Laserfiche WebLink
<br />results, For more complex hydraulic conditions, two-dimensional models also can be used <br />(Miller and Cluer, 1998), <br /> <br />Flood Frequency Analysis <br /> <br />Paleoflood information is used in flood frequency analysis to extend systematic and <br />historical record length and to gather information about the largest floods that have occurred at a <br />particular site, Fitting a statistical distribution to the composite flood record allows estimation of <br />flood magnitudes with annual exceedance probabilities smaller than 0,0 I, The choice of a <br />distribution function depends on the nature of the data, experience of the analyst, and established <br />scientific practice, <br /> <br />Several general approaches are available for estimating the parameters of a distribution <br />including the method of moments (IACWO, 1982), method of maximum likelihood, and <br />expected moments algorithm (Cohn et aI., 1997), The method of moments is not usually used <br />with data sets that include paleoflood information, The expected moments algorithm is a <br />modification of the method of moments approach that utilizes historical and paleoflood <br />information in a censored data framework, Maximum likelihood estimation procedures can also <br />be used to integrate systematic, historical, and paleoflood information (Stedinger et aI., 1993), <br /> <br />For cases where sufficient at-site information does not exist, regional flood frequency <br />approaches may be used to detennine the frequency of rare flood events (Jarrett and Costa, <br />1988), The National Research Council (1988) has proposed several general strategies including <br />substituting space for time for estimating extreme floods, In substituting space for time, <br />hydrologic information at different locations in a homogeneous region are used to compensate <br />for short records at a single site, Several approaches have been considered for regional flood <br />frequency analysis including the average parameter approach, index flood method, and <br />regression techniques (Cudworth, 1989). <br /> <br />Interpretation Issues <br /> <br />The monograph will address interpretation issues associated with systematic, historical, <br />and paleoflood data; hydraulic models; and flood frequency analysis approaches, The quality <br />and uncertainty of hydrologic data impacts the ability to predict extreme floods, The <br />assumptions inherent in modeling, either 1-0 or 2-D, need careful evaluation to detennine <br />appropriate interpretation, uses, and limitations of the resulting hydrologic infonnation, <br /> <br />Several issues are commonly raised in using paleoflood information for estimation of rare <br />floods (with annual exceedance probabilities smaller than 0,01), Caution must be used for <br />paleoflood reconstructions to address issues of channel stability, which can be addressed by <br />selecting bedrock or stable channels, and flow resistance changes for past floods (such as climate <br />induced riparian vegetation changes), Another problematic issue is that paleoflood data may <br />come from a probability distribution in the past that does not represent the current distribution, <br />Climate variability can present problems for all types of hydrologic data, Flood frequency <br /> <br />5 <br />