Question on applying calibrated parameters to original SWAT model

[Anh Nguyen Ngoc]

Dear professors and colleagues,

I have conducted calibration and validation using the SWAT-CUP model for a watershed at Subbasin 17, with a total of 5 parameters, including:

• 3 streamflow parameters: r__CN2.mgt, v__ESCO.hru, v__SOL_K(1).sol

• 2 sediment parameters: v__USLE_K(1).sol, v__SLSUBBSN.hru

After selecting the best parameter set, I would like to apply it back to the original SWAT model to simulate the entire period from 2010 to 2023.

My question is:

When applying these parameters to the original SWAT model, should I assign them to all HRUs and reaches across the watershed, or only to those within Subbasin 17, where the calibration was performed?

In particular, for physical parameters such as SOL_K and USLE_K, which can vary significantly across different soil types, would assigning a single fixed value (e.g., SOL_K = 20.42) to all HRUs distort the spatial characteristics of the model?

I would greatly appreciate any insights or advice from professors and colleagues with experience in SWAT to help ensure proper implementation and modeling accuracy.

Thank you very much, and best regards!

Sincerely,

[balcha yonas]

Hi,

If you calibrated the model at subbasin 17 only you have to change the default parameters only for that subbasin. And when you change it as you know soil parameters are calibrated in relative to default. Therefore, you should change all the parameters for the HRUs in subbasin 17 relative to what it was before.

regards,

Yonas

--
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/deaba24c-c806-4d02-9ef0-a09f14b7c4c4n%40googlegroups.com.

[Nguyễn Ngọc Anh]

Dear Yonas,

First of all, thank you very much for your kind and helpful response to my previous question. I would like to ask a few more follow-up questions to ensure I am applying the calibration procedure correctly in my SWAT model:

1. About parameter formatting:
   I am currently using the following 8 parameters in my calibration:

   • r__CN2.mgt

   • v__ESCO.hru

   • r__SOL_K(..).sol

   • r__SOL_AWC(..).sol

   • r__USLE_K(..).sol

   • v__USLE_P.mgt

   • v__SLSUBBSN.hru

   • v__CH_K2.rte
     → Are these formats correctly applied? In which cases should I use r__ instead of v__, especially for soil-related parameters?

2. Regarding the calibration scope:
   In my watershed, there is only one gauging station located at Subbasin 17. Does that mean the calibration and validation process should be applied only within Subbasin 17?

3. When to apply parameters to the entire watershed:
   Under what circumstances would it be appropriate to apply these calibrated parameters to the entire watershed instead of just Subbasin 17?

4. Parameter replacement methods:
   Is there a recommended method or tool to automatically apply these parameter values (from SWAT-CUP output) back into the SWAT input files? Or must this be done manually for each HRU or input file (.mgt, .sol, .hru, etc.)?

I greatly appreciate your further guidance and any suggestions from the community that can help ensure my modeling process is both accurate and efficient.

Best regards,

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/E7NPfFA94ok/unsubscribe.
To unsubscribe from this group and all its topics, send an email to swat-cup+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/swat-cup/CAD6MSQE6idTEY4hk6vgYNFgqgweCsMmfqqVCcchWTp%2B79hC%2Bdg%40mail.gmail.com.

--

Nguyễn Ngọc Anh

Bộ môn Quản lý đất đai & BĐS

Khoa Quản lý Tài nguyên, trường Đại học Nông lâm Thái Nguyên

ĐT: 0983.454.954

[balcha yonas]

Hi,

1. The parameter format is correct

2. If the gauging location is at the outlet of your watershed parametrization would apply to all the sub-basins.
3. If you specifically stated the parametrization (refer SWAT-CUP user manual page 51) for that specific sub-basin then the change only applies to that sub-basin.

4. There is an option in arcswat (Edit SWAT Input then subasins tab) or maybe also QSWAT where you write these changes into the SWAT database manually for each subbasin.

regards,

Yonas

To view this discussion visit https://groups.google.com/d/msgid/swat-cup/CACoMOnN687hrTcK6r44Uy7k0Tx8%3DDduLQ9%3Dha3YkOqDZ7609fw%40mail.gmail.com.

[balcha yonas]

you can watch these tutorial video on how to replace calibrated parameters into SWAT:

https://www.youtube.com/watch?v=WruBqQmWaf0&list=PLeFmp55BgZALRvapKUwwuiKJMKiNTLPuD&index=32

[Nguyễn Ngọc Anh]

Dear Yonas,

Thank you very much for your detailed response and clarification regarding parameter application in SWAT.

Your explanations on when to apply parameter changes to the entire watershed versus a specific subbasin were extremely helpful. I also appreciate the reference to the SWAT-CUP manual and the suggestion to use ArcSWAT or QSWAT to implement parameter updates directly in the database — this gives me a clearer direction for my next steps.

If I need further advice or support in the future, I sincerely hope to receive your guidance again.

Thanks again for your support and time.

Best regards,

To view this discussion visit https://groups.google.com/d/msgid/swat-cup/CAD6MSQFneWHXvJaQp07xVOeTmS-SopCi6RzduSBBi0Zbcs5yPg%40mail.gmail.com.

[Nguyễn Ngọc Anh]

Thank you again for your helpful explanations.

I would like to share a small observation regarding the example shown in the video tutorial, particularly at minute 13:59, where the parameter r__SOL_AWC(..).sol is used.

In the video, the value 0.516667 is directly assigned, which appears more suitable for a v__ (absolute) replacement. However, since SOL_AWC varies significantly between HRUs depending on soil type, using the r__ prefix (for relative change) would require adjusting values relative to each HRU's original value, not assigning the same absolute value.

I just wanted to confirm that, to ensure spatial variability is preserved, we should use r__ only when applying relative percentages (e.g., +10%), and not fixed values as shown in that example.

Looking forward to hearing your confirmation or advice on this matter.

Best regards,

To view this discussion visit https://groups.google.com/d/msgid/swat-cup/CAD6MSQFquH5TVhmrYmEwA4QeLGOu51WyMbZLxUsrk5MuqEBG2A%40mail.gmail.com.