SWAT-CUP+ Calibration Error and Static NSE Output

139 views
Skip to first unread message

Harprabhjot Kaur Dhaliwal

unread,
May 14, 2025, 12:28:57 PMMay 14
to SWAT-CUP

Dear Dr. Abbaspour,

I hope this message finds you well.

I am currently working on calibrating my SWAT+ model using SWAT-CUP+. As a first step, I used only one parameter (v__WDPQ.bsn) to test the initial setup. This run completed successfully, and I obtained an NSE value of -19.91, along with a basic hydrograph as shown in the first picture.

  Following this, I expanded the Par_inf.txt file to include several additional hydrologically sensitive parameters to initiate the actual calibration. However, despite these changes, the NSE value remained unchanged at -19.91, and the hydrograph also looked identical to the initial run. This led me to suspect that the parameter updates might not have been applied correctly.

When I attempted to manually run swat_edit.exe within the SWAT-CUP project folder, I encountered an error message(2.png). 

Could you please advise on what might be going wrong here?  I would truly appreciate your guidance on resolving this.

Thank you very much for your time and support. 

Best regards,

Harprabhjot

Screenshot 2025-05-14 112356.png
2.png

balcha yonas

unread,
May 14, 2025, 12:51:17 PMMay 14
to swat...@googlegroups.com
Try restoring from backup files.

--
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 visit https://groups.google.com/d/msgid/swat-cup/4ec2fc10-65d4-4634-aede-becead2fe121n%40googlegroups.com.

2w2e

unread,
May 14, 2025, 1:48:59 PMMay 14
to SWAT-CUP
Are you properly set up for calibration in file.cio?$
SWATCUP must be running in order to execute any program from the command line.

Harprabhjot Kaur Dhaliwal

unread,
May 14, 2025, 3:40:52 PMMay 14
to SWAT-CUP
Thank you so much for the help!!
  Yes, my file.cio file was missing calibration.cal, and codes.sft. I have attached the updated file.cio. Could you please confirm if it is correct?

Thanks!! 

file.cio

Harprabhjot Kaur Dhaliwal

unread,
May 19, 2025, 3:01:41 PMMay 19
to SWAT-CUP

Dear Dr. Abbaspour,

I hope this message finds you well.

I am currently working on calibrating a SWAT+ model for a watershed in Minnesota using SWAT-CUP+. While running the SPE_run.bat file, the batch process executes the SWAT simulation directly without showing the intermediate Swat_Edit step where parameter values are updated—unlike the standard process shown in several tutorials and documentation videos.

Additionally, when I attempt to run Swat_Edit.exe directly from the command line, I encounter the error as shown in the screenshot.

  Furthermore, the model output is currently underestimating flow compared to the observed data, and the calibration hydrograph does not align with reality.   

Could you please advise on what might be going wrong here?  I would truly appreciate your guidance on resolving this.

Thank you very much for your time and support. 

Best regards,

Harprabhjot


5.png
4.png
3.png
1.png

Abdurezak Ababor

unread,
May 20, 2025, 8:14:01 AMMay 20
to swat...@googlegroups.com
anyone who attach to me link of SWATCUP 5.2

--
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.

Abdurezak Ababor

unread,
May 20, 2025, 8:15:28 AMMay 20
to swat...@googlegroups.com
SWATCUP2019 5.2
Message has been deleted

Jeewanthi Sirisena

unread,
Jun 29, 2025, 5:29:04 AMJun 29
to SWAT-CUP
@Harprabhjot, Did you solve the problem of very low simulation results (underestimated flow)?
Thanks.
-Jeewa-

Harprabhjot Kaur Dhaliwal

unread,
Jun 29, 2025, 12:50:34 PMJun 29
to swat...@googlegroups.com
No, I'm still facing the same issue. The simulated flow results in SWAT-CUP+ are much worse compared to those from the SWAT+ Editor. Are you experiencing the same issue?

Reply all
Reply to author
Forward
0 new messages