Transcript Presentation of ENSEMBLES (FP6)

ENSEMBLES
GMES - ENSEMBLES 2008
Page 1
The ENSEMBLES Project
 Began 4 years ago, will end in December 2009
 Supported by €15M of European Commission funding,
coordinated by Met Office Hadley Centre
 67 partners from across EU, Switzerland, Australia, US
we welcome requests from new groups to participate (unfunded)
 Collaborates with other international projects
 Brings together a wide range of climate change-related
research communities
GMES - ENSEMBLES 2008
Who, where, when
Page 2
Output / products:
 Multi-model RCM projections for Europe at 25km
 Significant contribution to IPCC AR4
 Investigating probabilistic methods for “s2d2c” timescales
for range of models to explore impacts
 Earth system model simulations, will use a mitigation
emissions scenario developed within ENSEMBLES
 Providing information relevant to IPCC planning
 Web-based statistical tools for finer scale information
 Daily gridded dataset for Europe, + uncertainty estimates
GMES - ENSEMBLES 2008
Page 3
Project output 2004-2007
150 publications
20 book contributions
Approx 500 conference presentations
These numbers do not include the planned
project output (conferences, workshops etc)
GMES - ENSEMBLES 2008
Page 4
ENSEMBLES Strategic Objectives
Project Goal:
Overall goal is to maintain and extend European preeminence in the provision of policy relevant information on
climate and climate change and its interactions with society

Develop an ensemble prediction system based on global
and regional climate models, validated against
observations and analyses, to work towards a probabilistic
estimate of uncertainty in future climate at the seasonal,
decadal and longer timescales

Quantify and reduce uncertainty in the representation of
physical, chemical, biological and human-related feedbacks in
the Earth System

Exploit the results by linking the outputs to a range of
applications, including agriculture, forestry, health, energy,
water resources, insurance
We are aiming to increase availability of scientific knowledge and provision
of relevant information related to the impacts of climate change
How
GMES - ENSEMBLES 2008
Page 5
Comparison of calculated storm loss based
on ERA-40 with insurance data for
Germany. Correlations between calculated
loss and insurance date rate between 0.85-0.9
GMES - ENSEMBLES 2008
Page 6
Seasonal prediction of tropical
cyclones
Tropical Storm Frequency over the North Atlantic (JJASO)
Forecast starting on 1st May
Tropical Storm Frequency over the Eastern North Pacific (JJASO)
Forecast starting on 1st May
Correlation=0.62( 0.99)
Correlation=0.59( 0.98)
MULTIMODEL: ECMW F LODYC UKMO CNRM CERFACSRMS
MPI
SCNR
Error=
2.93( 3.65)
FORECAST
FORECAST
2 Standard Deviations
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
Tropical Storm Frequency over the western North Pacific (JJASO)
Forecast starting on 1st May
Correlation=0.72( 1.00)
MULTIMODEL: ECMW F LODYC UKMO CNRM CERFACSRMS
MPI
SCNR
Error=
2.73( 3.93)
1987
1988
1989
1990
1991
1992
FORECAST
39
38
37
36
35
34
33
32
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
1993
1994
1995
1996
1997
Observations
Year
1998
1999
2000
2001
Correlation=0.62( 0.99)
MULTIMODEL: ECMW F LODYC UKMO CNRM CERFACSRMS
MPI
SCNR
Error=
2.50( 3.17)
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
Year
Observations
FORECAST
1998
1999
2000
2001
2 Standard Deviations
South Pacific
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
Tropical Storm Number
Tropical Storm Number
2 Standard Deviations
Tropical Storm Frequency over the South Pacific (DJFMA)
Forecast starting on 1st November
2 Standard Deviations
Western North Pacific
Observations
Eastern North Pacific
Tropical Storm Number
North Atlantic
Tropical Storm Number
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
Observations
MULTIMODEL: ECMW F LODYC UKMO CNRM CERFACSRMS
MPI
SCNR
Error=
3.82( 4.56)
1987
1988
1989
1990
1991
1992
1993
1994
Year
1995
F. Vitart (ECMWF)
GMES - ENSEMBLES 2008
1996
1997
1998
1999
2000
2001
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
Year
Page 7
Seasonal crop forecast using an
ensemble crop yield model
25
63 ensemble members
20
Frequency
Model average
713 kg ha-1
Observed
775 kg ha-1
15
10
5
0
200 300 400 500 600 700 800 900 1000 1100 1200
-1
Yield (kg ha )
Multi-model ensemble for predicting seasonal groundnut yield in Gujarat, India, 1998, from Challinor et al. (2005).
GMES - ENSEMBLES 2008
Page 8
ENSEMBLES mitigation emissions
scenario
New emissions scenario developed
 IPCC SRES A1B baseline, stabilise
towards 450ppmv CO2eq
 provides information towards EU goal
of limiting warming to less than 2°C
above pre-industrial levels
 Uses proposed IPCC “AR5” design
 Earth system models will be driven by GHG concentrations, rather than
emissions. Carbon fluxes give implied emissions
 Will inform details of AR5 design and how to scientifically
exploit the runs
GMES - ENSEMBLES 2008
Page 9
T2m response E1 scenario
GMES - ENSEMBLES 2008
Taken from E. Roeckner, RT2A presentation
ENSEMBLES GA 2008, Santander, Spain
Page 10
Projected temperature extremes
Observations, 1864-2003
GMES - ENSEMBLES 2008
Page 11
Precipitation changes to 2090-2099
GMES - ENSEMBLES 2008
Page 12
Change in mean precipitation (%)
ECHAM
HIRHAM
July August September
GMES - ENSEMBLES 2008
Page 13
Changes in precipitation frequency (%)
Mean
GMES - ENSEMBLES 2008
> 99% percentile
July-August-September
Page 14
Projected change in summer-average
precipitation over Europe
- an ensemble of model simulations.
-15
0
15
% change in
2050s
GMES - ENSEMBLES 2008
Page 15
Response surface modelling of projected
risk of climate change impact
GMES - ENSEMBLES 2008
Page 16
Likelihood of low water levels in Lake
Mälaren, Sweden (perturbed physics exp.)
Use of a
response surface
Approach – regional
scale
GMES - ENSEMBLES 2008
Page 17
Risk probability of low crop productivity - Durum
wheat, perturbed physics experiment
Delta Risk
< -20 %
*
-10 %
0%
10 %
20 %
30 %
40 %
50 %
> 60 %
* Delta Risk was calculated as differences between the percentage of yields that
do not exceed yield threshold (20 percentile) in present and A1b scenarios
GMES - ENSEMBLES 2008
Page 18
Projections of bark beetle infestation in
northern European forests
GMES - ENSEMBLES 2008
Page 19
GMES - ENSEMBLES 2008
Page 20