FLOOD09736 - Laserfiche WebLink

Home Browse Search

, for rain and snowmelt events for a series of specific discharge values until the composite curve is defined over the desired frequency range, For example, consider a discharge of 3,000 cfs for a basin with a drainage area of 213 square miles and mean watershed elevation of 9,000 feet, The probability of an event of this magnitude being caused by rain is P(R) = 0.130 and by snowmelt is P(S) = 0.070. The combined probability for this event computed by Equation 3 is as follows: P(RUS) = 0,130 + 0.070 - (0.130)(0.070) = 0,019 . Peak flow for a return period of 100 years is plotted versus drainage areas for different mean watershed elevations in Figure 13 to illustrate the impact of watershed elevation on the 100-year flood. The significance of basin elevation becomes more pronounced for lower discharge frequencies. Sm~MARY AND CONCLUSIONS Stream gauging stations offer more information than just peak flows. By researching the original charts and identifying the cause of flooding, a separate annual series for each type of flood event can be made, The sta- tistical parameters which describe the probability distributions for each separate event type can be computed to describe frequency-discharge curves for each type event at the gauged site, The curves may then be statistically com- bined to give a composite frequency-discharge curve. . In a regional approach, several gauged subbasins which compose a homoge- neous region are first analyzed individually by separating the record into independent events, By regression analysis, the statistical parameters com- puted for each event type at each station are related to measurable subbasins and climatic variables. The statistical parameters are given weights equal to the station's length of record in the regression analysis. If the basin or climatic variables in a given site are known, the regional statistical parameters can re determined from the regression equations, and discharge-frequency curves for each independent population can be computed, After these curves for the independent events are statistically combined, the results will be compatible with other records in the region. The discharge-frequency curve may be adjusted to reflect long-term records in the vicinity of the site if present. The statistical parameters which were used to define a log-Pearson type III distribution in the regional analysis included the mean logarithm, coef- ficient of variation of logarithms, skew coefficient of logarithms, and skew coefficient of logarithms of peak flows. For rain events, it was found that the mean increases as the drainage area increases while the coefficient of variation and skew remain constant, For snowmelt events, the mean was found to increase as the drainage area increases and as the mean watershed elevation increases. Like rain events, the coefficient of variation and skew were found to be constants for snowmelt events, In all applications of the method, the peak discharges were found to increase with drainage basin area although the runoff per unit area tended to. decrease, Three factors are important in generalizing about the differences between snowmelt and rainfall type floods in Colorado, These are frequency, draiange basin size, and mean watershed elevation. In general for a given drainage basin area and mean watershed elevation, rain events tend to dominate the composite curve for rare events. For a given / 18