Correlation Plots
- Kester Scandrett
- Duy Pham
- Former user (Deleted)
A recent feature of EFDC_Explorer is the ability to plot correlations between the model and existing data to assist the user in calibration. The method for setting up correlation plots is similar to that for Model-Data Configuration in Time Series Comparisons The setting up of linkages is the same as that for Time Series Comparisons, however, with Correlation Plots the user is able to select which error statistics will be displayed on the plot, as shown in Figure 1.
Figure 1 Calibration tool: Model vs Data Correlation Plots.
Error statistics which may be displayed are explained in Section 6.1.1 and include:
- R Squared
- Average
- Absolute
- RMS
- Relative
- Nash-Sutcliffe
- Relative RMS
An example of a correlation plot is shown in Figure 2. Note that observed data is shown on the x-axis and the model data is shown on the y-axis.
Figure 2 Example model data Correlation Plots comparison for water surface elevation.
An additional feature of the Calibration plot function is that the box size for each plot may be saved. This means that even though screen aspect ratios may vary between different computers, they will still produce the same final plots. CTRL-W sets the size of the plot box size, and when the user saves the project the size of the plot will also be saved.