Simulations with best set of parameters yield different results in SWAT-CUP and ArcSWAT

536 views
Skip to first unread message

Mauricio Arias

unread,
Apr 3, 2014, 7:21:51 PM4/3/14
to swat...@googlegroups.com
Greetings,

I have been working with a relatively large model (approx 100,000 sq km, 140 subbasins) of tributaries of the lower Mekong. The model is well calibrated for water flows and now I'm calibrating it for sediment loads. After achieving a relatively decent calibration (considering the limited data we have) and a reduced range of values for the input parameters, I went ahead and loaded the parameters for the best simulation (out of 1000 after two SUFI iterations) into ArcSWAT. I did this by first running the best simulation in SWAT-CUP, and then copying the input files (HRU, RTE, SOL, etc.) into the default simulation folder in ArcSWAT. Surprisingly, the sediment monthly results in ArcSWAT were very different from what I had initially computed in SWAT-CUP for all 5 stations were I carried out my calibration. Are there any hints as two why the results might be different?
I have seen similar questions to this one in this and in the ArcSWAT forum but not answers to them.

Thank you very much in advance for your help.

Mauricio

______________________________________________
Mauricio E. Arias

Research Associate
Department of Civil and Natural Resources Engineering
University of Canterbury
Private Bag 4800, Christchurch 8140
New Zealand
    

Karim Abbaspour

unread,
Apr 4, 2014, 7:55:43 AM4/4/14
to swat...@googlegroups.com
... I don't understand why you need to put the files back in the ArcSWAT project?
But perhaps a different swat version ran in that directory or you did not  copy all files. Otherwise it makes no sense...
Best, Karim

 
-------------------------------------------------
Dr. K.C. Abbaspour
Eawag, Swiss Federal Institute for Aquatic Science and Technology
Ueberlandstr. 133, P.O. Box 611, 8600 Duebendorf, Switzerland
email: abba...@eawag.ch
phone: +41 44 823 5359
fax: +41 44 823 5375
http://www.eawag.ch/index_EN
--
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.
For more options, visit https://groups.google.com/d/optout.


Mauricio Arias

unread,
Apr 4, 2014, 10:37:18 PM4/4/14
to swat...@googlegroups.com
Dear Karim,

Thank you for your response.

I'd like first to answer your question of why I want to put the files
back in ArcSWAT For starters, I'm only using 5 stations to calibrate
sediment loads in SWAT-CUP, but actually I am analyzing my model
results at all individual subbasins and reaches, something which is
much simpler to do once ArcSWAT results are incorporated in the
databases. Second, my calibration datasets are only 3 years long at
monthly steps, but actually, the ArcSWAT simulations are 23 years long
at daily steps. Last, I will be running different land cover change
scenarios, which would make me reconfigure the model back in ArcSWAT.
I must say that I'm convinced of the geniality and practicality of
SWAT-CUP for calibration and uncertainty analysis, but at this stage I
think myself and many other users that have previously wrote in this
group see the need to go back and run SWAT after these analyses are
completed.

Going back to my original issue (same input parameters yielding
different results in ArcSWAT and SWAT-CUP), I wonder if you have done
that test of using the exact same parameters or copy the files from
SWAT-CUP into ArcSWAT.

Thanks,

Mauricio
> You received this message because you are subscribed to a topic in the
> Google Groups "SWAT-CUP" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/swat-cup/rajdqbrez6U/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to

Karim Abbaspour

unread,
Apr 7, 2014, 11:37:08 AM4/7/14
to swat...@googlegroups.com
If you copy all files in the swat-cup project (provided you have files with best parameters in them) into the TxtinOut directory of your ArcSWAT project and use ArcSWAT interface to run swat then you are running swat exe which resides in the C:\Swat\ArcSWAT directory. Please make sure this is the same version of swat as in the swatcup project directory. Otherwise there is no reason why they should produce different results.
Best, Karim
 

----------------------------------- --------------
Dr. K.C. Abbaspour
Eawag, Swiss Federal Institute for Aquatic Science and Technology
Ueberlandstr. 133, P.O. Box 611, 8600 Duebendorf, Switzerland
email: abba...@eawag.ch
phone: +41 44 823 5359
fax: +41 44 823 5375
http://www.eawag.ch/index_EN
> email to swat-cup+unsub...@googlegroups.com.

>
> For more options, visit https://groups.google.com/d/optout.
>
>
> --
> You received this message because you are subscribed to a topic in the
> Google Groups "SWAT-CUP" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/swat-cup/rajdqbrez6U/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to

> For more options, visit https://groups.google.com/d/optout.


--
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+unsub...@googlegroups.com.

Mauricio Arias

unread,
Apr 9, 2014, 7:17:25 PM4/9/14
to swat...@googlegroups.com
Dear Karim,

Thanks again for all your feedback. I have indeed checked that both
programs are running the same SWAT version (SWAT2012 rev. 591) and
that all files are copied across. I'm glad you point that out.

I think I have found an explanation and a partial solution to my
issue. I had underestimated the role of in-stream channel sediment
processes in my model, which I understand are largely affected by how
much sediment there is in the channel from previous time steps. My
original model runs for 5 warmup yrs+29 simulation yrs, and my
calibration dataset covers only the last 4. Thus, I decided to run
SUFI with 5 warmup yrs + 4 calibration yrs. Mistake. Results are only
comparable if SUFI and ArcSWAT are run for exactly the same number of
years, which in my case was 34 yrs (5 warmup+29 simulation).

Regards,

Mauricio
>> email to swat-cup+u...@googlegroups.com.
>>
>> For more options, visit https://groups.google.com/d/optout.
>>
>>
>> --
>> You received this message because you are subscribed to a topic in the
>> Google Groups "SWAT-CUP" group.
>> To unsubscribe from this topic, visit
>> https://groups.google.com/d/topic/swat-cup/rajdqbrez6U/unsubscribe.
>> To unsubscribe from this group and all its topics, send an email to
>> swat-cup+u...@googlegroups.com.
>> For more options, visit https://groups.google.com/d/optout.
>
>
> --
> 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.
> For more options, visit https://groups.google.com/d/optout.
>
>
> --
> You received this message because you are subscribed to a topic in the
> Google Groups "SWAT-CUP" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/swat-cup/rajdqbrez6U/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> swat-cup+u...@googlegroups.com.

Karim

unread,
Apr 12, 2014, 9:55:53 PM4/12/14
to <swat-cup@googlegroups.com>
Ok! Very good, good to know this!
Best, karim


>
>
> ---------------------
Reply all
Reply to author
Forward
0 new messages