README

This repo contains the workflows and software used for Data Driven Modeling, primarily applied to Groundwater Modeling.

The purpose of this work is to explore the application of Machine Learning in Hydrology. The main Machine Learning technique used is Support Vector Machine.

The notebook 'DDM_UQ_main_1' documents the workflow used to identify, analyze and correct the errors associated with using a physically based model (MODFLOW) for hydrological (groundwater/surface-water) simulation.

The workflow is based on the the description in Xu and Valocchi, 2015.

The repo also includes a Python version of the toolbox DDM-UQ.

I'm still working on the notebooks and the software.

References

DDM:

• https://wiki.illinois.edu/wiki/display/mlgwm/Home

• T. Xu, A. J. Valocchi, 2015. Data-driven Methods to Improve Baseflow Prediction of a Regional Groundwater Model. Computers and Geociences. doi: 10.1016/j.cageo.2015.05.016

• T. Xu, A. J. Valocchi, J. Choi, E. Amir, 2013. Use of Machine Learning Methods to Reduce Predictive Error of Groundwater Models. Groundwater. doi: 10.1111/gwat.12061

Contact author:

If you find any errors or have any questions, please contact me at [email protected]