Transcript The PRUDENCE project

COP10, 13 December 2004
The PRUDENCE
project
Jens Hesselbjerg Christensen
PRUDENCE coordinator
[email protected]
http://prudence.dmi.dk
The PRUDENCE Consortium
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
A.
B.
C.
D.
E.
F.
Danish Meteorological Institute, Copenhagen, DK
CINECA, Bologna, IT
Météo-France/CNRM, Toulouse, FRA
Deutsches Zentrum für Luft- und Raumfahrt e.V., Weßling, GER
Hadley Centre for Climate Prediction and Research, Met Office, Bracknell, UK
Climate Research ETH (Eidsgenössische Technische Hochschule), Zürich, CH
GKSS Research Center (Institute for Coastal Research), Geesthacht, GER
Max Planck Institut für Meteorologie, Hamburg, GER
Swedish Meteorological and Hydrological Institute, Rossby Centre, Norrköping, SWE
Universidad Complutense, Madrid, SP
Universidad Politecnica, Madrid, SP
International Centre for Theoretical Physics, Trieste, IT
Danish Institute of Agricultural Sciences, Foulum, DK
Risø National Laboratory, System Analysis Dept., DK
University of Fribourg, CH
Finnish Environmental Institute, Helsinki, FIN
University of Reading, UK
University of Lund, SWE
Centre International de Reserche sur l’Environment et Developpement,SMASH, Paris, FRA
Climate Research Unit, University of East Anglia, UK
Finnish Meteorological Institute, Associated to FEI (No. 16), FIN
Norwegian Meteorological Institute, Blindern, NO
Royal Dutch Meteorological Institute, De Bilt, NL
UQAM, Montreal, CAN
CSIRO, Victoria, AUS
Czech Republic, Israel, Greece, California, USA………………..
Munich-Re, Electricité de France, Elforsk, Hamburg Institute of International Economics,
Uni-Münster, DG-Research, STARDEX,
MICE
Overview this presentation
• The scientific objectives
• Aims and products
• Results
PRUDENCE objectives
• provide a series of high resolution (spatially
and in time) climate change scenarios for
2071-2100 for Europe;
• assess the uncertainty in European regional
climate scenarios resulting from model
formulation;
• in practical terms characterise the level of
confidence in these scenarios and the
variability in them related to model
formulations and climate natural/internal
variability;
PRUDENCE objectives
• quantitatively assess the risks rising
from changes in regional weather and
climate over all of Europe, and estimate
future changes in extreme events such
as flooding and wind storms, by
providing a robust estimation of the
likelihood and magnitude of the
changes;
PRUDENCE objectives
• demonstrate the value of the wide-ranging
climate change scenarios by applying them
to climate impacts models focusing on
effects on adaptation and mitigation
strategies;
• assess socio-economic and policy related
decisions for which such improved scenarios
could be beneficial;
• disseminate the results of PRUDENCE widely
and provide a project summary aimed at
policy makers and non-technical interested
parties
Regional information
 Regional
aspects
of
coupled
oceanatmosphere general circulation models
Temperature change relative to global mean
A2 & B2
(Giorgi et al. GRL, 2001)
Regional information
 Regional aspects of coupled ocean-atmosphere
general circulation models
 Time slice experiments utilizing high- and
variable- resolution atmospheric GCM’s
Regional information
 Regional aspects of coupled ocean-atmosphere
general circulation models
 Time slice experiments utilizing high- and
variable- resolution atmospheric GCM’s
 Limited area (regional climate) models RCM’s
 Statistical down scaling (see STARDEX)
A road to impact scenarios
the Delta Change approach
GCM
RCM
today
today
An interface
Impact
model
scenario
global
Impact
scenarios
scenario
scale
local
UNCERTAINTIES IN CLIMATE CHANGE
PROJECTIONS
• Uncertainty due to observational limitations
– use multiple means of validation
• Uncertainty in future emissions
– use a range of SRES emissions scenarios
• Natural variability
– use a number of different initial conditions (ensembles)
• Uncertainty in the response of the climate system
– use a range of climate modelling systems
– AND/OR assess confidence in climate projections (better models)
• Need for a large-scale coordinated effort
CO2 EMISSIONS PROFILES
under IPCC SRES scenarios
Source:
IPCC
GLOBAL TEMPERATURE RISE
due to four SRES emissions scenarios
Source: Hadley Centre
AGCM
exp forcing
HadCM3
SRES A2
HadCM3
SRES B2
HadAM3H
ARPEGE
ECHAM5
CCM3
3 ensemble members
150 km
BDY 1
1 member
150 km
BDY 2
2 mem
high res.
1 member
T106
2 members
T80 100km
BDY4
1 mem
high res.
ECHAM4/OPYC3
SRES A2
ARPEGE/OPA
SRES B2
RCM
50km
Input
BDY 1
BDY 1
BDY 1
ini
cond.
BDY 2
BDY 3
BDY 4
RCM
20 km
Input
BDY 1
P5
1 member
T106 T42
BDY 3
1 mem high res.
Meteo
P9
P1
P10
P6
ETH
P2/P12
IPCC
P8
MPI
P7GKSS
1 mem
1 mem
1 mem
1 mem
1 mem
1 mem
1 mem
1 mem
1 mem
Had Rossby DMI Es
3 mem
1 mem
3 mem
1 mem
1 mem
P5
P9
P1
P6
P8
1 mem
1 mem
1 mem
1 mem
1 mem
12 km
Input
Impact model
N.E. storm surge
GCM
C.E. river catchment
N.E. drainage area
S.E. agriculture
probed
all
All (2 for adaptive
responses)
one from each using
BDY1
well probed
all
all (only 2 for adaptive
responses)
N.E. agriculture
all (only 1 for
adaptive responses)
all
all
all (only 2-3 for
adaptive responses)
well probed
all
ecosystems
simple models and
indices
Mediterranean
agriculture and
hydrology
one mem
RCM 50 km
RCM 20 km
all members
all
all
Only one from
contractor/ partner
10 –UCMfor impact and
adaptive responses
all (only 2 for
adaptive responses)
all
One parent GCM and associated RCMs at 50km and 20km.
Comparisons can be made between (a) different forcings (SRES A2
and B2) (b) different ensemble members (c) different scales.
Attention will focus on the range of scenarios.
Relating to observed trends
• Flooding
• Heat wave
Recent European Summer Climate
Trends and Extremes
• Summer precipitation over much of Europe and the
Mediterranean Basin has shown a decreasing trend in
recent decades
• The intensity of summer precipitation events has shown
predominant increases throughout Europe
• The western European summer drought of 2003 is
considered one of the severest on record.
– Heat related casualties in France, Italy, the Netherlands, Portugal,
the United Kingdom, and Spain reached nearly 20,000.
– Many countries are experiencing their worst harvest since World
War II.
• In contrast, during 2002, many European countries
experienced one of their wettest summers on record.
– Weather systems brought widespread heavy rainfall to central
Europe, causing severe flooding along all the major rivers.
– The Elbe River reached its highest level in over 500 years of
record
What can PRUDENCE
say?
Change in JAS mean precip (2071-2100 minus 1961-1990)
Christensen&Christensen (2003;2004)
Sensitivity due to GCM and RCM resolution
ECHAM
Christensen & Christensen, Nature (2003)
Hadley 50km
Hadley 25km
Changes in heavy and mean precipitation
(1961-90 =>2071-2100)
Schär et al. (2004)
Schär et al. (2004)
Changes in Summer
500 hPa Geopotential Heights
NCEP Reanalysis
B2 Scenario
(1976-2000) minus (1951-1975) (2071-2100) minus (1961-1990)
Pal et al. (2004)
( meters)
( meters)
Change in Summer Precipitation
CRU Observations
(1976-2000) minus (1951-1975)
B2 Scenario
(2071-2100) minus (1961-1990)
(% change)
(% change)
Changes in Summer Extremes:
B2 Scenario
Max Dry Spell Length
Max 5-Day Precipitation
(2071-2100) minus (1961-1990) (2071-2100) minus (1961-1990)
( Days)
(% change)
Precipitation Distribution
REF
Drier
Summers
ref
B2
More
Droughts
B2
More
Floods
B2
ref
Conclusions I
• In both the A2 and B2 scenarios we find summer
warming and drying over most of the European
region.
• Maximum dry spell length (drought), maximum
precipitation intensity (flood) and interannual
variability increase in summer throughout most
of Europe
• Shift and change in shape of the precipitation
distribution
• The results from the climate change simulations
are consistent with trends of summer climate
observed over Europe in recent decades
Impacts
• Hydrology
Prudence
basins
Baltic
Basin
7 RCMs … A2
same GCM
boundary
7 RCMs ~50km … A2
2 RCMs ~25km … A2
same GCM
boundary
9 RCMs ~50km … A2
2 RCMs ~25km … A2
2 GCMs
9 RCMs ~50km … A2
2 RCMs ~25km … A2
3 RCMs ~50km … B2
2 GCMs
7 RCMs ~50km … A2
1 GCM ~150km … A2
same GCM
boundary
Conclusions 2
• Ensemble information from different
models provides valuable information
about the degree of uncertainty in the
impact signal
• Seasonal shift in hydrological cycle
confirmed
Impacts
• Hydrology
• Storm surges
Conclusions 3
• More intensive surge in warmer climate
– Up to 30% increase in high percentiles, no
change in mean.
– Magnitude of shift is highly dependant on
location.
• Valid for all four RCM simulations, driven
with same GCM (HC) and also similar
signal for simulation, driven with another
GCM
Impacts
• Hydrology
• Storm surges
• Simple indices
Potential shifts in
extreme climatic events
=> risks in society and ecosystems
CLIMATIC CHANGE IN:
IMPACT SECTORS
ADAPTATION OPTIONS
Maximum 1-day and 5-day
precipitation total
Water resources,
agriculture
Regulation guidelines, flood
gates, land use planning
Maximum length of dry
spells
Water resources,
agriculture
Increases in water-use
efficiency, water recycling
Total number of frost days
Ecosystems, transport,
heating, building
Preparedness for decreases
in energy consumption
Total number of days
crossing the 0ºC threshold
Wintertime road
maintenance
Timing of salting of roads
Frost-free season
Ecosystems, transport
Timing of cultivation
practices
Snow season
Recreation, tourism
Artificial snow in ski centres
Maximum ice cover over the
Baltic Sea
Wintertime shipping
Timing and efficiency of icebreaking
Changes in frost days and min temperature
1961-90 =>2071-2100
Impacts
•
•
•
•
Hydrology
Storm surges
Simple indices
Agriculture
Thermal suitability for grain maize (baseline + 2080’s)
Suitable area
Observed baseline
1961-1990 (CRU)
green – baseline suitability
red – suitability extension for all RCMs
blue – RCM uncertainty in extension
9 RCMs
A2, HadAM3H-driven
3 RCMs
B2, HadAM3H-driven
Winter wheat yield in 2080’s (example from 1 RCM)
Modelled 2080’s
Difference to CRU baseline
Nitrate leaching from wheat in 2080’s (example from 1 RCM)
Modelled 2080’s
Difference to CRU baseline
Conclusions 4
• General productivity increases for agricultural crops in
Northern Europe and decreases in Southern Europe
has low uncertainty, although the option to cultivate
crops during the winter in some Mediterranean
countries needs more consideration
• Impacts of nitrate leaching (and possibily other
environmental effects of agriculture) may have a
completely different spatial structure than the yield
effect
Impacts
•
•
•
•
•
Hydrology
Storm surges
Simple indices
Agriculture
Society
Methodology
• To compare the future climate of Paris and the present
climate of a gridpoint x’, we define 3 distances:
– dT measures the mean absolute distance between the 12
monthly mean temperatures
– dAP measures the relative distance between the annual
mean precipitations (to account for the total water
availability)
– dMP measures the mean relative distance between the 12
monthly mean precipitations (to account for the
precipitation seasonal cycle)
• 10 European cities: Athens, Barcelona, Berlin, Geneva,
London, Madrid, Marseille, Paris, Roma and Stockholm.
A global shift southward
Results based on ARPEGE
Impacts
•
•
•
•
•
•
Hydrology
Storm surges
Simple indices
Agriculture
Society
Surprises?
Dec 2001
Sept 2002
Thank you all
Oct 2003
Sept 2004
Utilisation of PRUDENCE data for
regional analysis
Ekstrøm et al. (2004)
Assessing uncertainty of regional
changes
• Construct a probability distribution function
(PDF) of climate change
• Combine PDF from
– global annual mean temperature increase
– change in regional temperature/precipitation
– per degree of global temperature increase (Jones, 2000)
• (Uniform distributions from within a range)
• Normal distribution* of PDF for the scaling
variables, log normal for global increase
• Full range of uncertainty
*(estimated from ANalysis Of VAriance (ANOVA) )
(2071-2100) wrt. (1961-1990)
Temperature
Precipitation
Ekström et al. (in press)