hydronetworks is a network-aware hydrological modelling tool using GeoPandas and rasterio, aimed at country-scale scriptable hydrological modelling.

The Jupyter notebook (example.ipynb) is split into several sections:

1. Create network arcs and nodes based on input GIS river network
2. Apply Shreve stream order to the network
3. Use rasterio to extract attributes from several raster layers
4. Calculate rainfall runoff from precipitatin
5. Use the network to calculate discharge contributions downstream
6. Estimate mini-hydropower potential in the network based on some parameters
7. Save the results as several geopackages

There's some more detail and pretty pictures in my blog post: Modelling hydrological networks at massive scale

Requirements

hydronetworks is confirmed to work with Python >= 3.6 with the following packages installed:

• numpy >= 1.15.4
• pandas >= 0.23.4
• shapely >= 1.6.4
• rasterio >= 1.0.13
• geopandas >= 0.4.0
• pyproj >= 1.9.5.1
• matplotlib >= 3.0.2

Install from GitHub

Download or clone the repository and install the required packages (preferably in a virtual environment):

git clone https://github.com/carderne/hydronetworks.git
cd hydronetworks
pip install -r requirements.txt

You can run make test in the directory, which will do an entire run through using the test data and confirm whether everything is set up properly.