Is calibration possible on these observed peaks?

[Justin Goldstein]

Colleagues,
  After performing some manual calibration, I am experiencing great difficulty matching the observed peaks in the attached image.  My watershed is in a semiarid region that experiences convective flash floods.  After the NCDC rainfall stations proved unusable because of the convection, I used nexrad radar data for precipitation.  Yet, a portion of my watershed is situated in a radar '"dead zone," which is leading me to conclude that the radar may have underestimated precipitation for some of the peaks.

    I can take care of the non-observed peaks without any problem through adjusting nutrients, baseflow, etc. but I have found it nearly impossible to match the observed peaks without greatly hindering my overall Nash-Sutcliffe value (currently -0.07).  Even by setting gwqmn to 0, revapmn to 1, alpha_bf to 0.99, rchrg_dp to 1, etc and changing eT and ICN, I only produce a little flow for the peak observed times while greatly oversimulating the non-peak periods.  Increasing cn2 by 10 also hinders my overall nash-sutcliffe although it would help me approach the observed peaks.
 
   I'm not asking for assistance in calibration; all that I'd like to ascertain is if SWAT can match these observed peaks without having to adjust precipitation values (e.g. through realistic fertilization values, etc.) or if I should go in another direction with this research.  I would rather not have to adjust precipitation since there are no realistic sources of the exact precipitation quantities outside of the xmrg NEXRAD radar products, which I am already using.

   Thanks very much for any insights.
   -JG

[Adam Freihoefer]

Justin,

The short answer to your question is no. In my opinion, I would not try to "force" model parameters to correct something that does not appear to be linked to them. If you eliminate the three high event periods, your calibration looks okay and minor parameter adjustments will take care of some of the imperfections. In the case of the three high event periods, it is likely a result of your precipitation inputs. 

Hopefully some else can offer you advice on how to best correct for your lack of precip data.

Cheers,

Adam Freihoefer

Wisconsin Dept of Natural Resources 

[Justin Goldstein]

Thanks so much, Adam!

--
You received this message because you are subscribed to the Google Groups "ArcSWAT" group.

To view this discussion on the web visit https://groups.google.com/d/msg/arcswat/-/ZmQM2LZ1U3oJ.

To post to this group, send email to arc...@googlegroups.com.
To unsubscribe from this group, send email to arcswat+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/arcswat?hl=en.