LAMTA example datasets and notebooks

This repository (LAMTA_examples) contains a collection of tutorial notebooks illustrating Lagrangian trajectory analysis and diagnostics using the LAMTA framework.

The full documentation, including rendered notebooks and workflow explanations, is available on Read the Docs:

Read the LAMTA examples documentation

Notes on Read the Docs

The notebooks displayed on Read the Docs are rendered using precomputed outputs. They are not executed online.

To reproduce, modify, or extend the analyses, the notebooks must be run locally in a suitable Python environment with LAMTA installed.

Data access model

The example notebooks rely on a set of reference datasets distributed via a dedicated GitHub release:

LAMTA examples data (v0.1)
Release tag: data-v0.1

The data archives are not committed to the repository. Instead, they are:

• Hosted as compressed .tar.gz assets in the GitHub release
• Downloaded automatically at runtime using pooch
• Verified using SHA256 checksums
• Extracted locally into a data/ directory

No manual download is required when running the notebooks locally or when building the documentation.

Data access is handled programmatically via the helper function ensure_dataset().

Contents of the data release

The data release provides reference datasets defining the expected file names, variable conventions, and directory structure used in the notebooks:

• Bathymetry

  • ETOPO 2022 (Mediterranean Sea)

• Sea Surface Temperature (SST)

  • GHRSST OSTIA L4 (global, sample)
  • Sentinel-3 SLSTR L2P (Mediterranean Sea, June 2023)

• Altimetry

  • CMEMS DUACS merged all-sat L4 (Europe, June 2023)
  • Near-real-time (NRT) all-sat L4 (Europe, May 2023)
  • Near-real-time (NRT) all-sat L4 (Global, September 2022)

• SWOT

  • SWOT L3 LR SSH Expert (June 2023)

Intended use

These datasets and notebooks are intended for:

• demonstration
• testing
• documentation
• development support

They are not intended for scientific production use.

Users applying LAMTA to their own datasets should use these examples as a reference for workflow and data structure, and adapt the loading routines accordingly.