作者: Milan Gocic , Lazar Velimirovic , Miomir Stankovic , Slavisa Trajkovic
DOI: 10.1007/S12145-020-00543-9
关键词: Generalized Pareto distribution 、 Water cycle 、 Precipitation 、 Mathematics 、 Mean squared error 、 Probability plot 、 Correlation coefficient 、 Generalized extreme value distribution 、 Statistics 、 Distribution (mathematics)
摘要: Precipitation is one of the key components in water cycle. To analyse changes precipitation at a specific location, it necessary to identify distribution that best fits data. For this purpose, three distributions i.e. generalized extreme value (GEV), Pareto (GPD), and logistic (GLO) were fitted annual data collected from 28 meteorological stations Serbia for period 1946–2019 using method L-moment. The goodness-of-fit selected was confirmed L-diagram measures namely relative root mean square error, absolute probability plot correlation coefficient. From results analysis, GEV as fitting Serbia. increasing trends are presented western part can cause higher risks floods than other parts.