Réf. Frei & al. 2006 - A

Référence bibliographique complète

FREI, C., SCHOLL, R., FUKUTOME, S., SCHMIDLI, J., VIDALE, P.L. 2006. Future change of precipitation extremes in Europe : intercomparison of scenarios from regional climate models. Journal of Geophysical Research Atmosphere, 111, D06105, doi :10.1029/2005JD005965. [Etude en ligne]

Abstract: An analysis of the climate of precipitation extremes as simulated by six European regional climate models (RCMs) is undertaken in order to describe/quantify future changes and to examine/interpret differences between models. Each model has adopted boundary conditions from the same ensemble of global climate model integrations for present (1961–1990) and future (2071–2100) climate under the Intergovernmental Panel on Climate Change A2 emission scenario. The main diagnostics are multiyear return values of daily precipitation totals estimated from extreme value analysis. An evaluation of the RCMs against observations in the Alpine region shows that model biases for extremes are comparable to or even smaller than those for wet day intensity and mean precipitation. In winter, precipitation extremes tend to increase north of about 45°N, while there is an insignificant change or a decrease to the south. In northern Europe the 20-year return value of future climate corresponds to the 40- to 100-year return value of present climate. There is a good agreement between the RCMs, and the simulated change is similar to a scaling of present-day extremes by the change in average events. In contrast, there are large model differences in summer when RCM formulation contributes significantly to scenario uncertainty. The model differences are well explained by differences in the precipitation frequency and intensity process, but in all models, extremes increase more or decrease less than would be expected from the scaling of present-day extremes. There is evidence for a component of the change that affects extremes specifically and is consistent between models despite the large variation in the total response.

Mots-clés

 

 

Organismes / Contact

• Institute for Atmospheric and Climate Science, Eidgenössische Technische Hochschule, Zurich, Switzerland
• Institute of Terrestrial Ecology, Eidgenössische Technische Hochschule, Zurich, Switzerland
• Center for Global Atmospheric Modelling, Department of Meteorology, University of Reading, Reading, UK

Federal Office of Meteorology and Climatology (MeteoSwiss), Zurich, Switzerland (christoph.frei@meteoswiss.ch)

 

(1) - Paramètre(s) atmosphérique(s) modifié(s)

(2) - Elément(s) du milieu impacté(s)

(3) - Type(s) d'aléa impacté(s)

(3) - Sous-type(s) d'aléa

 

 

 

 

 

Pays / Zone

Massif / Secteur

Site(s) d'étude

Exposition

Altitude

Période(s) d'observation

 

 

 

 

 

 

 

(1) - Modifications des paramètres atmosphériques

Reconstitutions

 

Observations

 

Modélisations

In the present study the authors undertook an intercomparison of precipitation extremes as simulated by six different European regional climate models, all with comparable model settings and driven with boundary data from the same global climate model. An evaluation of the model simulations for present climate in the region of the European Alps shows that RCMs are capable of representing mesoscale spatial patterns in precipitation extremes that are not resolved by today's GCMs. However, model biases are large in some cases, in particular in summer. Even for rare extremes (5-year return values), these biases were nevertheless similar to or smaller than those for wet day intensity or mean precipitation. The authors conclude that there is no evidence from this evaluation for model errors that are specific to precipitation extremes and that are not evident in simpler, average diagnostics. Moreover, comparison to earlier evaluations of similar RCMs, but driven by reanalysis data [Frei et al., 2003], implies that errors inherited from the driving GCM were small and did not alter the RCM specific error characteristics, which, in retrospect, attests to the quality of the GCM's present-day climate.

The simulated future change in European precipitation extremes shows a seasonally very distinct pattern: In winter, land regions north of about 45°N experience an increase in multiyear return values while the Mediterranean region experiences small changes with a general tendency toward decreases. Results are very consistent between the six RCMs, with the change in 5-year return values increasing by 0–11% in central Europe and by 10–22% in southern Scandinavia. The simulated increase of extremes is accurately explained by the increase in wet day intensity and frequency in a simple rescaling of the distribution for present climate.

The increase in wintertime precipitation extremes is a robust feature in RCM climate change experiments over Europe. Versions of the RCMs considered in this study, but driven by different GCMs, yield changes very similar to those found in this study [Durman et al., 2001; Räisänen et al., 2004]. Also, the simulated response in winter qualitatively conforms to the observed trends in heavy precipitation over Europe, which shows an increase in winter primarily north of 45°N [e.g., Klein Tank and Können, 2003; Fowler and Kilsby, 2003a, 2003b; Haylock and Goodess, 2004; Brunetti et al., 2004; Schmidli and Frei, 2005]. These similarities are worth noting, but it is premature to infer the detection of an anthropogenic influence on heavy precipitation in Europe [see also Kiktev et al., 2003, 2004; Hegerl et al., 2004].

In summer the character of change is more complex: The larger-scale pattern shows a gradient from increases in northern Scandinavia to decreases in the Mediterranean region and this is fairly similar between models, but the transition across the continent differs between models and the magnitude of the change in the 5-year return value varies considerably (−13% to +21% for central Europe and +2 to +34% for southern Scandinavia). The large model differences are well explained by differences in the change of average precipitation events as represented by wet day intensity and frequency. This suggests that it is primarily the response in the basic intensity and occurrence process of precipitation where models differ, and not the particular response in the extremes themselves. The prominent role of physical parameterizations (e.g., convection, land-surface atmosphere exchange, radiation, and clouds) may explain the large model spread in summer compared to winter where large-scale circulation exerts a stronger control [e.g., Noguer et al., 1998; Schär et al., 1999].

However, the models simulate a larger increase or smaller decrease of extremes than would have been anticipated from the simulated change in wet day intensity and frequency alone. The change in summer is therefore also governed by a factor affecting the frequency distribution more fundamentally and specifically at the tail. This factor tends to increase extreme quantiles in the simulation for future climate, independently of the sign of the total change in the quantiles. This tail-specific component is seen consistently in all RCMs and its magnitude is considerable, capable of reversing a decrease that would be expected from changes in mean conditions alone, into an increase.

The present analysis offers a more in depth statistical interpretation of results from previous studies on the future change of European summer precipitation extremes. In their RCM experiments, Christensen and Christensen [2003, 2004] and Pal et al. [2004] found increases in high precipitation quantiles in central Europe although mean precipitation was simulated to decrease. In this study, for central Europe we find an increase in five but a decrease in 2 models. In principle an increase of extremes would be possible even if the change at the tail was determined by the change in average conditions alone (see Appendix A). At sufficiently high return values an increase in wet day intensity always dominates a decrease in wet day frequency, even if the frequency decrease is larger than the intensity increase, i.e., when mean precipitation decreases. The present analysis suggests that this simple picture is not a sufficient explanation for the model results. The change in summer extremes indeed reflects a nontrivial change at the tail of the distribution. This excessive response for extremes is found consistently in all RCMs, even in those two models where extreme return values actually decrease. Also, the same behavior is noted further north in Scandinavia, where a response more similar to winter could have been expected. However, there is a large intermodel difference in the sign and magnitude of the change at a particular frequency level (return period), which is primarily related to the different responses in average conditions, and contributes to considerable uncertainty about the future change in European summer precipitation extremes.

Clearly, more research will be needed to understand the physical nature of the tail-specific response and the reasons for the different model responses in summer. Also, tests with more complex scaling models could shed more light in the statistical nature of the peculiar change in summertime precipitation pdfs. From the point of view of scenarios, the present analysis suggests that the formulation of regional models (e.g., the parameterization) contributes significantly to the uncertainty in scenarios of summer precipitation extremes. It is therefore not a waste of resources if multimodel ensemble systems, devoted to estimating scenario uncertainties, include a set of RCMs nested into the same GCM, alongside the nesting of RCMs in several different GCMs.

Hypothèses

 

 

Informations complémentaires (données utilisées, méthode, scénarios, etc.)

The purpose of this study is to compare scenarios of European precipitation extremes for the late 21st century between six different RCMs using consistent diagnostics. The idea is to isolate the contribution to scenario uncertainty, which is due to differences in the formulation of the regional models. Accordingly, all the RCM simulations being analyzed here are based on the same emission scenario (SRES A2 [Nakicenovic et al., 2000]), are nested into the same global climate model (HadAM3H [Pope et al., 2000]), and are operated at comparable grid spacing. Clearly, our analysis satisfies some obvious interest in scenarios of precipitation extremes; however, its results are also relevant for the design of multimodel ensembles, when it comes to estimating the full range of scenario uncertainty. High sensitivity of scenarios to RCM formulation may suggest the consideration of several different RCMs nested in the same GCM, whereas a low sensitivity may suggest that computational resources are used more efficiently in sampling GCM formulation, i.e., by nesting RCMs into several different GCMs.

The diagnostics of primary focus in this analysis are extremes of rainfall with return periods between 5 and 50 years. Their estimation is based on the technique of extreme value statistics [see, e.g., Coles, 2001; Katz et al., 2002] very similar to the studies mentioned above. Here this method is applied consistently to all the RCMs and results are compared quantitatively for specific regions. In addition, we also consider more direct diagnostics of average or intense events, which allows us to describe a wider range of the frequency distribution and to employ a simple scaling concept to interpret changes for rare extremes. All our analyses are carried out seasonally stratified in order to identify seasonal variations in scenarios and uncertainties [see also Wehner, 2004].

One part of this study is also devoted to an evaluation of the RCMs with respect to their representation of precipitation extremes. Unfortunately, there is currently no comprehensive high-resolution data set that would allow an evaluation for the whole European continent. In this study we consider the European Alps as a test ground. This region has at its disposal a very dense rain gauge network from which an accurate observational data set could be created that is compatible with the grid spacing of the models. Although the Alps cover only a limited part of the model's domain (typically 25 × 15 grid points), and results may not be extrapolated to other regions, this region is particularly interesting for assessing downscaling abilities because of its complex topography. Also, the evaluation in the Alps complements previous evaluation studies that have focused on more northern parts of Europe […].

The RCM integrations considered in this study were derived as part of a larger multimodel ensemble in the frame of the European project PRUDENCE [Christensen et al., 2006]. The present analysis forms part of an even broader intercomparison of downscaling methods for extremes, involving statistical and dynamical methods, in the frame of the European project STARDEX [Goodess, 2003].

 

(2) - Effets du changement climatique sur le milieu naturel

Reconstitutions

 

Observations

 

Modélisations

 

Hypothèses

 

 

Sensibilité du milieu à des paramètres climatiques

Informations complémentaires (données utilisées, méthode, scénarios, etc.)

 

 

 

(3) - Effets du changement climatique sur l'aléa

Reconstitutions

 

Observations

 

Modélisations

 

Hypothèses

 

 

Paramètre de l'aléa

Sensibilité des paramètres de l'aléa à des paramètres climatiques

Informations complémentaires (données utilisées, méthode, scénarios, etc.)

 

 

 

 

(4) - Remarques générales

 

 

(5) - Syntèses et préconisations

 

Références citées :

Brunetti, M., M. Maugeri, F. Monti, and T. Nanni (2004), Changes in daily precipitation frequency and distribution in Italy over the last 120 years, J. Geophys. Res., 109, D05102, doi:10.1029/2003JD004296.

Christensen, J. H., and O. B. Christensen (2003), Severe summertime flooding in Europe, Nature, 421, 805– 806.

Christensen, J. H., and O. B. Christensen (2004), Intensification of extreme European summer precipitation in a warmer climate, Global Planet. Change, 44, 107– 117.

Christensen, J. H., T. R. Carter, and M. Rummukainen (2006), Evaluating the performance and utility of regional climate models: The PRUDENCE project, Clim. Change, in press.

Coles, S. (2001), An Introduction to Statistical Modeling of Extreme Values, 208 pp., Springer, New York.

Durman, C. F., J. M. Gregory, D. C. Hassell, R. G. Jones, and J. M. Murphy (2001), A comparison of extreme European daily precipitation simulated by a global and a regional climate model for present and future climates, Q. J. R. Meteorol. Soc., 127, 1005–1015.

Fowler, H. J., and C. G. Kilsby (2003b), Implications of changes in seasonal and annual extreme rainfall, Geophys. Res. Lett., 30(13), 1720, doi:10.1029/2003GL017327.

Fowler, H. J., M. Ekström, C. G. Kilsby, and P. D. Jones (2005), New estimates of future changes in extreme rainfall across the UK using regional climate model integrations: 1. Assessment of control climate, J. Hydrol., 300, 212–233.

Frei, C., and C. Schär (1998), A precipitation climatology of the Alps from high-resolution rain-gauge observations, Int. J. Climatol., 18, 873– 900.

Frei, C., C. Schär, D. Lüthi, and H. C. Davies (1998), Heavy precipitation processes in a warmer climate, Geophys. Res. Lett., 25, 1431– 1434.

Frei, C., J. H. Christensen, M. Déqué, D. Jacob, R. G. Jones, and P. L. Vidale (2003), Daily precipitation statistics in regional climate models: Evaluation and intercomparison for the European Alps, J. Geophys. Res., 108(D3), 4124, doi:10.1029/2002JD002287.

Goodess, C. M. (2003), Statistical and Regional dynamical Downscaling of Extremes for European regions: STARDEX, EGGS, 6.

Haylock, M. R., and C. M. Goodess (2004), Interannual variability of extreme European winter rainfall and links with mean large-scale circulation, Int. J. Climatol., 24, 759– 776.

Hegerl, G. C., F. W. Zwiers, P. A. Stott, and V. V. Kharin (2004), Detectability of anthropogenic changes in temperature and precipitation extremes, J. Clim., 17, 3683– 3700.

Katz, R. W., M. B. Parlange, and P. Naveau (2002), Statistics of extremes in hydrology, Adv. Water Resour., 25, 1287– 1304.

Kiktev, D., D. M. H. Sexton, L. Alexander, and C.K. Folland (2003), Comparison of modeled and observed trends in indices of daily climate extremes, J. Clim., 16, 3560–3571.

Kiktev, D., D. M. H. Sexton, L. Alexander, and C. K. Folland (2004), Corrigendum: Comparison of modeled and observed trends in indices of daily climate extremes, J. Clim., 17, 2489.

Klein Tank, A. M. G., and G. P. Können (2003), Trends in indices of daily temperature and precipitation extremes in Europe, 1946– 1999, J. Clim., 16, 3665– 3680.

Nakicenovic, N., et al. (2000), Special Report on Emissions Scenarios: A Special Report of Working Group III for the Integovernmental Panel on Climate Change, 599 pp., Cambridge Univ. Press, New York.

Noguer, M., R. G. Jones, and J. M. Murphy (1998), Sources of systematic errors in the climatology of a regional climate model over Europe, Clim. Dyn., 14, 691– 712.

Pal, J. S., F. Giorgi, and X. Bi (2004), Consistency of recent European summer precipitation trends and extremes with future regional climate projections, Geophys. Res. Lett., 31, L13202, doi:10.1029/2004GL019836.

Pope, D. V, M. Gallani, R. Rowntree, and A. Stratton (2000), The impact of new physical parameterizations in the Hadley Centre climate model: HadAM3, Clim. Dyn., 16, 123– 146.

Räisänen, J., U. Hannson, A. Ullerstig, R. Do¨scher, L. P. Graham, C. Jones, H. E. M. Meier, P. Samuelsson, and U. Willén (2004), European climate in the late twenty-first century: Regional simulations with two global models and two forcing scenarios, Clim. Dyn., 22, 13– 31.

Schär, C., D. Lüthi, U. Beyerle, and E. Heise (1999), The soil-precipitation feedback: A process study with a regional climate model, J. Clim., 12, 722– 741.

Schmidli, J., and C. Frei (2005), Trends of heavy precipitation and wet and dry spells in Switzerland during the 20th century, Int. J. Climatol., 25, 753–771.

Wehner, M. F. (2004), Predicted twenty-first-century changes in seasonal extreme precipitation events in the parallel climate model, J. Clim., 17, 4281– 4290.