Shallow aquifer storage in SWAT

[Henning]

Hello All!

I am working with SWAT2005 on a watershed in Morocco. Since the
groundwater table is very low, I decided to minimize revaporation, by
decreasing gw_revap or increasing reavapmn. But the water seems to
remain in the aquifer instead of forming baseflow (See attached file;
I only changed gw_revap and reavapmn). Since this is not a realistic
behaviour, I was wondering wether SWAT could not "handle" no
revaporation conditions or am I missing other important/related
parameters?

Kind regards,

Henning Busche
Hydrology Research Group
University of Bonn

http://groups.google.com/group/swatuser/web/HRU_excerpt.xls

[willem vervoort]

My guess would be to decrease your groundwater storage, make the
aquifer smaller, or decrease the threshold for baseflow to occur. Have
a look at GWQMN: 2:4.2.3 GROUNDWATER/BASE FLOW
Willem

--
http://blogs.usyd.edu.au/waterhydrosu/

[Henning]

I also tried different initial values for aquifer storage and gwqmn is
set 0. The recession looks strange. My understanding of the baseflow
procedure is, that baseflow should be the same, everytime aquifer
storage is the same. This seems not to be the case. Does anyone work
in an arid region with low groundwater tables? If you turn off revap,
does the shallow aquifer storage react similar (check output.hru, non-
limited output)?
Henning

[Jim Almendinger]

Dear Henning --

This sounds similar to a problem we had with water infiltrated from surface-water bodies (ponds, wetlands, and channels) -- SWAT puts this water directly into shallow aquifer storage rather than considering it recharge.  The infiltrated water appeared to be "trapped" in shallow aquifer storage and was not released to baseflow, at least not that we could see during the period of our model runs.  We also made sure our gwqmin was zero.  I've looked at the theory manual and some baseflow should be released from shallow aquifer storage -- but it doesn't seem to be.  In any case, it seems that once SWAT puts water into shallow aquifer storage, it mostly stays there.  

Claire Baffaut also noted the problem, and she has written a modified version of SWAT that should fix the issue, at least for infiltrated ut your issue specifically).  I think her solution was to add the infiltrated water to the recharge component -- from which it can directly contribute to baseflow.  I'm not sure if it deals with the parallel problem, of how water is released from shallow aquifer storage, once it gets there.  I haven't tried Claire's version quite yet (I hope to soon), but you might get a copy of her version of swat20005 from her (Claire....@ARS.USDA.GOV) or Nancy Sammons, and give it a try.  

Cheers,

-- Jim

On Oct 22, 2008, at 4:11 AM, Henning wrote:

Hello All!

I am working with SWAT2005 on a watershed in Morocco. Since the
groundwater table is very low, I decided to minimize revaporation, by
decreasing gw_revap or increasing reavapmn. But the water seems to

remain in the aquifer ow (See attached file;

I only changed gw_revap and  reavapmn). Since this is not a realistic
behaviour, I was wondering wether SWAT could not "handle" no
revaporation conditions or am I missing other important/related
parameters?

Kind regards,

Henning Busche
Hydrology Research Group
University of Bonn

Dr. James E. Almendinger, Senior Scientist

St. Croix Watershed Research Station

Science Museum of Minnesota

16910  152nd St. N

Marine on St. Croix, MN  55047

tel: 651-433-5953 X 19fax: 651-433-5924

email: din...@smm.org

web: www.smm.org/SCWRS/

[Henning]

Dear Jim!

It is exactly the same problem. I checked the source code:

Transmission losses from tributary channels are added to storage:
tran.f; line 135: shallst(j) = shallst(j) + tloss

Whereas baseflow is calculated without regarding storage:
gwmod.f; line 102: gw_q(j) = gw_q(j) * alpha_bfe(j) + (rchrg2 -
gwseep ) * (1. - alpha_bfe(j))

This problem can be overseen under revap conitions, because
revaporation is subtracted from storage
gwmod.f; line 113: shallst(j) = shallst(j) - revapday

I am now changing the source code to add transmission losses to
recharge as well.

Thanks for your help

Henning

On 22 Okt., 16:45, Jim Almendinger <din...@smm.org> wrote:
> Dear Henning --
> This sounds similar to a problem we had with water infiltrated from  
> surface-water bodies (ponds, wetlands, and channels) -- SWAT puts this  
> water directly into shallow aquifer storage rather than considering it  
> recharge.  The infiltrated water appeared to be "trapped" in shallow  
> aquifer storage and was not released to baseflow, at least not that we  
> could see during the period of our model runs.  We also made sure our  
> gwqmin was zero.  I've looked at the theory manual and some baseflow  
> should be released from shallow aquifer storage -- but it doesn't seem  
> to be.  In any case, it seems that once SWAT puts water into shallow  
> aquifer storage, it mostly stays there.
>
> Claire Baffaut also noted the problem, and she has written a modified  
> version of SWAT that should fix the issue, at least for infiltrated  

> waters (I'm not sure about your issue specifically).  I think her  

> solution was to add the infiltrated water to the recharge component --  
> from which it can directly contribute to baseflow.  I'm not sure if it  
> deals with the parallel problem, of how water is released from shallow  
> aquifer storage, once it gets there.  I haven't tried Claire's version  
> quite yet (I hope to soon), but you might get a copy of her version of  

> swat20005 from her (Claire.Baff...@ARS.USDA.GOV) or Nancy Sammons, and  

> give it a try.
>
> Cheers,
> -- Jim
>
> On Oct 22, 2008, at 4:11 AM, Henning wrote:
>
>
>
>
>
> > Hello All!
>
> > I am working with SWAT2005 on a watershed in Morocco. Since the
> > groundwater table is very low, I decided to minimize revaporation, by
> > decreasing gw_revap or increasing reavapmn. But the water seems to

> > remain in the aquifer instead of forming baseflow (See attached file;