final_program_abstracts[1]

11 IMSC Session Program
Analysis of regional climate projections of extreme
precipitation and temperatures by using an Extreme Value
Theory non-stationary model
Thursday - Parallel Session 7
Barbara Casati 1 , Louis Lefaivre 2 and R. De Elia 1
1
Consortium Ouranos, Montreal, Canada
2 Canadian Meteorological Center, Dorval, Canada
Future climate projections of extreme events can help to forewarn society of highimpact
events and develop better adaptation strategies. Extreme Value Theory (EVT)
provides a well established and robust statistical framework to analyze the behaviour
of extreme events for the present climate and future projections. In this study a nonstationary
model for Generalized Extreme Value (GEV) distributions is used to
analyze the trend of the distributions of extreme precipitation and temperatures, in the
context of a changing climate.
The analysis is performed for the climate projections of the Canadian Regional
Climate Model (CRCM), under a SRES A2 emission scenario, over North America.
Annual and monthly extremes of 24h accumulated precipitation and daily minimum
and maximum temperatures are analyzed. Significant positive trends for the location
of the distributions are found, indicating an expected increase in extreme value
intensities, whereas the scale (variability) and shape (tail values) of the extreme
distributions seem not to vary significantly.
The trends have been tested for three different behaviours: linear, parabolic and
logistic, the latter representing an abrupt change between a present and a future
equilibrium states. Linear trends best fit the changes in the extremes, for both
temperatures and precipitation; temperature maxima exhibit sometimes also parabolic
trends; temperature minima exhibit logistic behaviours in the Arctic regions, possibly
due to a more ice-free future polar environment.
Abstracts 247