Microsoft Planetary Computer
Microsoft Planetary Computer is a platform designed to advance environmental sustainability and Earth science through the power of cloud computing. It provides a comprehensive Data Catalog that includes petabytes of Earth systems data, hosted on Azure and made freely available to users. The platform is equipped with APIs to facilitate easy searching and accessing of the data needed across this vast catalog.
It is similar in concept to Google Earth Engine. The main differences are that Planetary Computer relies on existing tools (e.g., python, JupyterLab, R, Rstudio, STAC) whereas Earth Engine has built there own processing libraries and integrated development environment (Code Editor).
The main Data Catalog includes petabytes of environmental monitoring data, in consistent, analysis-ready formats. All of the datasets below can be accessed via Azure Blob Storage, and can be used by developers whether you're working within or outside of our Planetary Computer Hub. You can use the Explorer to visualize and search for data in the catalog.
All Planetary Computer datasets are indexed using SpatioTemporal Asset Catalogs (STAC). This makes it easy to find and use data programmatrically with python or R scripting.
The Planetary Computer STAC API address is https://planetarycomputer.microsoft.com/api/stac/v1/.
You can browse the Planetary Computer STAC API using the Radiant Earth Browser
There are a few different ways to access the computing resources of Planetary Computer
Logging into the Hub, users are invited to start a virtual machine and have a few choices on the computing environment. You can launch a JupyterLab with geospatial python libraries pre-installed or you can launch a JuypterLab specifically with a GPU for machine learning. You also have the option to run R code within a Jupyter Environment.
You can also access Planetary Computer through your Local Visual Studio Code, through Github Codespaces, or through QGIS.
Once a cloud instance of JupyterLab has been launched, please read the 'README' document. It conatains many examples and tutorials you can check out to get started with python analysis. Any tools available in the Python or R ecosystem can be used in the JupyterLab.
Within a JupyterLab hosted in Planetary Computer, users have the ability to request larger processing resources if you are working with large datasets. Instead of using just one computing node, you can request multiple nodes using a tool called Dask. It is very easy to use and there are tutorials to show you exactly how to scale you analysis.
