FLOOD09620 - Laserfiche WebLink

Home Browse Search

2978 JARREIT AND TOMLINSON: REGIONAL INTERDISCIPLINARY PALEOFLOOD METIlOD . Max paleofloods KIrcher at at. Reservoir envelope curve for I (at dam) Gage \.. ... ...__.~Ik.~,:"d_C'~.~:~~~.9!..._____uu.____u.___u~~'~\.m_m__~",~..-\r:::_u c,-::::'~-==:::::::::: 7 EMA (1=1=5,000 yr.; ave. skew-o.03) 10000 PMF 1000 -!! ~ E .,; e' . ~ u . 'C ~ . . .. 100 .....~... "" I 10 1 1 1 0 100 Recurrence Interval, years 1 000 10000 x Peak oracharge 95% confidence limits EMA (P:S,OOO yr.; station skew) Figure 11. Flood-frequency relations for Eikhead Creek at Elkhead Reservoir [Kircher el al., 1985] with paleoflood data (shown as a reclangle) at the reservoir; near Elkhead gage (0924500) with paleoflood data (shown as a rectangle) using at-site and average skew with 95% confidence limits. P denotes the paleoflood length of record. The envelope curve '/aiue (Figure 9) for maximum flooding in the Holocene is also shown. The flood-frequency relation and probable maximum flood for Elkhead Reservoir (Ayres Associates, 1nc., Fort Collins, Colorado, unpublished data, 1996) are shown for comparison. [1985] that can be used to estimate these relations at an un- gaged site such as Elkhead Reservoir. Kirche- el oi. [1985] developed regression relations for 33 flow characteristics, which include mean annual and mean monthly discharges, flow-duration series, peak discharge, and minimum and maxi- mum 7-day discharges for recurrence intervals from 2 to 500 years. These relations were developed for four hydrologically distinct regions in western Colorado using records from 264 streamflow-gaging stations by relating gage flow characteristics with basin physiographic and climatic explana:ory variables. The regression relations developed from 67 gages in north- western Colorado [Kircher ef ai., 1985, Table 81, which are a function of drainage area and mean annual prec'~pitation, were used to estimate flood-frequency relation for Elkhead Creek at Elkhead Reservoir. The regional regression relations for northwestern Colorado have a mean standard enor estimate of 63% [Kircher el al., 1985]. At Elkhead Reservoir the drainage area is 531 km2, and the mean annual precipitation for the basin is 635 mm. The f1ood- frequency relation for Elkhead Creek at Elkhl~ad Reservoir (Figure 11 and Table 3) was estimated using the work of Kircher el ai. [1985]. The frequency curve in Figure 11 was extended linearly. The 100-year flood estimate for Elkhead Creek is about 61 m3 S-l at Elkhead Reservoir [Kircher et al., 1985] and compares with 64 m' s" at the gage using EMA (Table 3). The estimated recunence interval for the maximum paleoflood (135 m' S") at Elkhead Reservoir is slightly more than 10,000 years (Figure 11) [Kircher ef al., 1985), and the 10,000-year flood is about 110 m' S'I (Table 3). The 10,000- year flood estimate is 104 m' s" at the Elkhead gage (Table 3). The flood-frequency relation for the Elkhead Creek gage (Figure 11 and Table 3), which has about 30% of the draioage area at the reservoir, is similar to the regional relation. The similarity of the relations are likely because the majority of the basin bet\Veen the gage and the reservoir contributes little additional peak runoff from snowmelt and rainfall. Also shown on Figure 11 is the regional envelope value of maximum flooding for Elkhead Reservoir from Figure 9. For comparative purposes the PMF estimate shown for Elkhead Creek at the dam (Ayres Associates Inc., written communica- tion, 1996) is 1020 m' S-1, which is 4 times larger than the envelope curve discharge value for Elkhead Reservoir, and the recurrence interval for a flood the magnitude of the PMF exceeds 10,000 years (Figure 11). The maximum paleoflood in the last 5000 years for Elkhead Creek near Elkhead Reservoir is about 13% of the site-specific PMF estimate. 6. Discussion Paleoflood techniques and rainfall-runoff modeling (includ- ing PMP/PMF methods) have inherent assumptions and limi- tations that produce uncertain flood estimates. Although pa- leoflood estimates also involve uncertainties, the estimates are based on interpretations of physical data preserved in channels and on floodplains during the past 5000 to 10,000 years in northwestern Colorado. Paleoflood uncertainties primarily are related to possible postflood changes in channel geometry and flood heights interpreted from PSIs. Where possible, paleo- flood estimates are obtained in bedrock-controlled channels that minimize changes in channel geometry; there is little ev- idence that major changes in channel geometry have occurred in alluvial channels in the study area. Because beds of most ..., , , I I , i t < ~ i. J i I . ~ f ! '. , i