Land Use Change Scenario in SWAT

[Rajveer]

Hello every one 

I want to analysis the Land use change impact on Hydrology of my watershed. For that i have created Land Use maps of 1975, 1990, 2000 and 2016 from Landsat images. I have climate data from 1975 to 2005. I have ran the model with all these four Land Use maps with keeping the climate data same i.e. (1975-2005) 

I have some doubt, I am not getting which Land use map should i used for Calibration??? And how to fix the calibrated parameters for other Land use scenario ???

 

Please suggest me. Thanks in advance      

[Natalja C.]

Hi,
So this is how the workflow should be:
1) Create a model for the LU map (which corresponds to the years that you have observation data). Choose a baseline scenario (a landuse map and climate data and observations). For example, you have observations for flow for 1995-2010, then make the model with the map, which represents these years in terms of LU (maybe 2000). Call this model your baseline model.
2) Calibrate the model with this dataset;
3) Validate the model with a relevant dataset;
4) you are good to go with simulating changes on the LU. You do not need to "fix" the calibrated parameters for other landuse maps, because your model will be already calibrated and representing the baseline. From there, you just compare the different outputs.

There is another option, a more difficult one, no make a model and update your maps with a LULC change table. I never tried this, so maybe someone else will have an idea.


Hope this helps,
Natalja

[Rajveer]

Hello Natalja, 

Thanks for replying.

 I have climate as well as observed discharge data for the same period i.e. (1975-2005, both discharge and climate). In that case what would be my base line model?? 

As you said i do not have to fix the calibrated parameters, it means i have to change only the LU map in the calibrated model.?? or should i built a new model with different LU map

Regards 

[Natalja C.]

It is not for me to decide on the baseline scenario, this should be done by you depending on what you want to study. An example is: in year X there was a political decision and half of the forests were cut down, so then I will have a baseline before year X; Or in year Y there was a dam build and all the wetlands transformed into drylands, so I will model the baseline before this dam was built. See my point? So, as in the majority of questions, the answer is: it depends.

And you are right for the second part. Have your model calibrated for the baseline, and make the LULC update and see what happens.

Best,
Natalja

[Rajveer]

Thanks Natalja, 

Now it is clear to me how to choose the base line scenario. 

 I have calibrated my base line model in SWAT-CUP using SUFI-2 and got R2=0.85 and NSE=0.82 which shows satisfactory model performance. But i have some doubt about calibrated model that 

How can i take back these calibrated parameters in SWAT again?? Should i replace all the TaxtInOut files from Sufi-2 output folder to SWAT Scenario TaxtInOut folder. Or is there any other way??

Regards 

[Natalja C.]

I'm not sure what do you want to achieve?
Do you want to "restore" the old (uncalibrated) model? If so, this is a SWAT-CUP related question, but I think I know the answer. In SWAT-CUP there is a possibility to restore the files from backup. Just hit that option and it should do the trick.

Natalja

[Rajveer]

No, I don't want to restore uncalibrated  model. 

SWAT-CUP creates its own folder in which it copied all the files from SWAT scenario ==> Default folder. 

When we ran the SWAT-CUP it made changes in the files that are saved in SWAT-CUP directory folder only it does not make any change in SWAT scenario (Default) folder. 

I checked the files like output.hru and output.sub in SWAT scenario folder they were unchanged. 

So  I want to know how can we see the calibrated model results in GIS again???

Regards  

[Natalja C.]

This is explained in the SWAT-CUP user manual on page 66.
I think that section should answer your question.

Best,
Natalja

[Rajveer]

Thanks Natalja for your valuable suggestions that will be very helpful for me.

Regards 

Rajveer

[Rajveer]

Hi Natalja,

I have Calibrated my Base Line Model in SWAT-Cup and got the fitted values. I Replace the default SWAT parameters with these fitted values of the parameters using Manual Calibration Helper and re-run the QSWAT model. 

But i found a mismatch between SWAT-CUP output and QSAWT output.

Why this mismatch is happening?? I used the same calibrated values that were used in SWAT-CUP.

When i used these calibrated parameters with other Land Use then also my NSE is decreased and Base flow get changed ??

Could you please suggest me How to Resolve this issue?? 

   

[pragati sawant]

Hi rajveer did you got the answer regarding how to put calibrated values into swat again??

[galma godana]

Dear

Rajveer

My work is similar to yours. And I need your help. Could you explain how you changed these different land use classes of different years after you simulate your model with the baseline year? I have set my model with one year land use (Baseline period) and finished the swat simulation. Then now I want to use another land use of different year. So would you please explain how you overcome this? 

Thanks in advance!

Galma

[tamnva]

https://www.researchgate.net/publication/328420314_R_script_for_generating_TxtInOut_files_SWAT_for_simulating_dynamic_land_use_land_cover_change_please_see_the_documentation?_sg=e9gVT8xVuSkX-pmcf0YoYZiDdh-tTihFnhUiS7scK-NCuL_VYvNHmrML7Jo2NyAOdLWimaoj0tiILydNkEhyqvw4LIIwlVah-bbufBC_.HeNo2mq0411IItACkbhT5LHN0tbtMTNIE_uaPzPrlAcOcYwzamj8o0hvO62ha6eXx96dJJtuerNvfig9Ogkg4A

[nguyenduy1133]

Dear all,

Here is a way of land use change analysis in ArcSWAT:

https://youtu.be/AhdjLf-Ux0A

I hope it can be useful!

ND Liem

[ngaru...@gmail.com]

Do you find a method to have LULC update in qswat