Conservative design rainfall distribution for application in arid regions with sparse data

This research aims to analyze short duration rainfall data in arid regions, to propose design storm distributions representatives of the recorded storms. Frequency analyses were carried out on all storm durations. Design storm distributions are compared with published depth duration ratios or entire synthetic storm distributions such as the Bell ratios for less than 2-hour durations and ratio of 24-hour to 1-hour rainfall depths embedded in the Soil Conservation Service (SCS) type II storm distribution, known to be the most conservative among SCS distributions. Rainfall-runoff simulations were carried out to determine the impact of the use of storm distributions on generated peak discharges It was found that Bell ratios are adequate to represent rainfall patterns in arid regions for rainfall durations less than two hours. Furthermore, the SCS type II synthetic storm distribution can be safely used for rainfall stations where extreme storms last more than 3 h on the average. This is true for most of the studied locations. Nevertheless, a modified SCS design storm distribution is proposed to represent regions with more concentrated storms. Peak discharges calculated using the modified SCS storm distributions are equivalent or higher than those calculated using the entire Intensity-Duration-Frequency information. (C) 2011 Elsevier Ltd. All rights reserved.