Skip to content

A differentiable hydrological model for daily streamflow prediction in data-scarce basins

License

Notifications You must be signed in to change notification settings

xxcj-cyk/HydroDHM

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

62 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

A Differentiable Hydrological Model in Data-Scarce Basins

This repository contains the code for the paper "A Differentiable, Physics-Based Hydrological Model and Its Evaluation for Data-Limited Basins" (link to be added in the future). The code builds upon the torchhydro and the hydromodel packages. The former is a PyTorch-based hydrological modeling framework, while the latter provides implementations of traditional hydrological models, including the Xin'anjiang model, using Numpy.

The experiments conducted with these models can be found in the run_xaj, streamflow_prediction and data-limited_analysis directories. The calculate_and_plot directory contains the scripts used to generate the results and figures.

The model is trained and evaluated on the CAMELS dataset as well as on several basins located upstream of the Three Gorges Reservoir Area, in or near Sichuan Province, China.

For the CAMELS dataset, we will make the data available in the future. However, data sharing for the Three Gorges Reservoir Area may be subject to certain policy restrictions.

About

A differentiable hydrological model for daily streamflow prediction in data-scarce basins

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%