SWAT+ Parameter calibration or optimization using R-SWAT

[Dr.T.N. Wickramarachchi]

Hi

I am a beginner for SWAT+ and R-SWAT.

In R-SWAT, SWAT+ parameter sampling was done as the 2nd step, Run SWAT was done as the 3rd step and Evaluate Output was done as the 4th step.

Only in the 4th step we have to select the Observed Data file (ie after running the R-SWAT).

My question is ,

Parameter calibration is a process of adjusting model parameter values within their plausible ranges to minimize the differences between simulated and observed values.

If we select  the Observed Data file (4th step) after running RSWAT (3rd step), how the above could be achieved by the R-SWAT? 

Look forward to hearing from you.

Thank you

Thushara

[Tam Nguyen]

depend on which calibration/optimization approaches you select, some approaches (for example, SUFI2)  you just run the model with given parameter sets first and then evaluate the model with observed later to select the best parameter sets while other (e.g. DDS approach - it works like what you said) => so you need to input the observed first (R-SWAT will prompt a message if you don't provide observed)

[Dr.T.N. Wickramarachchi]

Many thanks for the clear explanation.