This repository is a collection of Jupyter notebooks for learning how to search, download, and work with oceanographic data through Argovis. The notebooks are organized into three folders:
Start here. These notebooks introduce the core data types Argovis serves and the general patterns you'll use to query them:
PROFILES.ipynb— working with point/profile datasets (Argo, GO-SHIP, etc.): querying by region and time, plotting maps and vertical profiles, and making T/S diagrams.GRIDS.ipynb— working with gridded products: querying a collection, subsetting in space, and loading results as an xarray.Argovis_JSON.ipynb— for users working outside Python or in specialized applications, a tour of Argovis's raw JSON API responses.
Hands-on, guided exercises that build on the introductory material. Each activity walks through a small scientific question end-to-end using the Argovis API.
DOXY_ACTIVITY.ipynb— characterizing the vertical structure of dissolved oxygen in the ocean using BGC-Argo profiles and gridded products.
Deep dives into individual datasets whose schemas or query patterns differ from the standard profile/grid examples covered in introduction/. Reach for these when you want to work with a particular product.
Argo_trajectories.ipynb— estimated Argo parking-depth trajectories and velocities.Argo_float_location_forecasts.ipynb— the Argone float location forecast API.Intro_to_Atmospheric_Rivers.ipynb— atmospheric rivers, and Argovis's extended objects schema for region-valued data.
To run locally with Docker, clone the repo and mount it into the prebuilt image. First, install Git (https://git-scm.com/install/) and Docker (https://docs.docker.com/desktop/) on your machine, then start the Docker application. Finally, run the following commands in the terminal:"
git clone https://github.com/argovis/demo_notebooks/
cd demo_notebooks
docker container run -p 8888:8888 -v $(pwd):/books argovis/notebooks jupyter notebook --allow-root --ip=0.0.0.0
On Windows, replace $(pwd) with the full path to your cloned demo_notebooks directory (e.g. C:\Users\username\python_files\Argovis\demo_notebooks) and run the command from PowerShell. In either case, copy the printed http://127.0.0.1 URL into your browser to access the notebook environment.
You can refresh the image periodically with docker image pull argovis/notebooks, or update just the helper package from inside a notebook cell with %pip install argovisHelpers.
- Argovis' EarthCube 2022 submission: illustrates dataset colocation, QC filtering, and interpolation.