SWAT models baseflow trend incorrectly

290 views
Skip to first unread message

Zack McVey

unread,
Jan 27, 2022, 10:25:01 AM1/27/22
to ArcSWAT
Dear SWAT users,

I have been working on my swat models for a while now and I got down to the route of the problem is that SWAT models the seasonal changes in baseflow incorrectly. As it artificially models higher baseflow during the summer period when the baseflow should be lower due to increased evapotranspiration rates. Additionally, the model simulates a lower streamflow during the winter. Any suggestions on which parameters to work with?

Data trend.pngevaporation.png

Stormbrise40

unread,
Jan 28, 2022, 10:38:46 AM1/28/22
to ArcSWAT
I tend to look at the GW and soil storage parameters and peculation through the soil layers. So perhaps these variables. Yes there is increased ET in the summer, but also an ability for the dry soil layers to quickly fill cavities with rain. If it is a strong event, which SWAT cannot predict, dry soils usually have a splash effect and can cause more overland flow and quick times to streams.

As for calibration some people do a quick and dirty calibration with the Curve Number. Is the flow from a gauge?

Jim Almendinger

unread,
Jan 28, 2022, 1:36:48 PM1/28/22
to Zack McVey, ArcSWAT
Zack --
You've got an odd phase shift in your model response.  Actually the model lines up pretty well with the precip data -- low in winter and larger in summer.  I'm curious as to why your observed flows decline over the summer, when the precipitation is ramping up.  Perhaps it is your ET, as you note, but ET is also driven by precip (the more precip, often the more ET because of easy water availability.  Temperature isn't the only factor driving ET.).  You might try calculating a 30-day moving average (or sum) of precipitation (to smooth that signal) and see how that lines up with your observed flows.  Perhaps my eye-ball estimation of what the smoothed precip record looks like (low in winter, high in summer) is not accurate. 

An obvious thing to check is to make sure you haven't got the dates wrong for your observed flows (or precip), by several months.  I realize this is not likely -- but all of us have made such errors before. 

Finally, you might change the GW_DELAY parameter.  This could shift the phase of gw recharge a bit.  If you increase the parameter (its default is about 30 days, I think) it tends to smooth out baseflow and reduce seasonal swings (ups and downs).  If you reduce it, it could increase the amplitude of seasonal ups and downs.  I'm not sure if it will change the phase of the seasonal signal or not.  Maybe, if you're lucky... 

Good luck,
-- Jim



--
You received this message because you are subscribed to the Google Groups "ArcSWAT" group.
To unsubscribe from this group and stop receiving emails from it, send an email to arcswat+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/arcswat/caf4b667-d2a1-4505-bebd-a3b11d4299dan%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages