European regional climate change and the PRUDENCE project

1
European regional climate change and the PRUDENCE
project
Ole Bøssing Christensen DMI
2
IPCC AR4 CH. 11 structure Jens H. Christensen
(CLA)
3
Chapter 11 structure Regional Climate
ProjectionsLength60 printed pages including all
refs and figures, excl. FAQCLAs Christensen and
HewitsonLAs Busuioc, Chen, Gao, Held, Jones,
Kwon, Laprise, Magana, Mearns, Menendez,
Räisänen, Rinke, Kumar, Sarr, Whetton
Executive summary (1-2 pages) 11.1 Introduction
(3 pages) 11.1.1 The importance of regional
projections 11.1.2 Summary of the TAR 11.1.3
Developments since the TAR 11.2 Assessment of
Methods 11.2.1 Generating regional information (5
pages) 11.2.1.1 AOGCM results 11.2.1.2 High
resolution AGCMs 11.2.1.3 Nested RCMs 11.2.1.4
Statistical downscaling 11.2.1.5 Pattern scaling
of climate model simulations 11.2.1.6 Other
methods 11.2.1.7 Inter-comparison of
methods 11.2.2 Quantifying uncertainties (3-4
pages) 11.2.2.1 Sources of regional
uncertainty 11.2.2.2 Methodological
developments 11.3 Regional Projections (30
pages) Details on following slides 11.4
Conclusions and discussion (1 page)
4
Chapter 11 structure Regional Climate
Projections
11.3 Regional Projections (30 pages) 11.3.1
Introduction to regions and relationship to WGII
regions (1 page) (Any sub-regions listed below
may be further sub-divided if authors feel this
is appropriate) (Length nominally 3-4 pages
each) 11.3.2 Africa Sahelian Africa Horn of
Africa / Arabian peninsula Equatorial
Africa Southern Africa 11.3.3 Mediterranean and
Europe Mediterranean Central and northern
Europe 11.3.4 Asia Central Asia South Asia East
Asia South east Asia / Maritime continent 11.3.5
North America North America 11.3.6 Latin
America Central America / Caribbean Northern
South America Southern South America 11.3.7
Australia and New Zealand Australia/New
Zealand 11.3.8 Polar Arctic Antarctic 11.3.9
Small Islands
5
Chapter 11 structure Regional Climate
Projections
BOX 11.1 Summary of AOGCM regional projections
(2 pages) Consistent method across regions, to
include uncertainty Probabilistic statements
based on AOGCMs, in coordination with Ch 10 BOX
11.2 Common aspects of small scale climate
change High altitude (1 page) BOX 11.3 Common
aspects of small scale climate change Coastal
(1 page) Table 11.1 Extremes (1 page) Summary
table in collaboration with Ch 3,4,5,9,10
WGII FAQ
Proposed FAQ a) Does this report say anything
about what will happen in my back yard? b) Will
the weather become more extreme? c) How can I use
regional information that is uncertain? (Why are
regional projections useful?) d) Whats
downscaling? e) Whats wrong with extending
recent regional trends?
6
IPCC WG1 schedule
7
PRUDENCE participants

• 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 Recherche sur
  lEnvironnement et le Développement, SMASH,
  Paris, FRA

8
PRUDENCE objectives

• A series of high resolution climate change
  scenarios for 2071-2100 for Europe
• Characterize level of confidence and variability
  related to model formulations and climate
  natural/internal variability
• Assess the uncertainty in European regional
  scenarios resulting from model formulation
• Quantitatively assess the risks arising from
  changes in regional climate over Europe, and
  estimate changes in extremes like heat waves,
  flooding and wind storms, by providing a robust
  estimation of the likelihood and magnitude of the
  changes
• Demonstrate the value of the wide-ranging
  scenarios by applying them to 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

9
A modelling system for detailed regional
scenarios the PRUDENCE method
Coupled GCM (300km atmosphere)
SST/sea-ice change from coupled GCM
Observed SST/sea-ice
150km global atmospheric GCM
12-50km RCM for relevant region
10
Quasi-ensemble probabilities
11
Precipitation change sources of uncertainty
C. Frei, ETH
95-confidence internal variability
12
Precipitation change sources of uncertainty
OBS Slightly different values since the changes
in precipitation have been scaled to a 3 K change
of the global mean temperature
13
Probabilistic precipitation change
14
Variability sources in sub-areas
1 British Isles
M. Déqué, Météo-France
2 Iberian peninsula
3 France
4 Central Europe
5 Scandinavia
6 Alps
7 Mediterranean
8 Eastern Europe
15
Temperature change sources of uncertainty
DJF
Depends on driving model
Also on RCM and scenario
16
Precipitation change sources of uncertainty
DJF
Driving GCM and RCM
RCM quite important
17
Baltic water balance
18
7 RCMs 50 km - 2 RCMs 25 km A2
19
9 RCMs (2 GCMs) 50 km - 2 RCMs 25 km A2
20
9 RCMs (2 GCMs) 50 km - 2 RCMs 25 km A2
- 3 RCMs 50 km B2
21
PRUDENCE work on extremes

• Better understanding of how European weather and
  climate extremes are likely to change
• Heat waves
• Precipitation heavy and low
• Wind storms and storm surges

22
Precipitation extremes
23
Changes in HIRHAM 5-year return levels
5-day Winter precipitation
Summer 1-day precipitation
HIRHAM
? Increases over Europe except for decreases in
south in summer
24
Sensitivity due to GCM and RCM resolution
ECHAM
HC 50km
HC 25km
25
JAS precipitation mm/day Resolution 50km
26
JAS precipitation mm/day Resolution 25km
27
JAS precipitation mm/day Resolution 12km
28
Wind extremes
29
change in 90th percentile of 10-metre wind speed
RCAO
? Increased wind speed intensity in core of
Europe north of Alps
30
A2 changes in max winter surge heights
HIRHAM
Changes (meters) in max surge heights from
HadAM3H / HIRHAM.
? Largest change of 0.3 metres on coasts near
German bight
31
Conclusions

• Warming of near-surface temperatures
• DJF west/east gradient with strongest warming in
  the east
• JJA north/south gradient with strongest warming
  in the south
• Changes in precipitation
• DJF mainly due to driving GCM but also due to
  RCM
• JJA dryer conditions in all but northern Europe
• Large ensemble of simulations allows for the
  generation of probabilistic regional climate
  scenarios
• Uncertainty of temperature changes
• DJF mainly due to driving GCM
• JJA also due to RCM and scenario
• Uncertainty of changes in precipitation
• DJF mainly due to driving GCM but also due to
  RCM
• JJA to a large extent due to RCM

32
Conclusions

• Significant changes of the discharge into the
  Baltic
• Increased magnitude due to enhanced winter
  precipitation
• Earlier peak due to earlier snow melt

33
Conclusions

• Heat waves increased frequency, intensity, and
  duration of summer heat waves
• Increase in interannual variability of
  temperature Summer 2003 could become more likely
• Heavy precipitation general increase except
  over S. Europe in summer. Central Europe will
  have less rainy days, but probably larger
  intensities
• Wind storms increased intensity and frequency
  of high wind speed events in winter
• Storm surges increase in maximum storm surge
  level of up to 0.3 metres especially near the
  German Bight.

34
(No Transcript)
35
Sensitivity of response to RCM
Winter precipitation
Summer precipitation

• More consistent 10 increase in winter
• Strong sensitivity to RCM formulation in summer
  HC models have so much drying that intensity
  increase is overshadowed

36
Mean fields
37
Precipitation change - DJF
38
Precipitation change - JJA
39
Near-surface temperature change - DJF
40
Near-surface temperature change - JJA
41
7 RCMs 50 km A2
42
Methodologies
43
Temperature extremes
44
Number of days/yr gt 30C
1961-90 HIRHAM CTL
2071-2100 HIRHAM A2
? Northward shift of heat waves e.g. Paris 9
days/yr ? 50 days/yr
45
Percentage change in 99th percentile
Fraction of change left after correcting for the
change in the location (median)
HIRHAM
After correcting for the change in the location
and scale (median and IQR)
? Changes in variance have a big impact on high
extremes
46
Schär et al. (2003)
47
Schär et al. (2003)
48
Uncertainty due to emission scenario
Winter 5-day maxima increase related but smaller
increases in means similar differences between
RCMs, GCMs, scenarios, natural variability. Summer
1-day maxima increase related decrease in
means largest differences due to RCM. Smaller
changes for B2 than for A2.
49
Change in mean sea-level pressure
CHRM
? More cyclonic low pressure conditions in winter
50
Schär et al. (2003)
51
Relative changes in heat wave indices
Duration
Intensity
Frequency
Number
HIRHAM
? Increased frequency, intensity and duration of
heat waves
52

• Regional Template
• Extremes, thresholds, land cover change, air
  quality (as there will be global projections in
  Ch7), and other effects to be folded into the
  discussion (coordinate with Ch 8,10 WGII)
• Coordination with WGII time frames
  (2020,2050,2080) where possible
• Key regional processes (current climate, with
  schematic)
• Regional simulation skill (Ch 8 coordination)
• Projections (begin with synthesis of PCMDI
  results regional map, followed by downscaled
  information where available)
• Cross-reference projections to context of Ch
  3,4,5,6 WGII Ch 1.
• Uncertainties (consistent terminology)