FLOOD09620 - Laserfiche WebLink

Home Browse Search

JARREIT AND TOMUNSON: REGIONAL INTERDISCIPLINARY PALEOFLOOD METHOD 2973 >- (continued) .. , Velocity, Qgas:lo QIA, DOed, DFB' Age, Q, Difference, Q, Reliability, m 5-1 m3 S-1 % % m3 5-1 km-2 mm mm Type RD Method years years Remarks. Basin 2.4 28 30 0.6 NI 59, W5, M6, LA, B8 100 1000 1,2 ! . 2.1 30 30 0.2 Nl 58, W7, M6, L7, B9 100 1000 1,2,4 3.1 340 25 0.2 FB,Nl 58, W9, M8, L7, B9 5000 =1000 1,2 2.3 158 5.3 HWM 150 gage 2.4 20 30 05 NI 55, W8, M6, 1.5, B7 100 1000 1,2,4 t. tial for dating recent flood surfaces, reveals the least informa- tion because of numerous factors affecting lichen growth. These factors made it difficult to estimate relative ages, and the mcthod is iikely limited to ages considerably iess than 3000 years for most of the study area. The weathering of flood boulders and lichen colonization of boulder surfaces may be complicated by the effccts of forest and range fires [Birkeland, 1984; Bierman and Gillespie, 1991J. If the outer part of the boulder spalls after a fire, the boulder weathering "clock" is essentially reset, and rock-weathering features record the time elapsed sincc the last episode of fire. if spall evidence is prescnt, then rock-weathering and lichen-cover data provide only minimum age estimates. No spalled surfaces on rocks or spall detritus was evident at sites in the study area. Age reli- ability (ranges) for alluvial channels with arroyo development (e.g., site 24 in Table 2) is an estimate of the conservative age. Because of the scope of this investigation, use of field soii morphology and development indexes [Bilzi and Cio/kosz, 1977; Burke and Birkeland, 1979; Meixner and Singer, 1981; Harden, 1982J were not used, but they could provide better age control. No evidence of substantial flooding was found in any inves- tigated stream in Elkhead Creek basin or streams in the north- western Colorado study area. If substantial flooding were com- mon in the study area, evidence should be present. The maximum paleoflood discharge for four sites in Elkhead Creek upstream from Elkhead Reservoir ranged from 79 to 95 m3 S-1 (sites 57, 59, 60, and 61 in Table 2). Considering the estimated total uncertainty associated with each paleoflood discharge ("Q, %" in Table 2), the best estimate is 85 m' S-1 :!: 25% in about 5000 years (:!:1000 years). No tree scarring or flood- transported woody debris was identified on floodplain surfaces, except associated with the Sage Creek Dam failure flood near Hayden (sites 22 and 23 in Table 2) and in lowland areas along the lower Elkhead Creek and Yampa River downstream from about Milner. The peak discharge resulting from failure of the dam was ahout 175 m' s,j (average for sites 22 and 23), which provides an analog for PSIs for a large flood in the study area. Locai residents reported that the dam failed in the mid-1980s from seepage through the dam and was not related to a me- , , , , I I i. i I . teorologic event. This is supported by the maximum paieoflood data of about 3 m' s' j for Sage Creek immediateiy upstream from the dam (site 21 in Table 2). Intermittent streams in northwestern Colorado show little evidence of substantial runoff and are undemt streams for the basin size and broad valleys. An underfit stream is one that appears too small to have eroded the valley in which it flows. Valley bottoms are relatively broad and completeiy covered with native grasses and often have well-developed soils. It is unlikely that a channel could undergo degradation and aggra- dation without leaving any evidence such as terraces and with- out showing evidence of substantial age (thousands of years) on the valley floor. A lack of chaonel development is due to (I) little seasonal snowpack [Doesken et aI., 1984] and tbus reia- tively little snowmelt runoff versus high mountain streams and (2) the basin location above the elevation for substantial rain- fall runoff. It is particularly noteworthy that for many channels having coarse-grained bed material, these sediments have not been mobilized and deposited as flood bars and slack-water depos- its. Data on maximum particle size in the channeis (Dbed) and on flood bars (DFs) presented in Table 2 help demonstrate iack of flow competence. In all but tbe very fined-grained channels, the largest particles in flood bars are smalier than particles available for transport in the cbannels, suggesting that large floods have not occurred during the Holocene. The few in-channel bars that exist are small, have low relief, and par- ticle sizes of cobble or smaller, and suggest insufficient Ilow to mobilize readily avaiiable streambed material (Figures 6 and 7). Similarly, for fine-grain material streams the in-channel bars also are poorly developed and exhibit low relief. Coarse material in the streambed of many streams in northwestern Colorado (derived from conglomerate, basalt, and Precam- brian rocks) is slightly reworked glacial outwash gravel (e.g., Figure 6). Had substantial flooding taken place, coarse streambed material would be transported onto the floodplain [e.g., McCain ef aI., 1979; Jarrett and Costa, 1988; Jarrett, 1990b; Waythomas and Jarrett, 1994] such as shown in Figures 3 and 4, which would be preserved until a larger flood emplaced higher deposits. ~. :.-" ,. L ~ . ,;' ,;