SWAT+ Toolbox Calibration Support - Next Steps

[Liliana Bettolo]

Hello! 

I completed the calibration, but I do not yet have an NSE above 0.5 for the daily river flow for my outlet. 

I conducted a sensitivity analysis with the Sobol method and selected the top 10 parameters, as shown below. I then conducted automatic calibration with the CALSI method with the following parameter settings. I am curious what next steps I should consider around data inputs or if should explore different calibration settings or methods. What have people done from here to understand what to consider for improving NSE in calibration?? 

Thank you so much for your thoughts! 

Best, 

Liliana

[Nadya Salma Azzahra]

Hi Liliana,

I came across your post from January in the user group about struggling to get an NSE above 0.5 for daily river flow. Since a few months have passed, I wanted to reach out and ask if you eventually figured it out and managed to improve your model's performance?

I am currently working on a similar watershed modeling project where my observation data is limited to discharge, and I am facing the exact same hurdle right now.

If you did manage to get a good NSE, I would love to know what your next steps were. How did you fix it? Did you end up correcting certain data inputs, trying a different calibration algorithm besides CALSI, or focusing on manual calibration first?

Any insights or lessons learned from your experience would save me a lot of time and be incredibly helpful. Thank you so much!

Best regards,

Nadya