This page features documented scientific use cases allowing users of these materials to read through the computational narative, experience the unfolding of the results, experiment with changes to the code, or further extend the analysis to address other problems. Each Jupyter book features a stand-alone example that illustrates an application of machine learning in the geosciences. These include image processing, pattern recognition, data fusion (combining diverse streams of observations), land cover classification, data assimilation and prediction. The use-case notebooks are hosted on GitHub allowing for users to clone or fork the use case repository for further use or modifications.

Look for these badges to identify books that make up the GeoSMART library.

Tutorial content can be integrated into jupyterbooks in one of two ways: do it yourself (use a template book and add your content) or by providing use your content (preferably in a github repo) and asking us integrate it. The barebones template contains the minimum amount of files for setting up an online jupyter notebook, including the automated CI/CD deployment to Github Pages. The regular template book contains everything the barebones template has as well as extensive instructions for getting started, setup for conda environments to promote reproducibility, and setup for building binder environments from via conda.

If you have any doubts about which to choose, pick the regular template book. The goal is to provide executable code on some platform. The contributor can choose between Binder, Google Colab, and Free AWS (for smaller cloud-based examples). If none of the above options work for you, please contact us directly to discuss further.