COP 10 13 December 2004 The PRUDENCE project

COP 10, 13 December 2004 The PRUDENCE project Jens Hesselbjerg Christensen PRUDENCE coordinator jhc@dmi. dk 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 A 2 & B 2 (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 An interface today 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

CO 2 EMISSIONS PROFILES under IPCC SRES scenarios Source: IPCC

GLOBAL TEMPERATURE RISE due to four SRES emissions scenarios Source: Hadley Centre

100 km T 42 Meteo Had Rossby DMI Es 12 km ETH IPCC MPI GKSS

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 50 km Hadley 25 km

Changes in heavy and mean precipitation (1961 -90 =>2071 -2100)

Schär et al. (2004)

Changes in Summer 500 h. Pa Geopotential Heights NCEP Reanalysis B 2 Scenario (1976 -2000) minus (1951 -1975) (2071 -2100) minus (1961 -1990) Pal et al. (2004) ( meters)

Change in Summer Precipitation CRU Observations (1976 -2000) minus (1951 -1975) B 2 Scenario (2071 -2100) minus (1961 -1990) (% change)

Changes in Summer Extremes: B 2 Scenario Max Dry Spell Length Max 5 -Day Precipitation (2071 -2100) minus (1961 -1990) ( Days) (% change)

Precipitation Distribution REF Drier Summers ref B 2 More Droughts B 2 More Floods B 2 ref

Conclusions I • In both the A 2 and B 2 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 … A 2 same GCM boundary

7 RCMs ~50 km … A 2 2 RCMs ~25 km … A 2 same GCM boundary

9 RCMs ~50 km … A 2 2 RCMs ~25 km … A 2 2 GCMs

9 RCMs ~50 km … A 2 2 RCMs ~25 km … A 2 3 RCMs ~50 km … B 2 2 GCMs

7 RCMs ~50 km … A 2 1 GCM ~150 km … A 2 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 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 A 2, Had. AM 3 H-driven 3 RCMs B 2, Had. AM 3 H-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: – d. T measures the mean absolute distance between the 12 monthly mean temperatures – d. AP measures the relative distance between the annual mean precipitations (to account for the total water availability) – d. MP 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)