Transcript Assessment of anthropogenic changes in wind storms and cyclones

Quantitative assessment of
wind storms and extreme extratropical cyclones under climate change
Gregor C. Leckebusch
Special Thanks to:
M. Donat, D. Renggli, T. Kruschke,
K. Nissen, P. Lorenz, T. Pardowitz, U. Ulbrich
Institute of Meteorology
Freie Universität Berlin, Germany
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
0) Outline
Introduction to extra-tropical cyclones
Gaussian and Lagrange Perspective of Cyclonic Activity
Recent climate conditions
Anthropogenic Climate Change
• Objective Wind Storm Identification
• Extreme Value Analysis
•
Anthropogenic Climate Change
Summary
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
1. Introduction
Extra-tropical Cyclones: Crucial part of atmospheric circulation;
Highly relevant for energy transports
Objective Measure of its strengths?
How to deduce information about potential changes of these extreme events?
Which processes influence the variability of cyclone occurrence?
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
1. Introduction
Potential Sources of Variability
~Decades
AMM, PNA,
AMO, ENSO
~Months
SST, NAO,
snow
~Days
Growths factors for
cyclones
latent heat
Cyclone & storm
Baroclinicity
Divergence
Geopotential 500 hPa
~Centenial
Anthropogenic
Change
Adapted fromLutgens & Tarbuck (Eds., 2007)
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
ECMWF Forecast for 20.05.2010 00 UTC
G.C. Leckebusch
2. Gaussian Perspective
MSLP: Stormtrack
NCEP-NCAR, 1961-2000
23 AOGCM Ensemble
Ulbrich et al., 2008
J. Climate
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
2. Gaussian Perspective
1000 hPa Stormtrack: ACC-Signal (SRES A1B)
Ulbrich, Pinto, Kupfer, Leckebusch, Spangehl, Reyers (2008), J Climate
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
2. Lagrange Perspective: Cyclone Tracks
ERA40 Re-analysis: Track density
Winter (ONDJFM), ALL Systems
Ensemble Mean: Track density
Winter (ONDJFM), ALL Systems
1961-2000
Unit: Systems per winter
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
2. Lagrange Perspective: Cyclone Tracks
A1B Climate Change Signal
All Systems
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
2. Lagrange Perspective: Cyclone Tracks
Definition: Extreme Cyclone Systems
Systems with a Laplacian of MSLP (
above the 95th percentile,
)
i.e. only the upper 5% of the distribution are
recognised here
NCEP or ERA40 re-analysis:
most of the historical relevant winter storms are included
(as far as storms are resolved by re-analysis data)
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
2. Lagrange Perspective: Cyclone Tracks
Weighting via: W
A1B Climate Change Signal
Strong Systems
Weight: W2
Leckebusch et al., 2008b
Weight: W4
For the region of e.g. NE Atlantic:
ca. 10-20% increase
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
2. Lagrange Perspective: Cyclone Tracks
Cyclone Intensity (Laplacian of MSLP) for ALL Systems:
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
2. Lagrange Perspective: Cyclone Tracks
European Storm Cyclones
Change Signal GCM-Ensemble
cyclone track density
ERA40, 1961-2000
intensity of storm cyclones
Max.Wind speed during storm days
Isolines: Inter-model std.-dev
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
(Donat, Leckebusch et al., 2010a)
G.C. Leckebusch
3. Wind storms: Identification
Introduction of an objective Storm Severity Index (SSI)
(cf. Leckebusch et al., 2008)
K Ger
SSI T , K Ger  
k
T

t

 max


 v max ( k , t )

 0,
 1   Ak 

v 98 ( k )



3
Event Tracking: related to impacts
98th Percentile
(ONDJFM)
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
m/s
G.C. Leckebusch
3. Wind storms: EVA
• Generalized Pareto Distribution

G ( z )  exp  

g (z) 

 z   

1   
  

d
1 / 



G (z)
ξ : shape
σ : scale
μ : shift
Probability density
function (pdf)
dz
Containing 3 families of tail distributions
•
Rotated Weibull (with upper bound)
ξ<0
•
Gumbel (no bound, exponential decay of pdf)
ξ=0
•
Fréchet (no upper bound, polynomial decay of pdf)
ξ>0
e.g. Coles, An Introduction to Statistical Modeling of Extreme Values (2001)
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
3. Wind storms: EVA
NO
All tracks are considered,
which have at least one
track point within the
according region:
NW
NORTH-EUROPE (NO-EU)
NORTH-WEST-EUROPE (NW-EU)
SO
SOUTH-EUROPE (SO-EU)
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
Extreme value analysis: Ensemble composition and threshold
Percentage of events from single model contribution
to the ensemble
Region: NW-EU
%
Threshold selection:
On average 1 event per
year and model:
= 1024 storms
 SSI threshold: 11.19
SSI-Threshold
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
3. Wind storms: EVA
Total numbers of events in each 30 year period
Events: SSI > 11.19
ENSEMBLES
Multi-model
ensemble in
A1B
scenario
Region:
NW-EU
Model composition for each 30 year period
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
3. Wind storms: EVA
Return Level of a 50-year Event
1971-2000
EVA:
stationary &
non-stationary
model
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
Region: NO-Europe
2069-2098
ENSEMBLES
Multi-model
ensemble in
A1B
scenario
Region:
NW-EU
G.C. Leckebusch
3. Wind storms: EVA
Relative changes of Return Levels by 2098 compared to 1971
Region
RL5
RL10
RL25
RL50
RL100
NOEUROPE
+53%
+59%
+67%
+73%
+79%
NWEUROPE
+13%
+14%
+15%
+15%
+16%
SOEUROPE
-19%
-27%
-36%
-43%
-49%
Red: statistical significant on 95% level
Trends (non-stationary model)
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
3. Wind storms: EVA
ECHAM5-OM1: 3x30 Years of 20C climate conditions
Germany !!!
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch
3. Wind storms: EVA
Results based on 3 x 30 year simulation (ECHAM5-OM1)
SRES A1B scenario
Region: Germany
20112040
20412070
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
20712100
G.C. Leckebusch
4. Summary
Different methods to assess the potential anthropogenic
changes in synoptic cyclone behaviour were applied or newly
developed
Stormtrack: A 23 MME reveals significant increase at the
downstream end of the climatological storm track over the
North-Atlantic under ACC conditions
Cyclone tracks: ACC:
Reduction in the number of all cyclones on an hemispheric scale
Increase in frequency of severe cyclones in winter in certain
“hotspots”, e.g. Northeast-Atlantic
Extreme Value Analysis identifies changes in the statistical
properties of the frequency of severe wind storms
UNESCO Extreme Meeting, Paris, 27.-29.9.2010
G.C. Leckebusch