Abstract
Five parametric Wakeby (L-moment) distribution has been compared with Gaussian, Gumbell and Generalized Extreme Value (GEV) for representing the rainfall and thunderstorm activities over four stations in the Northern areas of Pakistan (viz., Astore, Bunji, Garhi Dupatta and Muzaffarabad) for the period of 1961-2010. The test statistics of Kolmogorov-Smirnov test based on the empirical cumulative distribution function revealed that Wakeby modelling is quite suitable to model rainfall frequencies over all the considered stations while the same distribution also shows better model fittings for Astore, Garhi Dupatta and Muzaffarabad with exceptional case of Bunji where Gumbell distribution has a slightly better fit than the Wakeby. The intent of this paper is to present a contemporary statistical view of rainfall and thunderstorm investigation for the considered stations.
