FLOOD06897 - Laserfiche WebLink

Home Browse Search

I I I I I I I I I I I I I I I I I I RESULTS Paleoflood Investigations Paleoflood data were obtained at 134 sites for streams draining from the Palmer Divide (table 1). PSI data are readily identifiable onsite by coarse-grained flood deposits (figs. 2a and 3a). Flood deposits typically are coarse to very-coarse, sandy gravel and often have little or no soil- profile development. In contrast, Ihick, clay-rich, fine-grained, well-developed colluvial and alluvial soil (labelled colluvium on figs. 2b and 3b) generally occur in non-flooded areas in the valley. Because these colluvial soils have few particle sizes larger than sand, they were not deposited by main stream flooding; rather, the sediments primarily originated from hills lope (sheetflow) runoff and aeolian (wind-blown) sediments (R. Madole, USGS, pers. commun., 1997; E. Wohl, Colorado State University, pers. commun., 1997). These colluvial organic- rich soils, termed Piney Creek alluvium by Hunt, were dated from a minimum of 1,000 years to greater than 5,000 years (Hunt, 1954). thus, providing a minimum age since floods have inundated these surfaces. These colluvial soils are easily eroded by flood waters, thus, they also provide physical constraints on the present channel geometry (e.g., figs. 2b and 3b). Because streams in the study area are un glaciated, particularly those on bedrock, the age of paleofloods could be many tens of thousands of years old (Jarrett and Way thomas, in press). In addition, numerous large (3-5 ft diameter) cottonwood and ponderosa pine trees are growing on floodplains of many Palmer Divide streams. Although coring trees to determine their age was beyond the scope of this study, general relations of tree age and diameter for different species (Waythomas and Jarrett, 1994; Jarrett and olhers, in review) suggest tree ages in streams in this study range from about 50 to 150 years old. Flood-abrasion scars and accumulations of woody debris on trees are very common when flooding occurs (Matthai, 1969; Osterkamp and Costa, 1987; Hupp, 1988). Flood scars are notably absent along Cherry Creek, Box Elder Creek and their tributaries, except for trees located on the main channel banks. The lack of an accumulation of woody-flood debris around trees located on the floodplain of Cherry and Box Elder Creeks indicates flood waters have not reached those trees since they germinated (-50 to 150 years ago). Conversely. extensive stands of cottonwood trees generally germinate only within a few years after a flood has deposited sediments on a flood-plain surface (Waythomas and Jarrett. 1994; Friedman and others, 1996; Jarrett and others, in review); thus, flood waters have been at least as high as the margin of cottonwood trees in the valley bottom shortly before the trees germinated. Generally, there is very good agreement defining maximum PSI elevation for streams in the study area interpreted from the maximum height of flood-deposited sediments and margins of cottonwood forests. A sensitivity analysis of factors (Manning's n value, channel-bed and width changes) affecting paleoflood estimates was made. Manning's n values were varied by +/-25 percent, which is considered a reasonable range of uncertainty in flow-resistance coefficients (Jarrett and Waythomas (in press). Many streams draining from the Palmer Divide, notably Cherry Creek near the Franktown streamflow-gaging station (06712000) in Castlewood Canyon State Park and the headwaters of Box Elder Creek (Running Creek), have bedrock streambeds that provide a stable-channel geometry for flood estimation. Alluvial-channel geometry can change with time, thus, estimating paleoflood discharge can be affected by channel change. Onsite, maximum-channel width is constrained by undisturbed colluvial soils along the main channel 5