Google makes 40 years of Earth observations available to researchers with Google Earth Engine!

June 21, 2017
Mr. Stace D Maples
A False Color 432 Landsat composite image, made in Google Earth Engine

Last week I spent 3 days at Google for their annual Google Earth Engine Summit, learning about new features and applications of their Google Earth Engine technology. If you haven’t seen Google Earth Engine, I encourage you to go to https://earthengine.google.com and use the signup link to get an account. It’s absolutely free for non-commercial use and it’s capabilities are pretty mind-blowing. In a nutshell, Google is taking every bit of public domain satellite, climate, weather, and other datasets it can get it’s servers on and putting it into a browser-based platform that can leverage that data with traditional remote-sensing algorithms, at a global scale.

An excellent example of this is Matthew Hansen, et al’s Global Forest Change project, [ https://earthenginepartners.appspot.com/science-2013-global-forest ], which demonstrates deforestation and reforestation dynamics, at a scale of 30m per pixel, for the entire globe. Google estimates that the calculations necessary for the project, applied to a dozen years of LANDSAT imagery, would have taken over 300 years on a typical desktop computer, but in Google’s Earth Engine platform, those calculations took about 3 days. 

Here at Stanford, David Lobell and his team at the School of Earth Energy and Environment are using Google Earth Engine to build a “Scalable Satellite-based Yield Mapper” that provides reliable crop productivity data at the field or sub-field level, globally. [ http://www.g-feed.com/2015/05/introducing-scym.html ]  

With two interfaces: One the “Explorer” which provides a GUI for quick reconnaissance of imagery; The other a “Code Editor” where researchers can implement hundreds of functions and even create their own, using simple JavaScript programming, Google Earth Engine provides in “in point” for beginners and expert programmers, alike. In addition to the two interfaces, Google Earth Engine includes a Python API, as well. If you are interested in learning more about using Google Earth Engine, and how others are using it, there are fantastic tutorial materials to be found at https://developers.google.com/earth-engine/edu , and you can access the Google Earth Engine Summit 2017 Agenda site at https://events.withgoogle.com/google-earth-engine-user-summit-2017/#content

The Stanford Geospatial Center will be working with the Google Earth Engine team to bring them to the Stanford Campus for a workshop on Google Earth Engine during the Autumn Quarter, so sign up for the StanfordGIS Listserv [ https://mailman.stanford.edu/mailman/listinfo/stanfordgis ] to receive updates on that and other events of interest to Stanford researchers who need to leverage geospatial technologies and data in their research.