Impact

Making it simple to access and process satellite data

The quantity of publicly-available satellite imagery has grown in the last few decades, from Landsat-8 and NASA sensors such as MODIS, to the new series of Sentinel satellites from the European Space Agency (ESA). Each new type of data brings with it a new set of metadata, formats, processing requirements, and frequently, a new set of tools for using the data.

Several years ago we developed landsat-util, which proved to be a popular tool for searching, downloading, and fetching Landsat-8 scenes. Since then we’ve developed other utilities aimed at making it easier for end users to not only access and process data, but also to incorporate this functionality into their own software.

“Landsat imagery is among the most valuable open data produced by the U.S. Government.”

Sat-utils is a collection of libraries and command line interfaces (CLIs) that leverage existing tools where it can (e.g., GDAL, Shapely, Rasterio, Gippy) to provide a simpler interface for common imagery tasks. While Sat-utils are generally Python libraries, most contain a CLI, so users can quickly perform one-off tasks. Keeping each repository focused avoids heavy dependencies that might not be needed in downstream projects.

Innovation

A family of tools to cut down on routine tasks

We use sat-utils in several of our projects, as it provides common functionality that we routinely need.

Sat-api is an app that catalogs imagery and serves up an API for searching, which can be easily deployed to AWS using Cloud Formation. Currently, Landsat-8 and Sentinel-2 are available at our publicly available sat-api endpoint.

Sat-search is a Python-based client library for searching a sat-api endpoint. The library or CLI can be used to search the catalog, batch download any or all bands, and save results in GeoJSON for viewing, manual editing, or later downloading.

Sat-process is a library for processing satellite scenes into useful products, such as NDVI or color corrected images.

See the sat-utils organization for the complete list of sat-utils projects, then give our API a try, either on your own or using our sat-search client. If you want to develop your own applications, you may also want to look at geolambda, which provides a ready-to-go Dockerized container with common geospatial libraries, that you can easily deploy to AWS Lambda.