Multi-site deep learning for groundwater level prediction across worldwide datasets: toward scalable applications under data scarcity - data and code

This repository provides the data and Python code enabling the reproduction of published results (https://doi.org/10.2166/hydro.2025.095). The folders are structured as follows:- The folder "00Scripts" contains the Python code in Jupyter notebooks for running the models, analyzing the model results (saved to folder "02Results"), and creating plots and tables (saved to folder "03Plots").- The folder "01Data" contains all the data utilized in the Python code ("00Scripts").- The folder "02Results" contains the output of the model runs.- The folder "03Plots" contains plots and tables created from the model outputs ("02Results"). When using data or code from this repository, please cite both the repository and the associated publication. Refer to the publication for further details and context.