A year after releasing a prototype version at the climate talks in Copenhagen, Google has unveiled its much-anticipated “Earth Engine” at the climate talks in Cancun. Earth Engine is an on-line tool designed to help scientists access satellite data, analyze land-use trends and in support of efforts to halt emissions of carbon dioxide from deforestation.
Google Earth Engine represents the latest step in a scientific revolution that has allowed scientists to monitor changes in global land cover from space and provide initial estimates on tropical forest biomass. This is crucial data, given that nations are moving toward a regime in which industrialized countries would pay to reduce carbon emissions from deforestation in tropical countries. In order to do that, scientists and governments need to know how much carbon is locked up in their forests and be able to track changes over time.
—
Last year Nature included a sneak preview in a feature about a forest-carbon project in Peru. The Carnegie Institution of Science’s Greg Asner headed the project and has developed a comprehensive methodology to help tropical countries quickly and cheaply assess their forests. Asner’s system relies on aircraft-based laser measurements (see the three-dimensional map) of the standing forest, ground-based sampling and computer algorithms that pore over satellite data to produce automated estimates of deforestation and forest degradation. The results of the Peruvian project were published in September (see our coverage here).
UPDATED:
Although the original goal was to incorporate scientific algorithms into the final product, Google ended up designing its Earth Engine database to support satellite software used by independent scientists. At present, Google says it is working with Asner, Carlos Souza of Imazon, a nonprofit research organization in Brazil, and Matt Hansen of the Geographic Information Science Center at South Dakota State University. Asner says this route ended up making more sense because it provides the data while allowing scientists to continue managing and improving their systems.
“Google is providing the world’s first mega-mega-database of all satellite images,” Asner says. “It doesn’t sound very sexy, but it’s so large that it’s transformational.”
Google Earth Engine not only collects satellite images – some of which were manually uploaded from old government tapes – but processes them in order to provide a useable product for scientists. For instance, Asner says the database has provided the first complete images of the Columbian Amazon – an area that is notoriously cloudy – by simply merging images and picking out the cloudless pixels. “The Google approach is just blast it with data until you get the clear pixels,” he says.
For Asner, the idea is to disseminate these tools while providing some much-needed computing capacity for tropical countries that are ready to move forward. To that end, he says he will be making Carnegie’s satellite analysis software available for free on the web over the coming year. First, he needs to do a little ground validation work in order to make sure that the software performs in terrain where it hasn’t been tested before.
Read more about the project on Google’s official blog or go straight to the Earth Engine.