WSP08834 - Laserfiche WebLink

Home Browse Search

, nJ13~6 17 segments, etc.) for which satisfactory mathematical representation of the physical processes exist. The time distribution of storm water runoff for each drainage unit, and ultimately for the entire watershed,- is then generated by applying the rainfall event and any upstream runoff to these drainage units for each time step. It should be noted that the extent to which the watershed is divided requires e~perience and model- ing judgement; the increased accuracy of a very detailed representation of the watershed may not be worth the increased cost of simulation (9). uempster's (29) development of regional flood-frequency-urbanization equations is a good example of how physically-based rainfall-runoff models can be utilized. Although he does not address the time distri- bution of runoff. the writers feel that the development of general re- lat ionshi ps for the runoff hydrograph (defining dilllensionl ess runoff hydroyraphs for various frequency flood events) could have easily been included in the work. Dempster also illustrates that the use of physically-based models relaxes the assumption that the recurrence interval of the rainfall event must equal the recurrence interval of the runoff event. Unfortunately, the models do not account for the errors that are introduced when approximating the antecedent moisture conditions of the watershed. Regional flood-frequency-urbanization equations similar to Dempster's can be developed for other study regions given the availa- bility of the data listed in Table II-l. The development procedure begins with the verification and calibration (parameter estimation) of some (available) physically-based rainfall-runoff model using regional rainfall-runoff data for watersheds at various stages of development. Storm runoff events from each of these watersheds are then simulated by