FLOOD06897 - Laserfiche WebLink

Home Browse Search

12 I I I I I I I I I I I I I I I I I I in East Plum Creek basin (fig. 1). Moist Gulf air that flows through the saddle of the Palmer Divide (fig. 5) encounters these buttes producing localized rain cell:> in East Plum Creek basin. Maximum rainfall amounts decrease by 50 percent within 2 to 4 miles from the buttes (fig. 1), which support the large effect of localized topographic relief on rainfall and flooding. The results of the regional paleoflood study and computer-model simulations suggest that although the Palmer Divide has substantial topographic relief ranging from 2,000 to 3,400 ft in about 140 miles (15 to 24 ft/mi) , the locations of maximum convective ,'ainfall amounts appear to be more related to areas of localized topographic relief. Local relief as little as about 300 It such as the crest of the Palmer Divide, along basin boundaries, in incised valleys. and isolated buttes, which all seem to be formed by resistant bedrock, are at the centroid of maximum rainfall amounts for convective storms. The localized maximum rain cells in eastern Colorado such as those in figure 1 all occurred over areas of local topographic relief. Several computer model simulations have been made by Dr. Ed Torrlinson of Applied Weather Associates and Dr. Jan Paegle of the University of Utah (written commun., 1997), which provide additional independent documentation concerning topograpr ic influences of the Palmer Divide on orographic precipitation. "These simulations were made using the University of Utah mesoscale atmospheric model. The atmospheric conditions used in the simulations were based on atmospheric sounding data associated with the 1965 flooding event along the Palmer Divide. The result of these simulations indicate that: 1.) the Palmer Divide is a significant topographic feature for initiating extreme rainfall events; and 2.) the heaviest rainfall centers are located over the upslope and ridgeline regions of the eastern Palmer Divid3." Their modeling results did not show maximum rainfall amounts in Cherry and Box Elder Creeks. Eastern Colorado has long been believed to be a relatively homogeneous hydrologic area. However, study results demonstrate a large spatial variability of flcoding due to orographic effects from relatively small topographic relief (proposed as more ':han about 300 It in about 3 miles or less). Although large convective rainstorms have produced localized areas of rainfall and flooding in all of eastern Colorado, the largest rainstorms and flooding are closely associated with relatively low-topographic relief features in the Palmer Divide. Flood-frequency estimates for eastern Colorado using regional-regression equations (e.g., McCain and Jarrett, 1976) and various techniques using rainfall-runoff modeling have had large uncertainties due to limited rainfall and streamflow dala, whi.:h contributed to poor understanding about the spatial variability of rainfall-produced flooding in eastern Colorado. The present study results provide important new insight into the spatial variability of extreme rainfall flooding and a paleoflood approach that can help reduce the uncertainty of estimating flood-frequency relations for areas with limited hydrometeorologicai data such as eastern Colorado. Flood-frequency relations for Cherry Creek, which incorporate paleoflood data, indicate the 10,000-year flood ranges from about 44,000 1t3/S (nl~ar Franktown) to about 75,000 ft3/s (near Melvin). A new moments-based approach was recently developed by the Bureau of Reclamation that is more robust in using historical and paleoflood data in flood- frequency analysis (Cohn and others, 1997; England and others, 1998). The Expected Moments Algorithm (EMA) is a simple and efficient method for incorporating historical and paleoflood data. The EMA program was applied to the data for Cherry Creek ga'Jes. EMA results were within about 15 percent of the LPIII analyses.