DifferLand: A differentiable hybrid machine learning framework for retrieving unknown functional forms in terrestrial biosphere models

This repository contains the code and data accompanying the manuscript "Exploring Optimal Complexity for Water Stress Representation in Terrestrial Carbon Models: A Hybrid-Machine Learning Model Approach,"to be published soon by JAMES. Please refer to the README.md file for detail usage instructions. The calibrated parameters from the runs used in the manuscript can be found in the DifferLand_JAMES_calibrated_parameters.zip file.