Hey, Google, Find Some Water for My Cows

Earth Engine, a data analysis service, is helping people around the world in ways simple satellite imaging can’t

- Leslie Kaufman

In the Ferlo region of Senegal, climate change is making it harder for herders to find water for their cattle. In the Peruvian Amazon, indigenous tribes must remain alert at all times to chase off illegal loggers. In Los Angeles, city planners are struggling to figure out how best to plant 90,000 trees to help cool the hottest neighborhoods. To solve these problems, all three groups rely on a little-known arm of Google that works closely with environmental advocates around the world.

Google Earth Engine has created a vast trove of open source satellite imagery and augmented it with data-analysis software that makes it relatively easy for skilled outsiders to draw up their own interactive maps. More important, its team of staff scientists (Google won’t say how many) wrangles massive data sets to answer critical questions for a constellation of pro bono “clients” that includes conservation groups, city agencies, community advocates, and researchers.

The 20,000 image files added to the Earth Engine team’s collection each day are more than just static photos. Satellites gather, for example, detailed information on the soil composition more than a foot underground and the amount of water vapor rising from farmland. Staffers help clients distill relevant information and relay it to the field. Via radio, the nomads in Senegal learn where to find a drink for their cows; bright pink dots on interactive maps inform the indigenous Peruvians of the locations of logging activity; and a shared website targets the spots in LA where trees will likely do the most good.