final_program_abstracts[1]

11 IMSC Session Program
An analysis of changes in the extremes of temperature,
precipitation, and wind speed based on regional climate
projections for Germany
Friday - Poster Session 7
Thomas Deutschländer 1 , Achim Daschkeit 2 , Susanne Krings 3 and Heiko Werner 4
1
German Meteorological Service (DWD), Department of Climate and Enviromental
Consultancy, Offenbach am Main, Germany
2
Federal Enviroment Agency (UBA), Dessau-Rosslau, Germany
3 Federal Office of Civil Protection and Disaster Assistace (BBK), Bonn, Germany
4 Federal Agency for Technical Relief (THW), Bonn, Germany
Besides earthquakes, weather catastrophes have by far the greatest societal effect of
all natural disasters in terms of civil damages. Therefore, changes in the climatic
extremes are definitely one of the most important aspects of climate change.
However, while an increase in the frequency of hot days and warm nights over land
masses can be regarded as highly probable, e.g. regional projections of heavy summer
precipitation events in Central Europe vary from a slight decrease up to an increase of
30%. Generally, changes in frequencies and intensities of climatic parameters
particularly affecting the general public are still very uncertain on a regional and
especially local scale. Consequently, the Alliance of German Federal Agencies
comprising the Federal Environmental Agency (UBA), the Federal Office of Civil
Protection and Disaster Assistance (BBK), the Federal Agency for Technical Relief
(THW), and the German Meteorological Service (DWD) has recently commenced a
research project aiming to analyze regional climate projections for Germany focussing
on weather extremes and their changes.
The project is subdivided into tow major parts. In the first stage the data will be
examined by means of a selection of climate monitoring and climate change indices
as recommended by the Expert Team on Climate Change Detection and Indices
(ETCCDI) and used within the European Climate Assessment & Dataset (ECA&D)
project. The selected indices were primarily chosen according to their importance for
risk and emergency management. Furthermore, statistical robustness was taken into
accounts, i.e. indices too far into the tail of the distribution are not employed
(typically, the 10 th or 90 th percentiles are used). For the examination of extreme wind
speed changes we principally follow Benistion at al. (2007) [1], who works with 10-m
winter wind speed percentiles, winter sea-level pressure percentiles, and Beaufort
thresholds.
In the second stage we will employ extreme value statistics for the determination of
the modelled changes in return values/return periods for the meteorological
parameters temperature, precipitation, and wind speed. As a great variety of – in our
belief equally promising – approaches for fitting theoretical distributions to empirical
or modelled data exists, we intend to compute several different methods in order to
compare the results. One of the most important issues we will focus our attention on
is the question of what part of the data is best to be used for the fits. Besides the
typically employed block maxima and peak over threshold (POT) approaches, a study
by Jones et al. (2005) [2] proposes that the use of the entire sample (i.e. not only the
data which is extreme in some kind of way like, for instance, annual maxima)
Abstracts 306