Laserfiche WebLink
<br />e <br /> <br />e <br /> <br />" <br /> <br />\- <br /> <br />with any continuous simulBtion modBl, such as the HSP model (Hydrocomp, Inc., <br />1976; Crawford, 1971). <br /> <br />In many cases, the length of record fDr streamgage data is only a few <br />years. Rainfall data are often available for much longer periods for the <br />same watershed, and hourly data in the U,S, are frequently available for a <br />period of 35 to 70 years. By simulating a streamflow record with these data, <br />sufficient data for a discharge-frequency relationship can be Dbtained. This <br />relationship is based on the response of the basin during the calibration <br />period, and thus no correction for changes in basin characteristics over the <br />period of record need to be made unless it is known that the <br />discharge-frequency relationship has changed due to urbanization or other <br />effects, such as channelization or water storage projects (Cermak, 1979). <br /> <br />This method has also been used on completely ungaged basins when data <br />are available from adjacent basins with known characteristics to establish <br />the modeling parameters. This type of model requires a similar subdivision <br />of the basin into small subbasins to properly account for the various soil <br />and land use characteristics within the basin. <br /> <br />2.4 Single Event Simulation with Hypothetical Precipitation Data <br /> <br />In many cases, neither streamflow nor precipitation data records are <br />long enough to determine flood-flow return periods that are less frequent <br />than a few years, If it is assumed that the frequency of a given flood is <br />the sBllle as the frequency of the storm producing the flood then hypothetical <br />storms can be developed on the basis of meteorological analyses to give <br />rainfall data for a particular frequency of occurance. <br /> <br />Analysis of total storm rainfall data on a regional basis permits the <br />development of generalized rainfall-frequency curves for relatively large <br />areas. These are usually based on point rainfall-frequency analyses which <br />are adjusted to account for the areal extent of the stOl'l1ls. These <br />generalized storms are usually refered to as "design storms," <br /> <br />The total storm rainfall can then be distributed in time and used as <br />precipitation input to a hydrologic model (such as HEC-l) to determine the <br /> <br />9 <br />