Machine Learning to Reliably Define Hydrologic Calibration Parameters

Development of a Machine Learning Algorithm


Kyle Thomson

Get in touch ➔

The calibration or validation of hydrological models is a critical first step in the process of calculating design flows. WMS has developed a machine learning algorithm to reduce the uncertainty of hydrologic model calibration (Figure 1 and Figure 2) and benchmark results (Figure 3) against the algorithms’ performance (Figure 4).

Figure 1
Figure 3
Figure 2
Figure 4
Figure 5

The model runs fast and provides a rapid turn-around, it will help to inform on issues related the storm or catchment files used or if a reasonable calibration cannot be achieved (i.e. due to poor gauge data).

Reach out to WMS if you would like to incorporate this tool into your next project!