved men in the fishing industry being freed. In a related presentation David Kroodsma – research program director at Global Fishing Watch – showcased an interactive tool that tracks commercial fishing vessels throughout the world.
- Jeff Larson, a reporter for ProPublica, gave a timely overview of his Hell and High Water project, looking at the possible effects of a large-scale hurricane striking Houston.
- Michael Corey, a news applications developer for CIR, explained how he used satellite imagery to help uncover the biggest water user in Southern California’s Bel-Air during the drought.
NASA’s Amber Jean McCullum gave an overview of the space agency’s satellites and what data they collect. She also provided a list of useful tools and resources, such as the NASA Water Resources program, and the NASA Capacity Building program, and NASA Worldview.
Rhiannan Price, a representative of DigitalGlobe, showcased her firm’s satellites, capable of capturing images of the Earth at much higher resolutions than what is publicly available from government organizations such as NASA and the European Space Agency. She outlined a number of inspiring projects that have leveraged their data, including working to eliminate malaria, tracking displaced populations, and following the supply chain of exotic timber. She also shows some timely images that were just taken of the recent flooding in Houston after Hurricane Harvey.
Obtaining data is only part of the equation. Journalists working with imagery data must also know the techniques used to analyze it and the science behind them. Google’s Herwig and Nick Clinton demonstrated some of the algorithms and machine learning methods used to help classify images and extract information from them, such as the percentage of land covered by healthy vegetation or man-made construction. Herwig also gave a thorough how-to on Google Earth Engine, a tool developed to simplify some of the complexity in analyzing remote sensing data.
— Lindsay Irving (@lichen_lindsay) September 8, 2017
Satellite imagery can be beautiful; a kaleidoscope of colors draped over the Earth’s surface. But journalists must exercise caution so the results of their analysis are not lost in a sea of colors, said Rob Simmon from Planet Labs, a private startup with more than 150 satellites in orbit. Simmon, who specializes in data visualization for Planet, talked about best practices for presenting visual information to viewers using clear and compelling imagery.
Pulling it all together, Google’s Vanessa Schneider moderated a panel with scientists and journalists focused on how to enable collaboration across disciplines. Panelists included journalists Larson and Mendoza along with Forrest Melton, a NASA researcher, and John Amos, co-founder of the non-profit SkyTruth. They discussed the importance of accuracy and deadlines but also touched on what journalists do best: storytelling and humanizing highly scientific topics.
Attendees then broke into small groups to discuss where space journalism should be going, what problems to focused on and how to begin to overcome them. These conversations led to an outline for The Space Journalism Handbook. Participants will continue to work over to develop the guide, with the goal of publishing it online later in the year.