Author(s): S. Sadri; H. Madsen; P. S. Mikkelsen; D. H. Burn
Linked Author(s): Henrik Madsen
Keywords: Climate change; Trend analysis; Extreme rainfall; Mann-Kendall test; Non-stationarity; Cross-correlation
Abstract: Changes in precipitation patterns can be seen as evidence of climate change. In recent years, flooding of urban areas in Denmark suggests changes have been occurring in precipitation patterns. Trends in hydroclimatic data can be considered on two spatial scales: at individual sites and regionally. This paper focuses on analyzing extreme rainfall data from 1979-2007 at 66 stations in Denmark. The rainfall intensity data were analyzed for 11 durations ranging from one minute to 48 hours; for each duration, the partial duration series method was used to extract extreme rainfall events. The Mann-Kendall (MK) non-parametric test was then used both for at-site and regional analysis to detect statistically significant trends. For the analysis of the entire region, the effect of spatial cross-correlation of the data for the 66 stations was accounted for. On a regional scale, rainfall intensities of 2,5, and 60 minute durations showed a significant increasing trends at a 5% significance level. On an at-site basis, out of the 726 data sets, 51 showed a significant trend: 39 increasing and 12 decreasing. This methodology is applicable in other parts of the world with similarly long extreme precipitation records.