Unit Hydrograph generation problem

[maryo]

Hello everyone,
I will start with the fact that I would like to do a floodplain analysis with HEC-HMS and HEC-Ras.
I am a new user to both softwares and I write to you because I can???t figure out the following problem related to HEC-HMS:
I have one rainfall gauge station on my study area (of about 30 km2). The rain data is daily (i.e. one record per day).My flash flood spans for 12 days. I would want to generate the hydrograph on the basis of this rain data, using the Loss SCS Curve Number method and the SCS  Unit Hydrograph Transform method.  I determined the CN following the Purdue university tutorial, along with all the hydrologic parameters needed. I also determined the Basin Lag and the Time of Concentration according to the HEC HMS users??? manual.
My problem is that the hydrograph generated has values 10 to 15 times smaller than the streamflow measured data.
I have no idea at all what do I am doing wrong, considering that a few months ago, on the same river basin, with the same  type of data, I managed to generate a hydrograph that fitted the measured streamflow hydrograph.
Any suggestions would help a lot.
Thanks,
Mary

[MattMead]

Mary,
It could potentially be any number of things.  The first thing I would do is a sensitivity analysis.  Start with your CN values and increase them to well above the values you calculated and run the model.  Compare this to your original model results and see how much difference it made.  Reset the CN back to the calculated value and then do the same with your Tc and then your lag times.  This will give you an idea as to which factor is having the most impact and may require further analysis.  Also, is your 30km2 study area a single basin or is it broken up into subbasins?  30km2 is really large for a single basin model and that may be your issue if that is the case.

[maryo]

hello,
First of all thank you Matt for your suggestions. I am sorry for not being able to post earlier a reply.
Thing is I understand what your suggestion implies and I am interested in the sensitivity analysis, just that I don t know a methodology to do this, and I would need something like a tutorial from the internet. Didn't work a lot with statistics before so I was wondering if you can suggest some software or even more, a useful tutorial in order to investigate my problem the way you said.
Thank you in advance and hope to hear from you soon.
Mary

PS: One more thing, I changed the time interval of the simulations in HEC-HMS and I got a good approximation of the peak flow for my flash floods just that the ascending and the descending time of the hydrograph don't correspond with what I have on field and neither does the volume simulated with the one measured at the gauge station. Any suggestions on that as well? 🙂