During SWAT model calibration, the simulated flow, sediment, nitrogen and phosphorus results are frequently compared with observed values to evaluate the model performance. SWAT Output Viewer has the ability to quickly calculate several model performance indicators on-the-fly (as shown below) once the model simulation is done.
To be able to compare with observed data, the observed data needs to be imported first, which could be done in project view. Take flow data as the example. Observed flow data is available for reach 5. Then the observed flow data needs to be associated with the simulated flow from reach 5. To do this, select reach 5 on the map and chose flow (m3/s) from the Observation drop down list (as shown below). Then click the the Load button to load the observed data from selected file. The loaded data is then associated with the flow from reach 5. This process only needs to be done once.
Once the observed data is loaded, it could be used to calculate the performance indicators. These indicators are automatically calculated and displayed in a table when entering the performance view as shown below. The default performance indicator is NSE. It can be change to R2, RMSE, Bias, etc. If observed data is available at more than one location or for more than one variable (flow, sediment, N and P), the performance indicator would be calculated for all the locations and variables and list in a table. The table could be sorted by any columns.
Furthermore, the performance indicator of each year is also calculated when one row is selected in the table above, with which the “best” and “worst” year could be identified.
For each year, the simulated and observed data is also plotted as shown below, where the simulated data is in read and the observed data is in green.
In SWAT Output Viewer, the performance indicators are available simply by clicking once without generating any additional files. It’s quite useful when the model is calibrated manually.