Home

TOAD (Tipping and Other Abrupt events Detector) is a Python framework for detecting and clustering spatio-temporal patterns in gridded Earth-system datasets. It is presented in detail in Harteg et al. (2026), available as a preprint on EGUsphere.

Note

For general information, project overview, and the latest updates, visit the TOAD GitHub repository.

If you’re new to TOAD, start with the Installation guide and then follow the Quick Start tutorial.

What’s in this documentation?

This documentation provides comprehensive guides for using TOAD in your research:

The TOAD Pipeline

TOAD pipeline workflow

TOAD provides a structured workflow for analyzing Earth-system data:

  1. Shift Detection: Identify abrupt transitions at individual grid cells using configurable detection methods

  2. Clustering: Group detected shifts spatially and temporally to reveal cohesive patterns

  3. Aggregation & Synthesis: Aggregate results across multiple datasets, models, or methods to produce consensus clusters

About

TOAD is developed at the Potsdam Institute for Climate Impact Research (PIK) and the Max Planck Institute of Geoanthropology. The project originated from early prototype work by Sina Loriani in 2022. Since 2024, Jakob Harteg has led the full development of the package as part of his PhD project. Over time, numerous contributors have played important roles at various stages, including Lukas Röhrich and Fritz Kühlein. The project has also benefited greatly from scientific advice and guidance from Sina Loriani, Jonathan Donges, Ricarda Winkelmann, and many others. Community contributions, such as feature suggestions, bug reporting, or even extensions like new detection algorithms, are very welcome.

Getting Help

  • Documentation: Browse the sections above or use the search function

  • GitHub Issues: Report bugs or request features on GitHub

  • Source Code: View the source code on GitHub