作者: Victor Ntegeka , Patrick Willems
DOI: 10.1029/2007WR006471
关键词: Scale (map) 、 Precipitation 、 Temporal clustering 、 Climate change 、 Quantile 、 Environmental science 、 Block (meteorology) 、 Climatology 、 Series (stratigraphy) 、 Global warming
摘要: Investigation was made on whether the recent historical changes in frequency and amplitude of rainfall extremes can be considered statistically significant under hypothesis no trend or temporal clustering extremes. The analysis based a 107-year time series 10-min Peaks-Over-Threshold data obtained from Uccle station Belgium. Rainfall intensities were aggregated at levels ranging 10 min to monthly scale, defined for different seasons block lengths between 5 15 years using sliding windows. Perturbations derived, which represent empirical quantile changes. Significant deviations quantiles found, persisted periods years. In winter summer seasons, high clustered 1910s-1920s, 1960s recently 1990s. This highlights difficulty attributing "change" climate anthropogenically induced global warming. Research variety phenomena is essential attribution climate.