FLOOD03622 - Laserfiche WebLink

Home Browse Search

lU~(O~ ENGINEERING, INC 8100 S Akron street, SUite 300, Englewood, CO 8011('- Phone (303) ('('1080(' I Fax (303) ('('1-4019 March 30, 2001 Planning rlJ-,.,;.I. ~. , "'~ Design Management Mr. Tom Browning, PE. Colorado Water Conservation Board 1313 Sherman Street Denver, Colorado 80203 RE: Grand Junction, Rainfall Depth-Duration-Frequency Curves Dear Tom: In 1992 Henz Meteorological Services (HMS) prepared a rainfall depth-duration-frequency study for Mesa County. This study resulted in a 100-year, 24-hour rainfall value of2.01 inches, compared to a value of2.6 inches in NOAA Atlas 2. The 1992 rainfall was adopted by the County and the City of Grand Junction for the design of storm system improvements. However, this study was never submitted to FEMA for review and approval. We are in the process of getting FEMA approval of the lower rainfall values, and are asking for you concurrence. The following summarizes the analyses to date regarding the rainfall issue. Supporting information is contained in the attached appendices. RAINFALL ANALYSIS The 1992 study by HMS is based on the National Weather Service rainfall records at the Walker Field Airport, for the years 1949 through 1988. The airport is located on the northeast side of the City. Based on an initial meeting with Dr. John Liou, FEMA, the study was updated to include data through 1999. The update to the study includes revised probability curves by HMS based on the Log-Pearson III distribution, as well as probability curves based on the Gumbel distribution prepared by ICON Engineering, Inc. John Liou suggested the Gumbel distribution, and a review of NOAA Atlas 2 revealed the Gumbel distribution was used for that 1973 study. The Gumbel distribution curves computed are for both the raw data, as well as the logarithms of the data. Table I displays the ordered data and the Weibull Plotting Position. Figure I displays the 24-hour and 3-hour data used in the analyses (using the Weibull Plotting Position), and all of the compttted probability curves. Figure 2 is the same as Figure] only without the Henz Log-Pearson III curves (Henz, 2000) and the Gumbel CUrves based on the raw data (ICON, 2001), which do not appear to fit the data as well. Table 2 presents the results of the Gumbel Distribution analyses. Appendix A contains supporting information for this analysis, as well as the HMS analyses. The Gumbel distribution analyses are based on the following references: "Handbook of Applied Hydrology," Ven Te Chow, 1964, and "Statistical Computer Techniques in Hydrology and Water Resources," Jose D. Salas, et aI., 1990. The cumttlative distribution function of the Gumbel distribution is F(x) = exp{- exp[- (x- xo)/ex]), where the scale parameter, ex = 0.7797 (J, the location parameter, Xc = 11- 0.57720, and IL and (J are the data mean and standard deviation, respectively. Substituting and rearranging yield the following equation x ~ II - (0.450043 + 0.7797In(- In(F(x)))) CT, where x is a value with a non-exceedance probability ofF(x). (For example, the 1 OO-year rainfall value, x has a non- exceedance probability of 0.99.)