|
<br />PURPOSE AND SCOPE
<br />
<br />In 1995, the Colorado River Water Conservation District undertook a study to reassess the size of the
<br />spillway and flood-control requirements for Elkhead Reservoir on the Elkhead River, which is located near
<br />Craig, Colorado. Generally, the design hydrology for dams and spillways utilize the probable maximum
<br />precipitation and probable maximum flood (World Meteorological Organization, 1986; Hansen and others,
<br />1988; Cudworth, 1989). Because of questions regarding the validity of methods and data used to
<br />estimate probable maximum precipitation estimates (Jarrett and Crow, 1987; Jarrett and Costa, 1988;
<br />Jarrett, 1989, Jarrett, 1990, 1994; in press; Baker, 1994; Jarrett and Waythomas, in press), the Colorado
<br />State En!;lineer organized a task committee to guide development of new criteria to estimate extreme
<br />precipitation and flooding in Colorado. In the interim, site-specific PMP/PMF studies combined with
<br />paleoflood studies are used for design hydrology such as for Elkhead Reservoir.
<br />
<br />The objective of this study was to conduct an interdisciplinary analysis of questionable extreme
<br />rainstorm and flood data. The analyses included onsite paleoflood investigations and a hydroclimatic
<br />evaluation of the available precipitation and streamflow data for sites located within the region. The study
<br />was conducted as part of a site-specific PMP/PMF study of Elkhead Reservoir. A secondary objective is
<br />to enhance the awareness of the need to assess the accuracy of historical flood data, because they are
<br />such a critical factor in flood-plain management, design of hydraulic structures in floodplains, and related
<br />environmental studies. Examples of reassessments of other questionable rainfall and flood data also are
<br />discussed, Finally, the interdisciplinary paleoflood methodology can be used to assess the accuracy and
<br />validity of flood data in other environmental settings.
<br />
<br />BACKGROUND
<br />
<br />On June 3, 1952,5,25 inches of rain was reported to have occurred in about 1.5 hours at the U.S.
<br />Weather Bureau precipitation gage (Cimarron 3SE) near Cimarron, Colorado (U,S. Department of
<br />Commerce, 1952; volume 13), Cimarron is located in central Colorado between the cities of Gunnison and
<br />Montrose near the confluence of the Cimarron River with the Gunnison Rivers (figure 1). The standard 8-
<br />inch precipitation storage gage was located about 3 miles southeast of Cimarron at an elevation of 6,904
<br />feet (figure 2). Daily observations were taken and recorded at 5 pm. The climate at Cimarron is semiarid
<br />and the average annual precipitation is about 13 inches (Colorado Climate Center, 1984; Collins and
<br />others, 1991). Vegetation, primarily grasses, sage, and aspen, cover moderately steep to rugged
<br />mountainous terrain in the lower part of the basin at and near Cimarron. Soil cover is relatively thin in the
<br />lower part of the basin, The June 3, 1952 rainfall amount is one of the largest known rainstorms reported
<br />to have occurred in westem Colorado, The maximum convective rainfall known to have occurred in
<br />western Colorado above about 7,000 feet is 2 to 3 inches (Jarrett, 1987, 1990; Jarrett and Waythomas, in
<br />press). The use, reliability, and transposition of the 1952 Cimarron storm in extreme precipitation studies
<br />in Colorado and elsewhere in the Rocky Mountains has been debated for many years (V.R. Leverson,
<br />Bureau of Reclamation, personal commun., 1995).
<br />
<br />Most approaches used to reassess the accuracy and validity of historic data utilize interdisciplinary
<br />paleohydrologic techniques. Paleoflood hydrology is the study of floods that occurred prior to
<br />systematically gaged observation (Baker, 1987; Stedinger and Baker, 1987). Paleoflood data is obtained
<br />from the study of the movement of water and sediment and their effects on channel morphology and
<br />vegetation on floodplains (Costa, 1987a; Jarrett, 1991). The benefit of paleoflood hydrology is that it
<br />provides information about the number, magnitude, areal extent, and age of past floods (Jarrett, 1990).
<br />Paleoflood studies can best be utilized to provide flood data to complement and extend the length of
<br />gaged data, These studies also provide invaluable data, which is otherwise unattainable, at ungaged
<br />sites or basins. Until about 10 years ago, the accuracy of paleodischarge estimates was about an order-
<br />of-magnitude (Costa, 1987b). Paleoflood discharge estimates made in mountainous streams have been
<br />shown to have an accuracy of about 25 percent (Jarrett and Waythomas, in press; Jarrett and others, in
<br />review). Although evidence of some small to moderate floods are not preserved, generally because larger
<br />floods disturb the deposits, the maximum flood that has occurred since glaciation or about 10,000 to 15,000
<br />years is preserved in channels (Jarrett and Costa, 1988; Jarrett, 1990, 1991). Flood deposits are
<br />preserved in channels in unglaciated areas many tens of thousands (Baker, 1973; Patton and Baker,
<br />1982; Jarrett and Malde, 1987; Jarrett, 1991; Ely and others, 1993; Levish and others, 1994; O'Connor
<br />and others, 1994) to a hundred thousand years (Waythomas and Jarrett, 1994; Jarrett and Waythomas,
<br />in press), In paleoflood investigations, discovering a lack of evidence of the occurrence of extraordinary
<br />floods is as important as discovering tangible onsite evidence of such floods (Jarrett 1987a, 1990; laVish
<br />
<br />2
<br />
|