The Nash-Sutcliffe model efficiency coefficient (NSE) is commonly used to assess the predictive power of hydrological discharge models. However, it can also be used to quantitatively describe the accuracy of model outputs for other things than discharge such as nutrient loadings, temperature, concentrations etc.
Nash-Sutcliffe efficiencies can range from -infinity to 1.0. An efficiency of 1.0 corresponds to a perfect match between model and observed data, whereas an efficiency less than 0.0 occurs when the observed mean is a better predictor than the model.
Below is a screen capture of the Script element used to calculate the mean of all values in the time series of the observed record.
Note - this model assumes the observed data is regularly spaced in time. If the spacing of observation times is not constant, please consider one of the following models:
Steps to calibrate:
- Run the model as is then view the result plot called "Fitted Model"
- Note how the predicted values are not inline with the observed points.
- Calibrate the model using GoldSim's built-in optimization by following the next steps:
- In the Run menu, select "Optimization..."
- Click on the Optimize! button
- When the optimization run is complete, click the Close button then check the box to "Copy optimized values to element definitions".
Run the model again and view the result plot to see how the model results compare to the observed data.
Contact:
Charles Ikenberry, P.E., Ph.D.
cikenberry@fyraengineering.com
www.fyraengineering.com
Comments
3 comments
Version update:
1. Fixed typo indicating that the objective function is to be minimized. Changed to say "maximized." This change does not affect the results in any way.
Hi
You can use these online calculators for Nash-sutcliffe coefficient
http://agrimetsoft.com/calculators/Nash%20Sutcliffe%20model%20Efficiency%20coefficient.aspx
It is free.
Thanks Hassan for the link. This is a great site and nice to know I have something to compare answers to.
Please sign in to leave a comment.