Assessing the adequacy of probability distributions for estimating the extreme events of air temperature in Dabaa Region

الملخص EN

Assessing the adequacy of probability distributions for estimating the extreme events of air temperature in Dabaa region is one of the pre-requisites for any design purpose at Dabaa site which can be achieved by probability approach.

In the present study, three extreme value distributions are considered and compared to estimate the extreme events of monthly and annual maximum and minimum temperature.

These distributions include the Gumbel/Frechet distributions for estimating the extreme maximum values and Gumbel/Weibull distributions for estimating the extreme minimum values.

Lieblein technique and Method of Moments are applied for estimating the distribution parameters.

Subsequently, the required design values with a given return period of exceedance are obtained.

Goodness-of-Fit tests involving Kolmogorov-Smirnov and Anderson-Darling are used for checking the adequacy of fitting the method/distribution for the estimation of maximum/minimum temperature.

Mean Absolute Relative Deviation, Root Mean Square Error and Relative Mean Square Deviation are calculated, as the performance indicators, to judge which distribution and method of parameters estimation are the most appropriate one to estimate the extreme temperatures.

The present study indicated that the Weibull distribution combined with Method of Moment estimators gives the highest fit, most reliable, accurate predictions for estimating the extreme monthly and annual minimum temperature.

The Gumbel distribution combined with Method of Moment estimators showed the highest fit, accurate predictions for the estimation of the extreme monthly and annual maximum temperature except for July, August, October and November.

The study shows that the combination of Frechet distribution with Method of Moment is the most accurate for estimating the extreme maximum temperature in July, August and November months while the Gumbel distribution and Lieblein technique is the best for October.