Transcript Slide 1

Analysis of extremes in a changing
climate in support of informed
decisions for adaptation
[email protected]
EURANDOM & KNMI, May 2009
Agenda
• Changing climate extremes
• Traditional practices
• WMO guidance document
• Analyses of extremes
• Take home message
EURANDOM & KNMI, May 2009
EURANDOM & KNMI, May 2009
Changing climate extremes
• IPCC-AR4: ‘confidence has
increased that some extremes
will become more frequent,
more widespread and/or more
intense during the 21st century’
EURANDOM & KNMI, May 2009
EURANDOM & KNMI, May 2009
Changing climate extremes
• Definitions?
– High impact events
– Exceedence of a relatively low threshold (e.g., 95th
percentile of daily precipitation amounts)
– Rare events (long return periods)
– Unprecedented events (in the available record)
EURANDOM & KNMI, May 2009
Changing climate extremes
• Definitions?
– High impact events
– Exceedence of a relatively low threshold (e.g., 95th
percentile of daily precipitation amounts)
– Rare events (long return periods)
– Unprecedented events (in the available record)
• Wide range of space and time scales
– From very small scale (tornadoes) to large scale (drought)
EURANDOM & KNMI, May 2009
Extremes table IPCC-AR4, WG1 report
EURANDOM & KNMI, May 2009
Traditional practices
• Design criteria for safety
of infrastructure are
typically based on
historical observations
of extremes, assuming
a stationary climate
EURANDOM & KNMI, May 2009
Traditional practices
• Design criteria for safety
of infrastructure are
typically based on
historical observations
of extremes, assuming
a stationary climate
• Methods fit extreme value
distributions to selected
observations of extremes
EURANDOM & KNMI, May 2009
Annual maxima of daily precipitation amounts
Station De Bilt, the Netherlands, 1906-2003
EURANDOM & KNMI, May 2009
Map of daily rainfall, 31 July 2002
EVT estimates:
 70 mm: once in
50yr
 50 mm : once in 5yr
EURANDOM & KNMI, May 2009
11 August 2002
EURANDOM & KNMI, May 2009
31 July 2002
24 August 2002
11 August 2002
EURANDOM & KNMI, May 2009
31 July 2002
Traditional practices
• Climate change makes it likely that there will be
change in some extremes that lies outside the
envelope of constant variability assumed under
stationary climate conditions
EURANDOM & KNMI, May 2009
Traditional practices
• Climate change makes it likely that there will be
change in some extremes that lies outside the
envelope of constant variability assumed under
stationary climate conditions
• Adaptation strategies should begin to take into
account the observed and projected changes in
extremes
EURANDOM & KNMI, May 2009
WMO Guidance document
• A new WMO/WCRP guidance document on this
topic is now in press:
– Albert M.G. Klein Tank, Francis W. Zwiers and Xuebin
Zhang, 2009: Analysis of climate and weather extremes in
a changing climate in support of informed decisions for
adaptation. WMO/TD-No ????
EURANDOM & KNMI, May 2009
WMO Guidance document
• Targeted at National Meteorological and
Hydrological Services (NMHSs) around the world
• Aim is:
– to help build capacity to identify and describe changes in
extremes, and
– to improve the information services on extremes under
climate change conditions
EURANDOM & KNMI, May 2009
WMO Guidance document
• The focus is on climate extremes defined as rare
events within the statistical reference distribution
of particular weather elements that are monitored
daily at a particular place, such as temperature and
precipitation
• More complicated weather elements that involve
compound factors, such as tropical cyclones or
storm surges, fall outside the scope
EURANDOM & KNMI, May 2009
Analysis of extremes
• Need long and quality controlled observational
series with high time resolution
3
2
TN, Amos
(Canada)
°C
1
0
-1
-2
1927
-3
1910
1920
1963
1930
1940
1950
1960
EURANDOM & KNMI, May 2009
1970
1980
1990
2000
Analysis of extremes
• Tests for nonclimatic jumps and/or gradual shifts
due to changes in station location, environment
(exposure), instrumentation or observing practices
EURANDOM & KNMI, May 2009
Analysis of extremes
• For analyzing moderate extremes an internationally
coordinated set of descriptive indices can be used,
which describe frequency, amplitude, persistence
• User-friendly R-based software (RClimDex) for their
calculation is available from
http://cccma.seos.uvic.ca/ETCCDI
EURANDOM & KNMI, May 2009
Analysis of extremes
• For analyzing moderate extremes an internationally
coordinated set of descriptive indices can be used,
which describe frequency, amplitude, persistence
• User-friendly R-based software (RClimDex) for their
calculation is available from
http://cccma.seos.uvic.ca/ETCCDI
• One key approach involves counting the number of
days in a season or a year that exceed specific
thresholds
EURANDOM & KNMI, May 2009
Analysis of extremes
upper 10-ptile
1961-1990
the year 1996
lower 10-ptile
1961-1990
EURANDOM & KNMI, May 2009
Analysis of extremes
upper 10-ptile
1961-1990
the year 1996
lower 10-ptile
1961-1990
EURANDOM & KNMI, May 2009
“cold
nights”
Analysis of extremes
“warm
nights”
upper 10-ptile
1961-1990
the year 1996
lower 10-ptile
1961-1990
EURANDOM & KNMI, May 2009
“cold
nights”
Analysis of extremes
De Bilt, the Netherlands
EURANDOM & KNMI, May 2009
Analysis of extremes
• The results (in a
series of journal
papers) contributed
to IPCC-AR4, Ch3
AR4 map of observed trends
(days per decade) for 1951 to 2003
(Alexander et al., JGR, 2006)
EURANDOM & KNMI, May 2009
Analysis of extremes
• The results (in a
series of journal
papers) contributed
to IPCC-AR4, Ch3
– Workshops held
post-AR4
– New workshops
organized in 2009
EURANDOM & KNMI, May 2009
AR4 map of observed trends
(days per decade) for 1951 to 2003
(Alexander et al., JGR, 2006)
Analysis of extremes
Linking the trends in
extremes indices to
regional circulation
changes
Example for Europe
(van den Besselaar
et al., Theor. Appl.
Climatol, in press)
EURANDOM & KNMI, May 2009
Analysis of extremes
• Classic EVT provides a framework for analyzing
extremes further in the tails of the statistical
distributions
EURANDOM & KNMI, May 2009
Analysis of extremes
• Classic EVT provides a framework for analyzing
extremes further in the tails of the statistical
distributions
• Possible to account for “non-stationarity”, but the
best way to do this is still under debate
EURANDOM & KNMI, May 2009
Analysis of extremes
• Classic EVT provides a framework for analyzing
extremes further in the tails of the statistical
distributions
• Possible to account for “non-stationarity”, but the
best way to do this is still under debate
• One option is making the parameters of the GEV
models time-dependent
EURANDOM & KNMI, May 2009
EURANDOM & KNMI, May 2009
Analysis of extremes
• Among others, a user-friendly R-based toolkit
(extRemes) is available from
http://www.assessment.ucar.edu/toolkit
(Stephenson and Gilleland, 2006;
Gilleland and Katz, 2005)
EURANDOM & KNMI, May 2009
Empirical
0.4
0.0
0.2
0.4
0.6
0.8
1.0
20
40
50
Model
Return Level Plot
Density Plot
60
0.00
30
f(z)
50
70
0.04
Empirical
10
Return Level
30
0.02
Model
0.0
10 20 30 40 50 60
Quantile Plot
0.8
Probability Plot
0.1
1
10
100
1000
Return Period
EURANDOM & KNMI, May 2009
10
20
30
40
z
50
60
EURANDOM & KNMI, May 2009
EURANDOM & KNMI, May 2009
160
120
80
40
Empirical
0.2
0.4
0.6
0.8
1.0
40
60
80
100
140
Model
Return Level Plot
Density Plot
f(z)
100
0.000
50
0.010
Empirical
150 200
0.0
Return Level
Quantile Plot
0.0 0.2 0.4 0.6 0.8 1.0
Model
Probability Plot
0.1
1
10
100
1000
Return Period
EURANDOM & KNMI, May 2009
20
40
60
80
120
z
160
EURANDOM & KNMI, May 2009
EURANDOM & KNMI, May 2009
Kharin et al., J.Climate, 2007
EURANDOM & KNMI, May 2009
Take home message
• For providing information services on changing
climate extremes the combined use of different
techniques is recommended, as this will also
provide information on the uncertainties
EURANDOM & KNMI, May 2009
Take home message
• For providing information services on changing
climate extremes the combined use of different
techniques is recommended, as this will also
provide information on the uncertainties
• New infrastructural works should be designed on
the basis of both historical information on changes
in extremes and projected future changes
EURANDOM & KNMI, May 2009
the end
EURANDOM & KNMI, May 2009
http://eca.knmi.nl
EURANDOM & KNMI, May 2009
http://eca.knmi.nl
EURANDOM & KNMI, May 2009
http://eca.knmi.nl
EURANDOM & KNMI, May 2009
Alexander et al.,2006; in IPCC-AR4
EURANDOM & KNMI, May 2009