Landsat AWS Sandbox

An ODC Sandbox for the Global Release of the Landsat Archive on the Amazon Cloud

The Landsat AWS Sandbox is a new initiative to create an online cloud-based data cube “sandbox” to support the Landsat Collection-2 release (September 2020) using the ODC infrastructure and application algorithms. This will enable new opportunities for training and capacity building to the global audience. To be released in September 2020!

In September 2020, USGS is expected to release the global Landsat archive in level-2 cloud-optimized GEOTIFF format (called Collection-2) on the Amazon Cloud (AWS). This release is ideally suited to the Open Data Cube (ODC) architecture as the ODC has demonstrated effective analyses and product generation using similar datasets on AWS throughout the world. 

In preparation for this release, NASA's Committee on Earth Observation Satellites (CEOS) Systems Engineering Office (SEO) is developing an online ODC User-Interface and an on-demand Jupyter notebook hub for running Python algorithms. These free and open demonstration tools, available to any global user, will allow the rapid creation of science application products anywhere in the world without the need to download and process the satellite data. Some example products that will be possible from these tools include: custom cloud-filtered mosaics, spectral indices, vegetation fractional cover, historic water extent, and water quality.

Example ODC products are shown above, including water quality variance over Lake Manyara, Tanzania (left), fractional cover over Lake Chad, Africa (center), and the ODC User Interface (right).

The ODC applications demonstrated by the SEO can be easily duplicated, modified and implemented by anyone in the world as the entire software system is free and based on open source software. Users will be able to take advantage of powerful cloud computing resources while also utilizing many other geospatial datasets available thru AWS. In addition, there is a growing community of ODC users that have an expanding set of application algorithms that are ready to apply to common applications including some of the UN Sustainable Development Goals (SDG). 

 

The release of these newly processed datasets is a significant milestone for the global user community and the next significant step in Landsat data science. The impact of this data release can only be compared to the global release of free and open Landsat data in 2007. Within a few short months, the world of satellite data will change, again. 

Contact: Brian Killough, Ph.D.   |   brian.d.killough@nasa.gov  |   CEOS Systems Engineering Office