FLOOD00079 - Laserfiche WebLink

Home Browse Search

hydrologically meaningful criterion, the record shall be treated as coming from one population. 7. Out1iers--Out1iers are data points which depart significantly from the trend of the balance of the data. The retention, modification, or deletion of these outliers will significantly affect the statistical parameters computed from the data. All procedures for treating outliers ultimately require judgement involving both mathematical and hydrologic considerations. The recommended use of a generalized skew coefficient tends to reduce the adverse effect of outliers. The treatment of high and low outliers is described separately. If the station skew is positive, high outliers are considered first. If station skew is negative, low outliers are considered first, and if station skew is near zero, both high and low outliers should be considered simultaneously. If a peak is considered a high outlier, a comparison with historical flood data and flood information at nearby sites should be made. With this information a plotting position is assigned to each outlier and it is then treated as a historic flood peak using procedures described later in this guide. The specific treatment used to handle outliers should become a matter of record. If it is not possible to assign a revised plotting position to the outlier, it is retained in the basic computations as is. The existence of low outliers is to be judged using the following relationship as a criterion: I (Xn-X)/S I > [2.5 + 1.2 log (N/10)] (1-O.4Gl (5) in which Xn = the logarithm of the lowest value (values) in sample of N items X = mean logarithm of all the data G = generalized skew coefficient 5 = standard deviation of logarithm of all the data When the criterion is true, i.e., I (Xn-X)/S I is greater than the expression on the right, then the observation Xn is considered a low outlier. This test is applied to one or more of the smallest values of the systematic 16