| Abstract: | 水文頻率分析在水資源之規劃與設計上 扮演十分重要之角色,諸如壩高、堤防高度之 設計流量等之決定,完全取決於頻率分析之結 果。然而頻率分析涉及之問題十分複雜,致使 水文學家對一水文量應屬何種機率分布,至今 仍無一共識。其中,資料偏態係數(Skewness Coefficient)之推估值,因其精確度不高且為 一具偏差性之推估值(Biased Estimator),故 Bobee及 Robitaille(1976)建議一修正公式加 以修正其推估值。同時,虞氏(1990)利用合成 資料(Synthetic Data)研究並指出當偏態係數 修正時,頻率分析所得之結果精確度較精確。 本論文利用合成資料研究偏態係數推估 值之一致性與偏差性,並提出另一修正公式。 同時,研究結果顯示,其精確度較Bobee及 Robitaille所提之修正公式為佳,此新修正方 法,可供水資源規劃與設計之參考。
Frequency analysis plays a very important role on the water resources planning and design. The design discharges of the height of a dam or an embankment are determined by frequency analysis. However, there is no general agreement among hydrologists as to which of the various theoretical distribution available should be used. Bobee and Robitaille (1976) suggested a modification function for the estimated skewness coefficient, because it is not an unbiased estimator. Meanwhile, a synthetic study by Yu (1990) found that the accuracy of frequency analysis is higher when the estimated skewness coefficient is modified. In this paper, the biased property of the estimated skewness coefficient was investigated by using synthetic data. A modification function was then proposed. The results indicated that the accuracy of this proposed function is better than that of Bobee and Robitalle. |
