Wednesday, March 17, 2010

Google's Earth Engine


In recent months Google has released some information on a new remote sensing analysis platform that they are developing called Earth Engine.  The platform will allow users to run remote sensing algorithms against petabytes of remote sensing data stored in the Google cloud.

The initial announcements were made right before the International Climate Change Conference (COP15) held Copenhagen in 2009.  The initial demo application was monitoring deforestation in the Amazon rain forest.

Source: Seeing the Forest through the cloud (The Official Google Blog)

More recently, an interview with Rebecca Moore, Engineering Manager of Earth Engine, describes more of the project's background, goals, and the intended launch (later this year in advance of COP16).

Many details of this platform are currently unknown, and those details will determine how useful the platform is to remote sensing scientists.  What level of processed data will be available? Will data be available before or after atmospheric correction (and if corrected, what algorithm)? Will the data be geolocated or resampled to a grid (and if so, what grid)?  What languages can be used to define the remote sensing algorithms?  What sensor datasets will be made available? Etc, etc, etc...

Despite the unknown details, it looks like this platform could be a valuable tool for remote sensing education, by enabling users to run and distribute remote sensing analyses to anyone with a web browser.  AmericaView will be watching this project to see how it progresses.