the differences in swat cup and manual calibration helper in arcswat

[hurem dutal]

My question is that don’t manual calibration helper and swatcup  work similarly? Is this the reason of mismatch when we set back best parameter to arcswat? I also would like to say that I won’t set back parameters to arc swat after calibration. I only want to understand the reason of the mismatch when we set back parameter to arcswat. I read the following:

If you have calibrated your parameters at the hru level, meaning you have assigned different values to a soil parameter in different subbasins having the same soil unit, then it is just impossible to bring these values into ArcSWAT. Because here you have to basically redrawn the soil map. 

Karim

,and

Some can be easily changed in ArcSWAT, but some will require some knowledge of SQL queries. Basically, if you have any parameters that are spatially variable, like CN2, any of the .sub, .rte, .sol, .hru, etc. - then SQL tutorial is your helper :) 

Best,

Natalja

 

According to these explanations, manual calibration helper is insufficient to change parameters compared with swatcup. For example, we found best parameter value of cn2 as 0.10, and its range 0.08 to 0.12 in swat cup.  Swat cup assigns cn2*1.10 to all hrus or assigns different coefficient from 0.08 to 0.12 to each hru. When we used the manual calibration helper, arcswat assigns cn2*1.10 to all hrus. Is this the reason of the mismatch?

[Natalja C.]

Hello,

In short, if you set back all the calibrated parameters of the "best fit" from SWAT-CUP back into arc-swat (I do not know why you would do that, but anyway) there will be no difference. 

You see, both ArcSWAT and SWAT-CUP are just a GUI that works with editing the necessary text files and calls the one and only swat executable. Which means, that you have a model setup, and it doesn't matter if it is stored in your ArcSwat project folder location or in the swat-cup project location, if your parameters are set in the same way you will get the same result. If you get a different result - it means that your parameters are set in a different way.

To put it simpler, if you download a video (totally legal one), and put it on your C drive and a copy on the D drive. You will open one on the C drive in the windows media player and the other (on the D drive) in vlc player - they will run thru different interfaces, but the content will be the same. If it is not the same - then you have a different video (maybe a different episode :D ).

So...if you see a "different video" it means that your models are not identical, which further means that you did not set back all the calibrated parameters back into the arcswat.

And again, I must ask, why do you want to set the parameters back to ArcSWAT. Just continue to work with your calibrated model from wherever it is stored (just play the video, and doesn't matter if it is on C or D drive or thru which player it gets played, what matters is to get the episode number correct :D).

Best,

Natalja

[hurem dutal]

Dear Natalia,
Firstly I world like to thank you for your answer and interest.
As l mentioned before, l wont set back so called parameters to arcswat. Because l know that my calibration results are parameter range, not so called best parameters. And l understand the stochastic approach of the swatcup very well. I agree with you and Karim on the opinion that a parameter may has different values in natural processes based on other factors. So, it is not correct to give a certain value to a parameter.

The reason why l asked this question is just a cruosity. Because l set a model in arcswat, then l run a simulation in swatcup without any change in parameter.l set cn2 as relative, min max 0. Because this doesnt lead to any change.Then l compared the results(flowout) obtained swatcup and SWAT. The results were the same. After that, l copied the files in backup and pasted them the main swatcup file. Then l run 50 simulation. I set best parameters to arcswat. The results were not the same. That is all.
Have you tried this before. If yes, how was it? Or l did something wrong if manual calibration helper has the same ability as swatcup.

Thank you again

My best regards

[Natalja C.]

Hi,

I really do not understand your issue, sorry :)

The manual helper doesn't have the same ability, as swat-cup does. And, again, if you received different results, that means that you did not set the parameter values back into arcswat, or did not do it in a correct way.

BTW, I would be very careful in stating that "it is not correct to give a certain value to a parameter". There are some people that might rightfully disagree, but this is a totally different discussion topic and would not lead to any objective results. 
It is nice that you experiment, hope you will find your answers :) . Best,

Natalja

[hurem dutal]

Dear Natalja,

what is the correct way? Actually I wonder it. Could you please explain how I can set back to arcswat to get same results

My best

1 Ağustos 2019 Perşembe 08:44:38 UTC+3 tarihinde Natalja C. yazdı:

[Natalja C.]

Well, it depends on which parameters, which files, which values/factors, etc. This can be as easy as "placing a tick" in a specific location, to as complex as database queries can be.

Basically, you have to modify your database tables in Access and then rewrite the input files from the SWATEdit interface (or is it ArcSWAT interface). For this you have to have an understanding of DB queries .

Or you can rewrite the input files using scripts (R, Python, Matlab...). <-- this is what I do, but it cannot be learned or reproduced by following a simple post.

I cannot show you the "correct way", because there is none. You have to decide for yourself which is the most convenient way for you. 

Best,

Natalja

[hurem dutal]

Dear Natalja,

According to your answer, I cant do it with only using manual calibration helper. 

Best

1 Ağustos 2019 Perşembe 16:38:11 UTC+3 tarihinde Natalja C. yazdı:

[Natalja C.]

I did not use the manual calibration helper so much. But as I said, it all depends on the chosen parameters :)

What you can do is follow the suggestions from this thread: https://groups.google.com/forum/#!topic/swat-cup/SRyTyrM6C0A and/or https://groups.google.com/forum/#!topic/swat-cup/uc1kPGZ4MHw

Best,

Natalja

[hurem dutal]

Thank you very much for your consultancy. It is valuable and helpful recommendation for me.

have a nice day

My best regards

1 Ağustos 2019 Perşembe 17:01:42 UTC+3 tarihinde Natalja C. yazdı: