Thematic code based approaches#
Developing your own code can be time consuming, but offers the most flexible way of working Earth observation data and opens up a near-unlimited range of options for visualisation. R, and Python are two of the most commonly used open-source languages for geospatial work, and there are a large number of libraries that support working with satellite data in both cases.
Both R and Python are supported by Project Jupyter, allowing them to be used to construct Jupyter Notebooks which facilitate training on using code. The landscape of available code options is ever changing, and publishing a comprehensive list of them here will be immediately out of date.
However, we can recommend some start points.
A catalogue of Jupyter Notebooks for working with Copernicus marine and atmosphere data can be found in the WEkEO Copernicus DIAS catalog
If you are working with EUMETSAT data you will find all of our Python based training code on our EUMETLAB GitLab group. This contains a number of repositories containing Jupyter Notebooks showing how to work with our marine and atmospheric composition data.
Jupyter Notebooks on dust, aerosol and fire detection can be found on the EUMETSAT TrainHub portal.
The SatPy python package offers extensive options for data visualisation for those working with weather satellites.
Radiant Earth provide Jupyter Notebooks and training data to help you work with Earth observation data and machine learning techniques.
There are also a variety of cloud based infrastructures that can support you in connecting code to data.
The Copernicus WEkEO Data and Information Access System (DIAS) cloud platform offers a free Jupyter Hub that you can use to develop cloud-based code to exploit many Copernicus data records, including those from the Sentinel satellite series. It also offers a scalable cloud infrastructure for more advanced data processing.
Google Earth Engine can be used to explore a wide catalogue of free and open satellite data, particularly when the application is focussed on land applications.
