This release contains the codes to train models with differentiable parameter learning (dPL) applied to HBV backbone as shown in these papers below. If any of the codes is useful for your research, please cite these papers.
Feng, D., Liu, J., Lawson, K., & Shen, C. (2022). Differentiable, learnable, regionalized process-based models with multiphysical outputs can approach state-of-the-art hydrologic prediction accuracy. Water Resources Research, 58, e2022WR032404. https://doi.org/10.1029/2022WR032404
Feng, D., Beck, H., Lawson, K., & Shen, C. (2022). The suitability of differentiable, learnable hydrologic models for ungauged regions and climate change impact assessment. Hydrology and Earth System Sciences Discussions, 1-28, accepted. https://doi.org/10.5194/hess-2022-245
The original full code release was made in this Zenodo link (https://doi.org/10.5281/zenodo.7091334). Please check this link out because it contains the necessary Potential Evapotranspiration data for reproducing the paper results. This Zenodo release also includes other detailed information related to running the codes. We presently use this GitHub repository to update our code release.
Please read this Instruction PDF File which includes detailed instructions for running the released codes, training and testing dPL+HBV models. Please also check out this Bug Fix File (https://bit.ly/3TOKmqK) where we log bug fixing information in historical versions of this release and answers to some frequently asked questions.
If you have any questions for this code release, feel free to contact us by duf328@psu.edu (Dapeng Feng) or cshen@engr.psu.edu (Chaopeng Shen)