This Zenodo record provides the reproducible code and selected results of PoNHy (Potential for Natural Hydrogen), a Python-based modelling framework developed to quantify natural hydrogen (H₂) generation associated with the serpentinization of mantle and ultramafic rocks. The workflow integrates geological and petrophysical constraints with thermodynamic and kinetic parameterizations to estimate H₂ production rates, solubility effects, and associated uncertainties for regional-scale case studies (e.g., the Western Pyrenees and Northern California). We strongly recommend visiting the project repository for the latest updates. Contents Source code: Python modules and scripts implementing the PoNHy workflow, including data preprocessing, PGI SimPEG inversion, Monte Carlo uncertainty propagation, univariate sensitivity analyses, and post-processing for figure-ready outputs. Model configuration: Parameter and configuration files defining thermodynamic assumptions (e.g., temperature–pressure ranges, water–rock ratios), petrophysical assumptions, and statistical sampling settings used in the published case studies. Results: Selected derived datasets and summary outputs required to reproduce the main figures and tables reported in the manuscript. Large intermediate files are omitted due to size constraints but can be regenerated by rerunning the workflow. Documentation: Instructions to install the pinned software environment and to reproduce the core results. Reproducibility A pinned software environment is provided to ensure numerical reproducibility. To create the environment, run conda env create -f environment.yml from the repository root and activate it with conda activate hydrogen. This environment includes all required dependencies to run the PoNHy workflows and reproduce the results. How to cite Please cite this Zenodo record when using PoNHy or the results provided here. If you use the results in publications, also cite the associated manuscript describing the methodology and case studies. License Code is released under the MIT License. Derived results and datasets are released under CC BY 4.0, unless stated otherwise in individual folders. Project homepage / related links Project repository and documentation:https://github.com/RodolfoChristiansen/PoNHy