FLOOD09620 - Laserfiche WebLink

Home Browse Search

JARREIT AND TOMLINSON: R!3GIONAL INTERDISCIPLINARY PALEOFLOOD METIlOD River basins, subalpine forests consisting of aspen, lodgepole.' pine, Douglas fir, and Engelmann spruce are common. In most lower parts of Elkhead Creek basin anc;l Yampa River basin below Steamboat Springs, vegetation consists primarily of pi- non pine, juniper, sagebrush, rabbit brush, and na':ive grasses. Mean annual precipitation in the basin varies from about 405 mm at Elkhead reservoir to slightly more than 760 mm near the headwaters [Doesken et al., 1984]. Most annual pre- cipitation falls as snow in the winter months. The largest pre- cipitation amounts are limited to small parts of the basin at the highest elevations around the east and north rims of the basin. Within the larger context of the regional study area, mean annual precipitation ranges from over 1525 mm in the Park Range (the wettest area in Colorado) to 305 mm near Maybell to 406 mm at Meeker [Doesken et 01., 1984]. Frequent, local- ized convective rainstorms occur during the summer months. Convective storms have produced moderate flooding in small, steep basins with little vegetation at lower elevatio:ls in northM western Colorado [JafTett, 1987]. These basins are. located in the western parts of the regional study area, particularly Piceance Creek and Yellow Creek basins. General rainstorms in northwestern Colorado can cover large areas but have not produced substantial flooding in historic times. Flood-frequency analysis has long been done as:;uming staM tionarity of climatic and hydrologic processes [Interagency Ad. visory Committee on Water Data, 1981; NRC, 1999]. There is growing concern that climate naturally varies witn time and that climate may be responding to anthropogenic effects; how- ever, at present (2000), there is little information to assess these impacts on average or extteme conditions [NRC, 1999]. Lills and Slack [1999] analyzed contemporary peak streamflow data in the United States and were not able to detect signifi- cant trends. Knox [1993] documented increases in flood mag. nitude due to climate change in the upper Mississippi River basin, but these results are site.specific. Ely [1997] suggests evidence for climate and flood flow change over the last 1000 years, particularly the last 2 centuries in the scuthwestem United States. England [1998] noted that no expl,.nation was provided for the apparent increase in floods and hat only six of Ely's [1997] 19 sites have records equal to or greater than 1000 years. Of equal importance, is that Ely's [1997] study does not account for large uncertainties in estimating flood dis- charge averaging 60% as noted by Ja"ett [1986, 1S'87, 1994]. Few climate change studies have been done for the Rocky Mountain region. Paleoclimatic studies by Madole [1986], Elias [1996]. Menounos and Reasoner [1997]. Valero-Garces et 01. [1997], Fredlund and Tieszen [1997], Smith and Betancourt [1998], and Vierling [1998] used dendrochronology, pollen, di. atoms, fossil bettle assemblages, carbonate geochemistry, iso- topic evidence, and sediment stratigraphy, respectively, to docM ument mean temperature and precipitation fluctuations. These research results are in good agreement that warming temper- atures resulted in deglaciation of high mountain are as between about 13,000 and 12,000 years B.P. and that mod'~rn climate began about 9500 to 9000 yrs B.P. In this period of modern climate, average temperatures varied :t IOta 20C from cantem. parary mean summer temperatures. Average annual precipi- tation and average annual streamflow have varied by up to about 50% of modern mean values, but nearly similar varia- tions have occurred during contemporary records [Jal7ett, 1991). However, inferring extreme variations from studies of average climate change is problematic. For example, extreme floods in Colorado such as resulted from the 1935 Hale and 2961 1976 Big Thompson floods were imbedded in one of the worst droughts in contemporary records, whereas other extreme floods have occurred during wet periods, such as the 1965 Plum Creek flood [Collins et 01., 1991]. The paleoflood data de. scribed i~ this report provide a means to aSSess the effects of climate change on large floods during the Holocene. 4. Methodology This section describes methods used to estimate paleoflood discharge, determine the paleoflood chronology, analyze re- gional precipitation and streamflow data, and conduct flood. frequency analyses. The strategy of a paleoflood investigation is to visit the. most likely places where evidence of substantial flooding, if. any, might be preserved. Because glaciation and glacial outwash "erases" evidence of floods, paleoflood eviM dence gen"erally can be no older than the time since the last period of glaciation or about 12,000 years ago in the Rocky Mountains [Madole, 1991a]. For unglaciated basins the objec- tive was to identify the largest flood that has occurred within the longest time period but limited to the Holocene when the climate variability has been relatively constant. Thus paleo- flood investigations can identify physical evidence for the OCM currence or nonoccurrence of substantial floods for very long time periods. Paleoflood data were collected for sites on Elk. head Creek and its tributaries and from numerous streams in the Yampa, White, and Little Snake Rivers basins in north- western Colorado. 4.1. Paieoflood Discharge Floods leave distinctive deposits and landforms in and along stream channeis, as well as botanic evidence [Baker, 1987; Baker et 01., 1988; Jal7ett, 1987, 1990b, 1991; Hupp, 1988]. Slack-water deposits of sand-sized particles [Kochel and Baker, 1982; Baker et 01" 1988J, flood scars on trees, erosion scars, and bouldery flood-bar deposits commonly are used as indicators of past flood levels called paleostage indicators (PSIs) (Figure 2). When flows are large enough, streambed and bank mate. rials are mobilized and transported [Costa, 1983; Komar, 1987; Wilcock, 1992]. Such mobilization and transport are a function of channel gradient. As gradient increases, smaller velocities and depths are required to move sediment on the bed of a stream [Costa, 1983]. When stream velocity, depth, and slope decrease, flowing water often is no longer competent to trans- port sediments, which are deposited as flood bars and slack- water deposits in the channel or on the floodplain (e.g., Figures 3 and 4). The types of sites where flow competence decreases and flood deposits commonly are found and studied include (1) locations of rapid energy dissipation, where transported sediments would be deposited, such as tributary junctions, reaches of decreased channel gradient, abrupt channel expan- sions, or reaches of increased flow depth; (2) locations along the sides of valleys in wide, expanding reaches where fine- grained sediments or slack-water deposits would likely be de- posited; (3) ponded areas upstream from channel contractions; (4) the inside of bends or overbank areas on the outside of bends; and (5) locations downstream from moraines across valley floors where large floods would likely deposit sediments eroded from the moraines. Flood-transported sediments and woody debris can scar trees (Figure 5) and also accumulate on trees and other obstructions to provide a good indicator of flood height. The height of tree scars and the top of woody debris are used as indicators of approximate flood height [Har- ~ 'I II I. I