Use HEC-1 model in the WMS package. It is a relatively simple model that can give you the results you required as long as it is calibrated properly. 

Use the surface-baseflow separation analysis. A simple stright line from baseflow at the beginning of the storm to the end of the surface hydrograph works well for basins with deep aquifers and negligible subsurface flow. Draw an increasing line at the beginning of the surface flow hydrograph to the centroid of the hydrograph and then the line declines to meeto the residing surface flow hydrograph. This method works well for catchments with aquifers near the soil surface and where subsurface flo and, saturated overland flow are common flow processes.

Once you draw the line just divide the surface flow hydrograph in discrete time intervals; evaluate Q for each time interval and sum for the entire surface flow hydrograph. Divide the Qtotal in cm (by dividing Qtotal/area of watershed) between total precipitation. This index can be contrasted to the indexes of the other basins in the area to find hydrological differences associated to different vegetation types or different geological attributes.

Hope these comments help you to clarify your concerns

Kind Regards

Sincerely

Jose Navar, Ph.D

Thank you all. I find your response particularly helpful, Jose - that is the method I decided to use.  

Regards,

Luther

hello Dear Luther,  i suggest to use recessions curves, Qb=Q.k^t , wich k is the recession factor of the hydrograph of the event.

you can find theses papers helpful

good luck

https://pubs.usgs.gov/sir/2012/5034/sir2012-5034.pdf

https://www.mdpi.com/2073-4441/5/4/1797/pdf

Article Surface/groundwater interactions in the Bani and Nakambe riv...

Article A Regression Model to Estimate Regional Ground Water Recharge