作者: Gabriele Villarini , James A. Smith , Francesco Serinaldi , Alexandros A. Ntelekos
DOI: 10.1016/J.JHYDROL.2011.01.007
关键词: Environmental science 、 Extreme value theory 、 Hydrology 、 Discharge 、 100-year flood 、 Flood myth 、 Seasonality 、 Heavy-tailed distribution 、 Physical geography 、 Generalized extreme value distribution 、 Climate change
摘要: Summary Annual and seasonal maximum daily discharge time series for 55 stations in central Europe (Germany, Switzerland, Czech Republic, Slovakia) are used to examine flood frequency from a regional perspective. In this study we temporal nonstationarities the peak records, characterize upper tail scaling properties of distributions. There is marked seasonality record, with large fraction annual peaks occurring during winter western part domain, summer southern portion region. The presence abrupt slowly varying changes examined by means non-parametric tests. Change-points mean variance distributions using Pettitt test, while monotonic patterns Spearman Mann-Kendall Abrupt changes, rather than trends responsible violations stationarity assumption. These step can often be associated anthropogenic effects, such as construction dams reservoirs river training. Given profound that these catchments have undergone, it difficult detect possible climate change signal record. estimates location, scale, shape parameters Generalized Extreme Value distribution location scale exhibit power law behaviour when plotted function drainage area, parameter decreases log-linearly increasing catchment area. findings suggest records heavy behaviour.