Transcript Geen diatitel - European Climate Assessment & Dataset

Global/European Analysis of Extremes
- recent trends Albert Klein Tank
KNMI, the Netherlands
11 June 2002
acknowledgements:
Lisa Alexander (Met Office, UK)
Janet Wijngaard, Aryan van Engelen & Günther Können (KNMI)
36 ECA-participants (Europe & Middle East)
European Study:
http://www.knmi.nl
/samenw/eca
What types of extremes?



Trends in extreme events characterised by the size of
their societal or economic impacts
NO
Trends in “very rare” extreme events analysed by the
parameters of extreme value distributions
NO
Trends in observational series of phenomena with a
daily time scale and typical return period < 1 year
(as indicators of extremes)
YES
Approach



Focus on counts of days crossing a threshold; either
absolute/fixed thresholds or percentile/variable
thresholds relative to local climate
Using derived climate indices as proposed by the joint
CCL/CLIVAR Working Group on Climate Change
Detection (Peterson et al., WMO-TD No. 1071, 2001)
Standardisation enables comparisons between results
obtained in different parts of the world
(e.g. Frich et al., Clim. Res. 2002; also in IPCC-TAR)
Motivation for choice of “extremes”


The detection probability of trends depends on the
return period of the extreme event and the length of
the observational series
For extremes in daily series with typical length ~50 yrs,
the optimal return period is 10-30 days rather than
10-30 years
120
Treturn
Trend  c
N N
100
80
example: 80% detection prob. (5%-level)
60
Event return period
T
40
365 days
100 days
20
30 days
0
10
10 days
20
30
40
Series length
50
N
60
(see also: Frei & Schär,
J.Climate, 2001)
Temperature indices I

Example of “frost days” as an extreme index based on
an absolute temperature threshold
Frich et al. (Clim.Res., 2002) in IPCC-TAR
Temperature indices II

Example of “warm nights” as an extreme index based
on counts of events over a (seasonally varying)
percentile threshold
After: Jones et al. (Climatic Change, 1999)
Yan et al. (…, 2002, IMPROVE- issue)
Minimum temperature at De Bilt
20
20
10
10
“warm
nights”
upper 10-ptile
1961-1990
0
0
the year 1996
-10
-10
-20
-20
1
2
3
4
5
6
7
Month
8
9
10
11
12
lower 10-ptile
1961-1990
Frich et al. (Clim.Res., 2002)
Kiktev et al. (J. Climate, submitted)
“warm nights”
HadAM3
atmosphere
only GCMsimulation
IPCC-TAR (Ch.2, Folland and Karl)
Europe-average trends (86 ECA-stations)
1946-1999
increase in mean temperature: 0.08 (-0.03  0.18) °C/decade;
increase in diurnal temperature range: -0.05 (-0.07  -0.02) °C/decade
cold extremes DEcrease
days/decade
warm extremes INcrease
days/decade
TN10%
2.0 (0.6  3.4)
TN90%
2.5 (0.9  4.1)
TN90%(expect)
2.5
TX10%
0.9 (-0.9  2.7)
TX90%
1.8 (0.0  3.6)
TX90%(expect)
1.0
1976-1999
increase in mean temperature: 0.44 (0.09  0.79) °C/decade;
increase in diurnal temperature range: 0.05 (-0.04  0.13) °C/decade
cold extremes DEcrease
days/decade
warm extremes INcrease
days/decade
TN10%
3.7 (-0.6  8.0)
TN90%
11.2 (5.8  16.6)
TN90%(expect)
4.6*
TX10%
4.1 (-1.7  10.0)
TX90%
10.8 (5.3  16.4)
TX90%(expect)
5.2*
Precipitation indices I
Easterling et al. (BAMS, 2000) in IPCC-TAR
see also Groisman et al. (Clim.Change, 1999)
Linear trends in rainy season over ~50 years
Precipitation indices II

Example of “R95%tot” extreme index for the
precipitation fraction due to very wet days
Precipitation fraction due to very wet days “R95%tot”
1) Identify very wet days using a
site specific threshold = 95th
percentile at wet days in the
1961-90 period
2) Determine fraction of total
precipitation in each year that is
due to these days
3) Trend analysis in series of fractions
“R95%tot”
Annual
amount
Frich et al. (Clim.Res., 2002) in IPCC-TAR
IPCC-TAR conclusions


Temperature extremes
“lengthening of the freeze-free season in most mid- and
high latitude regions”
“reduction in frequency of extreme low monthly and
seasonal average temperatures and smaller increase in
frequency of extreme high average temperatures”
Precipitation extremes
“in regions where total precipitation has increased ... even
more pronounced increases in heavy precipitation events”
“2 to 4% increase in frequency of heavy events in mid- and
high latitudes of the NH”
Key-issues to be discussed I




Are good quality observational series available with daily
resolution and adequate data coverage?
Is there a need for improved methods of homogeneity
testing and gridding the daily series?
Can trend analysis of extremes in daily series go beyond
phenomena that occur on average ~10 times per year?
Can trend analysis of extremes do without standardised
indices if we want a global picture of changes?
YES, but possibly NO
Key-issues to be discussed II


Are the standardised simple indices adequate for:
- improving our understanding of the physical
mechanisms behind changes in extremes?
- evaluating the capacity of climate models to
reproduce occurrences of extremes?
- constructing scenarios of future occurrences of
extremes for impact assessment?
Do we communicate the results of extremes analysis
in such a way that they can be used for climate change
detection/attribution and impact analysis?
???

the end...