FLOOD10502 - Laserfiche WebLink

Home Browse Search

Several trial computer simulations using the HEC-l model indicated that the estimates using the Snyder's method are within about 22% of those obtained using the Kirpich method to estimate times of concentration (USBR, 1977), The Kirpich method, though often used, has been found to result in relatively high peak flows (Prakash, 1986), This method was included in the 1977 edition of Design of Small Dams (USBAR, 1977) but its reference has been deleted from the 1987 edition of Design of Small Dams (USBR, 1987). This suggests that the time of concentration simulated by the Snyder's method may be overly conservative (i.e" too small), This method requires two input parameters, i.e., Snyder's standard lag time tp and a coefficient Cpo Physical visualization of these parameters for given watershed conditions is not straightforward, Therefore, previously calibrated values are generally used, This method does not produce the complete unit hydrograph. The method incorporated in the HEC-lIHMS model involves estimation of initial Clark's parameters from input values of Snyder's tp and Cp. A unit hydro graph is then computed using the estimated Clark's parameters. The peak ordinate and time to peak for this unit hydro graph are used to estimate the corresponding Snyder's parameters. If these parameters are significantly different from the input values, then the Clark's parameters are adjusted until the estimated Snyder's parameters closely match the input values. In view of the above, lag times for each sub-watershed are computed using several alternative approaches. These approaches include the following: (i) Kirpich Method (USBR, 1977) k (hrs) = 0.6 x ( 11.9 L3 I H)0.385 in which L = sub-watershed length (miles) and H = elevation difference between the divide and outlet of the sub-watershed (ft). (ii) Soil Conservation Service (SCS) Method (USDA, 1972) k(hrs) = LO,g (S +1)0,71 [1900 --.J Y] in which L = sub-watershed length (ft); Y = average slope of sub-watershed ( percent); S = [{IDOl CN} -10]; and CN = curve number. (iii) Kerby Method ( Kerby, 1959) tdhrs) = 0.01 x (2 L n 13 --.J s )0.467 in which L = sub-watershed length (ft); s = average slope of the sub-watershed (ft/ft); and n = retardance coefficient varying from 0.20 for poor grass, cultivated row crops or moderately rough bare surface to 0.40 for pasture or average grass, Assuming pasture and average grass conditions, a value of 0.40 is used for the sub-watersheds in Frenchman Creek basin, Draft Letter Report Independent Review of Hydrologic Analysis for Frenchman Creek Contract No. OACW 41-00-0-0026-0001 Flood Plain Management Services Special Study Holyoke, Colorado June 18,2001 Page 5