Fit4CAtS---An Advanced Python Framework for Atmospheric Inversion

Kurzfassung

Fits4CAtS is a flexible Level-2 framework currently in development and designed for retrieving atmospheric gases and temperatures. It leverages the `Py4CAtS` package for forward modeling and the `lmfit` package for optimization routines. Its modular design allows for quick prototyping and could serve as a blueprint for operational environments. The framework includes a range of inversion schemes, from data-driven methods such as the Matched Filter to non-linear solvers such as the Levenberg-Marquardt algorithm and the VarPro (Variable Projection) separable solver. The software will support various Py4CAtS compatible forward models, including scattering, non-scattering, and proxy setups, enabling comprehensive atmospheric analysis. The poster presented at the AGU24 showcases the current state of the framework, highlighting retrieval results from various solvers and forward models. At the time of presentation the package was in the early beta phase. Once in a stable release version the software will include a comprehensive documentation and testing routines.