<br />20
<br />
<br />LAWN LAKE DAM AND CASCADE LAKE DAM FAILURES, COLORADO
<br />
<br />In general, velocities increased in the downs team
<br />direction, except through Horseshoe Park and down-
<br />stream from river mile 10.28. Maximum depths ranged
<br />from 6.4 to 23.8 ft; maximum widths ranged from 97 to
<br />1,112 ft (table 4). Cross-sectional areas ranged from 710
<br />ft' to 2,980 ft', generally decreasing downstream,
<br />except through Horseshoe Park. Cross-sectional data for
<br />14 cross sections and water-surface elevations from
<br />which depths, widths, and areas were determined are
<br />shown in figures 55 to 70 in the Supplemental Cross-
<br />Section Data at the end of the report. These cross sec-
<br />tions reflect conditions after the flood. Elevations in
<br />most cross sections were surveyed to approximate mean
<br />sea level (from topographic maps), unless stated as be-
<br />ing mean sea level. Field-estimated Manning's n-values
<br />are shown. A question mark follows the values, if con-
<br />ditions (generally scour or debris) indicated a very ap-
<br />proximate estimate.
<br />
<br />FLOOD PROFILES AND BOUNDARIES
<br />
<br />In addition to the indirect-measurement site cross sec-
<br />tions, high-water marks were obtained at 12 cross sec-
<br />tions from which flood profiles could be determined and
<br />are shown in Supplemental Cross-Section Data Because
<br />of the relatively shallow depths of flow and extremely
<br />high-channel gradient (fig. 2), flood profiles cannot be
<br />shown adequately at a realistic scale. The Colorado
<br />Water Conservation Board conducted a study to outline
<br />the flood profiles and boundaries downstream from
<br />Rocky Mountain National Park to Lake Estes (w. P.
<br />Stanton, Colorado Water Conservation Board, written
<br />conunun., 1982). Flood profiles and boundaries for Estes
<br />Park for the July 15, 1982, flood are shown in a report
<br />by the Colorado Water Conservation Board (1983).
<br />
<br />FLOOD FREQUENCY
<br />
<br />It is of interest to know the relative frequency of the
<br />occurrence of a flood. Since this flood resulted from dam
<br />failures rather than natural causes, flood frequency is
<br />not directly applicable. However, knowing the relative
<br />frequency can be useful in illustrating the catastrophic
<br />nature of this type of flood and its channel-changing pro-
<br />cesses, related to historic and natural floods of this
<br />maguitude. A summary of flood frequencies for the
<br />Roaring River and the Big Thompson River is given in
<br />table 5.
<br />Floods in the mountainous regions of Colorado gen-
<br />erally are from three meteorologic causes: (1) Snowmelt
<br />floods, (2) rainfall floods, and (3) rain-on-snow floods.
<br />Many times, runoff peaks on a given stream originate
<br />from all three causes. but conventional hydrologic
<br />
<br />TABLE 5.-Discharge-{requency summary for the Roaring River at
<br />mouth and the Big Thompson River at Estes Park
<br />
<br />Discharge, in cubic feet per second
<br />Roaring River at alluvial fana Big Thompson River at Estes Parkb
<br />¤ce (drainage area, (drain~ area.
<br />intervlll 12.0 square miles) 137 square miles)
<br />
<br />10 year
<br />50 year
<br />100 year
<br />500 year
<br />July 15,
<br />1982
<br />
<br />205 1,540
<br />287 2,020
<br />317 2,200
<br />393 2,620
<br />12,000 5,500
<br />
<br />aBIl.ged on regional regression equation (McCain and Jarrett, 1976).
<br />heased on Log-~ 1YPe III flood-frequency analysis (Interagency Advisory Committee
<br />on Water Data, 1981).
<br />
<br />analysis fails to account for the mixed population of
<br />runoff peaks contributing to the total population of
<br />flood peaks in mixed population flood areas. When
<br />snowmelt- and rain-generated peaks are examined
<br />separately, flood-frequency analysis shows different
<br />trends based on elevation (Jarrett and Costa, 1983). In
<br />the Estes Park area above about 7,500 ft, snowmelt
<br />dominates, with rainfall generally not contributing to
<br />the flood potential for recurrence intervQls greater than
<br />the lOO-year flood. Where rainfall does contribute
<br />significantly to flooding above about 7,500 ft, unit
<br />discharges are small (generally less than 20 ft3/S/mi')
<br />(cubic feet per second per square mile), compared with
<br />lower elevation floods resulting from rainfall (which com-
<br />monly exceed 1,000 ft'/S/mi'). Below about 7,500 ft,
<br />rainfall-produced floods predominate.
<br />The flood-frequency curve for the Big Thompson River
<br />at Estes Park was analyzed for mixed population flows
<br />by the methods described in Jarrett and Costa (1983)
<br />(fig. 13). Inspection of the plotted rainfall and snowmelt
<br />flood-frequency curveS indicates that snowmelt flood
<br />peaks predominate in the Big Thompson River and
<br />tributaries above Estes Park, that is, above an elevation
<br />of 7,500 ft. This was consistent with the results of Jar-
<br />rett and Costa (1983) in which streamflow records of 69
<br />unregulated streams in the South Platte River, the
<br />Arkansas River, and the Colorado River basins were ex-
<br />amined to separate peak discharges from snowmelt and
<br />rainfall runoff during each water year. Above about
<br />7,500 ft, snowmelt is the maj or cause of flooding, not
<br />large, intense rainstorms.
<br />The failure of the Lawn Lake dam resulted in a peak
<br />flow of 5,500 ft'/s at streamflow-gaging station
<br />06733000 Big Thompson River at Estes Park (Site 6),
<br />just upstream from Lake Estes. Data presented in figure
<br />13 indicate that the 1982 dam-break flood was 2.1 times
<br />the 500-year flood for this location. In Estes Park. flood
<br />peaks were 2.5 ft above the level of the 500-year flood
<br />(Colorado Water Conservation Board, 1983). The
<br />
|