zumbro-flow-prediction is a series of Jupyter Notebooks that import and process a variety of environmental data to create a machine learning model that predicts streamflow based on 10 years of streamflow data for the South Fork Zumbro River near Rochester, MN.
This code is only accurate for the South Fork Zumbro River near Rochester, MN as the outside data is all based near that river.
- Carson DeSotel ([email protected], [email protected])
- Martha Burket ([email protected], [email protected])
Please reach out to either author if any concerns arise.
The project is built in Jupyter Notebooks. If you want to modify / run these Notebooks on your own machine, please visit Jupyter's Installation page for information on how to install Jupyter on your own machine.
Please also ensure that you're using Python 3 which can be downloaded here.
This project also uses pip to manage package installation. pip can be downloaded from here.
Once the basic software has been installed, this project can be downloaded through cloning the git repository.
git clone https://github.com/Carson-DeSotel/zumbro-flow-prediction.gitor by downloading the .zip file from the GitHub repo.
A large part of the data used in this project is handled through .csv files that have been pre-processed and are parts of the GitHub repo. Data required for watershed-delineation-land-use.ipynb is too large to be handled through GitHub, so extra downloading must be done.
Please download further required data from this Google Drive folder and export the files to the data/ folder in your local repository.
Ideally, this project should remain closed to change. Its goal is to be a proof of concept for a potential ML model that can be applied to streams.
Contact Carson DeSotel if you'd like to modify the code.