The Environmental Source Apportionment Toolkit (ESAT) is an open-source software package that provides API and CLI functionality to create source apportionment workflows specifically targeting environmental datasets. Source apportionment in environment science is the process of mathematically estimating the profiles and contributions of multiple sources in some dataset, and in the case of ESAT, while considering data uncertainty. There are many potential use cases for source apportionment in environmental science research, such as in the fields of air quality, water quality and potentially many others.

The ESAT toolkit is written in Python and Rust, and uses common packages such as numpy, scipy and pandas for data processing. The source apportionment algorithms provided in ESAT include two variants of non-negative matrix factorization (NMF), both of which have been written in Rust and contained within the python package. A collection of data processing and visualization features are included for data and model analytics. The ESAT package includes a synthetic data generator and comparison tools to evaluate ESAT model outputs.