Estimation of One Day Probable Maximum Precipitation for Fafan Zone of Somali Region, Ethiopia.

Abstract

Drought and floods are common in Fafan zone, since study of probable maximum precipitation is exceedingly imperative for design of
water resources project. Development of one day probable maximum precipitation (greatest depth of precipitation for a given duration
which is physically possible over a given size storm area at a particular geographical location and at a particular time) for Fafan zone
using daily annual extreme values of 5 stations by statistical method of Hershfield formula were implemented. The objectives are to
estimate point probable maximum precipitation (PMP) for station and their return period and to select the best fit probability distribution function. Missing data were reconstructed and inconsistency was checked using normal ratio method and double mass curve analysis,respectively. Frequency factor (Km) and PMP were derived using annual maxima, means and standard deviation. Km values varied from 2.93 (Gursum station) to 7.08 (Jigjiga stations) and PMP varied from 74.21mm (Harshin station) to 189.82mm (Awbare station) with an average 126.09mm. The ratio of one day PMP and highest observed rainfall varied from 1.07(Gursum station) to 1.42 (Jigjiga station) with an average 1.19. Normal, Log normal, Log pearson type III and Gumbel distributions were used to predict extreme values. The obtained values from probability distribution functions were tested by chi-square (χ2) test and coefficient of determination (R2). Results revealed that the log Pearson type III distribution performed the best (50%) and Gumbel performed second (34%) was the next. PMP estimates for one day durations using log pearson type III have minimum, maximum and average value of 2127.66, 3225.81and 2690.78 return period of years, respectively and the observed variability was found as 5.5%. The depths of rainfall for 5, 10, 50, 100, 1000 and 10000 return periods were found to vary between 37.07 mm and 241.49 mm. The predicted PMP values to a depth of various years return period ratios were found to vary between 0.76 (at 10000 years) and 2.89 (at 5 years). The estimated PMP, which is very uncertain values for 100, 1000, and 10000 years and reasonable for designing of hydraulic structures for return periods in the order of between 5, 10, and 50 years for some areas. In order to get very precise results meteorological stations has to be distributed evenly with adequate data range (quality) and length and the study needs to be updated with recent information.