Laserfiche WebLink
underestimated v a I ue increases with the degree of scatter about the <br />rating curve and can reach as high as f ifty percent (Ferguson, <br />1986). The statistical regression model for the sediment load Qs <br />(tons/day) is: <br />Qs = a Qb (1) <br />where a = regression coefficient <br />b = regression exponent <br />Q = water discharge (cfs) <br />This model can be improved by applying an unbiased correction factor: <br />C = e(2.65s2) (2) <br />n <br />where s2 = E (log Qs - log Qs )/(n-2) (3) <br />i=1 m c <br />and <br />Qs = measured sediment load <br />m <br />Qs = calculated sediment load (predicted from eq. (1)) <br />c <br />n = number of data poi nts <br />The bias correction Is made by multiplying the regression coefficient a <br />by the correction factor C. This simple correlation based on statistical <br />considerations removes most of the bias when the log-log rating plot Is <br />approximately linear with normally distributed scatter. It improves the <br />accuracy of the sediment load estimate. When the average of the measured <br />sediment I oad for al I three databases are compared w ith predicted val ues, <br />further corrections can be made by adjusting the coefficient a. <br />With updated and revised databases it was possible to determine a base <br />flow for both the Little Snake and Yampa Rivers separately. The water year <br />was divided into a base flow period (September 1 to February 28) and high <br />f I ow period (March 1 to August 31). The mean flow for this f al I and winter <br />period for each river was determined using the entire period of record. <br />This mean f I ow was designated as a base flow for this analysis. <br />To analyze potential impacts of flow reduction in a method that retains <br />the shape of the seasonal hydrogr aph, exceeda nce probability by drogr aph s <br />were developed. The exceedance hydrographs were computed from the daily <br />discharge record for each river, Yampa at Maybel I and L i ttl a Snake at L i l y <br />gaging stations, based on Wiebul I probabil ity distribution. An exceedance <br />probability hydrograph Implies that for the given percentage exceedance the <br />f low w i I I be equal to or greater than that corresponding discharge (e. g. 75 <br />percent exceedance probability means that three out of every four years the <br />f low w i I I be eq ua I to or greater than the indicated f I ow for that day). To <br />determine a particular exceeda nce by drogr aph the daily flows for every day <br />5