期刊名称：Journal of Environmental Hydrology (ältere Jahrgänge)
印刷版ISSN：1058-3912
电子版ISSN：1996-7918
出版年度：2008
卷号：16
出版社：IAEH
摘要：Return period analysis is a common statistical approach used for predicting flooding in urban centers. It is a very important tool for building and designing the safest possible flood control structures. The analysis provides information about when a flood of a given magnitude is expected to occur, be equaled, or exceeded. Flooding is common in urban cities as a result of unplanned development and poor drainage systems and recent flooding in the Accra, Ghana, metropolitan area has been attributed largely to these factors. In this study return period analysis of annual rainfall and maximum 1-month rainfall has been conducted for 30 years of data for the Accra plains. The analysis used the Weibull plotting position. The prediction equations were used to forecast flood magnitudes for return periods of 5, 10, 20, 25, 50 and 100 years. The highest rainfall occurred between May and June. Annual rainfall of 1223.5 mm and maximum 1-month rainfall of 420.6 mm are expected to occur every 31 years. Extrapolation of the return period based on a linear regression model suggests that a maximum 1-month rainfall and maximum annual rainfall of 550.5 mm and 1612.98 mm is expected to occur every 100 years, respectively. The study recommends that urban planners and structural engineers take into account all necessary statistically derived flood magnitudes as inputs during the design of drainage systems in the region.
