TEDA: A lightweight Python framework for educational data assimilation

We present TEDA (Teaching Ensemble-based Data Assimilation), a lightweight Python library designed to support teaching, experimentation, and prototyping of ensemble-based data assimilation (DA) techniques. TEDA includes educational implementations of various Ensemble Kalman Filter (EnKF) variants, such as LETKF, ETKF, EnSRF, and shrinkage-based filters. The framework emphasizes modularity and ease of extension: users can register custom assimilation methods without modifying the source code, making it ideal for classroom use and rapid experimentation. Unlike its previous version, which offered a limited and hardcoded interface, the new release introduces an extensible factory-based registry, support for structured experiments via JSON configurations, and a standardized interface for background models, observations, and inflation. TEDA facilitates the integration of custom toy models and allows users to visualize forecast–analysis errors and ensemble spread through simple plotting tools. TEDA is particularly well-suited for undergraduate and graduate-level courses in atmospheric sciences, oceanography, or machine learning, and provides a foundation for understanding the algorithmic structure of modern DA systems.