Marginal distributions of daily precipitation in Serbia | Marginalne raspodele kumulativnih dnevnih padavina na prostoru Srbije
2010
Mihailović, V., Šumarski fakultet, Beograd (Serbia) | Radić, Z.M., Građevinski fakultet, Beograd (Serbia)
This paper presents a part of results of the continous efforts of the authors to promote, improve and popularize structural analyses of daily hydrological and meteorological series, first introduced by professor Vujica Yevjevich. The main objectives of this paper are: (1) to present a new method which allows the same approach to both meteorological and hydrological series analyses; (2) to determin, in a systematical and objective way, which is the best fitting distribution to describe marginal distributions of cumulative daily rainfall on 25 climatological stations in Serbia for reference period 1961-1990. As precipitation series are discrete stochastic series (with a lot of zero daily values), for the droughts analyses is common to use cumulative rainfall values. In this study cumulative rainfall for past 30, 60 and 90 days (updated daily) are analyzed. The ability of various two and three-parameter probability distributions to describe marginal distributions of cumulative daily precipitation series, and preliminary selection of candidate distributions was performed using L-moment diagrams. Further analysis has shown that neither of three-parameter distributions are suitable for moddeling marginal distributions of cumulative daily precipitation over given range of time scales. Among two-parameter models, rival candidate distributions were gamma (GAM2) and Weibull (W2) distributions. The results of performed goodness of fit tests, (the mean absolute deviation - MAD, the root mean square deviation - RMSD and probability plot correlation coefficient - PPCC), showed that the most suitable model is GAM2.
Afficher plus [+] Moins [-]Mots clés AGROVOC
Informations bibliographiques
Cette notice bibliographique a été fournie par Matica Srpska Library
Découvrez la collection de ce fournisseur de données dans AGRIS