Derivation of new design rainfall in Qatar using L-moment based index frequency approach

For stormwater system design, flood estimation and many other environmental assessment tasks, design rainfall is an essential input. Estimation of design rainfall is generally made using a regionalization technique based on a regional database of observed rainfalls. Many countries have derived their own generalized design rainfall data, which are generally expressed in the form of intensity-duration-frequency (IDF) curves. In Qatar, situated in an arid region, the existing IDF data were developed in 1991 using a limited data set. This paper presents the development of new IDF data for the State of Qatar using the method of L-moments and the index regional frequency analysis approach. The daily rainfall data from 32 stations located in Qatar and nearby Gulf countries have been used to form a homogeneous region. It has been found that the Pearson Type 3 distribution best fits the 24-h duration annual maximum rainfall data in the Qatar region. For the ungauged case, a prediction equation is developed where mean annual maximum rainfall is expressed as a function of climatic and physiographic characteristics. From a leave-one-out validation, it has been found that the developed prediction equation can estimate mean annual maximum rainfall with a median relative error of about 5.5%. Finally, an approximate method is used to obtain design rainfalls for other durations due to the limitations of continuous pluviograph data in Qatar. The new set of IDF curves is based on a much bigger dataset than the existing 1991 IDF curves. It is expected that the new IDF curves will have wider application in Qatar and will provide a statistically sound basis for storm water design, flood and environmental studies. The method can be applied to other middle-eastern states and similar arid countries in the world.