Laserfiche WebLink
<br />Testing the sensitivity of the expected annual damages to the <br />system hydrology (event centerinq) is appropriate and necessary. The <br />alternative to the representative hydrograph procedure is the use of <br />all historical floods of record. However even this more laborious <br />process may introduce some bias in computing expected annual damages <br />if most historical floods were, by chance, centered over a certain <br />part of the basin and not over others. For instance one reservoir <br />site may have experienced several severe historical floods while <br />another site immediately adjacent to that area may, due to chance, <br />not have had any severe floods. <br />While it is possible in the program, HEC-5C, to use only a Single <br />flood event and several proportions (ratios) of that flood in computing <br />expected annual damages, this procedure could introduce considerable <br />bias in the results. A good approach is to use several historical <br />floods with storm centerings throughout the basin and to use several <br />proportions of those floods to obtain flows at the damage centers <br />representing the full ranqe of the flow-frequency-damage relationship <br />for base conditions and for regulated conditions. Another approach <br />is to synthesize events that have consistency in volumes of runoff and <br />peak flows and be reasonably representative regarding upstream con- <br />tribution to downstream flows. Table 1 contains sensitivity information <br />developed in studies of the Susquehanna Basin, Pennsylvania. <br /> <br />29 <br />