Help to select parameters for calibration.

[Sednas Stha]

Hello all,

I need help to select appropriate parameters to calibrate my model. Below is my initial model before calibration.

How do I initially select parameters to do sensitivity analysis?

Which method of sensitivity analysis is preferred?

By looking at the graph above, what parameter do you think I should focus on?

I know that the manual and prior studies have addressed all these questions, but I would like to hear what you have to say which can make it easy to understand.

Thank you so much.

[Salman Khan]

Hi,

It's not necessarily required to check number of parameters for sensitivity analysis, you can select parameters for calibration based on the behavior of your hydrograph. I guess you can fix CN2 , Esco, Epcot and Sol_AWC parameters to capture your observed peaks. By looking into your hydrograph these parameters might be sensitive. But, before going for that make sure, you have defined elevation bands into your model and fixed driving variables like laps rates and snow parameters if your watershed have significant altitudnal variation and there is snow contribution into stream flow.

Regards

Salman Khan 

--
You received this message because you are subscribed to the Google Groups "SWAT-CUP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to swat-cup+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/swat-cup/0de4ea71-fb96-4ac9-8444-bcfea0a5336cn%40googlegroups.com.

[2w2e]

have a look here:

https://www.sciencedirect.com/science/article/pii/S0022169415001985

[Sandes Stha]

Thank you so much. 

[Sandes Stha]

Hello again,

By following your suggestion and the article, I defined elevation bands, fixed laps rate, and snow parameters. Then, I fixed the value range for CN2, SOL_AWC, ESCO, GWQMN, GW_REVAP, and REVAPMN by following the article (https://www.sciencedirect.com/science/article/pii/S0022169415001985). I also put other 12 parameters for calibration with their respective ranges. 

I ran two iterations (600 simulations). But my p-factor  (0.51) does not seem to improve to an acceptable range (>0.7) and instead, it decreased a little bit in my second iteration. 

What should I do to improve the p-factor?