Instrumental observations and reconstructions of global and hemispheric temperature evolution reveal a pronounced warming during the past approximately 150 years. One expression of this warming is the observed increase in the occurrence of heatwaves. Conceptually this increase is understood as a shift of the statistical distribution towards warmer temperatures, while changes in the width of the distribution are often considered small. Here we show that this framework fails to explain the record-breaking central European summer temperatures in 2003, although it is consistent with observations from previous years. We find that an event like that of summer 2003 is statistically extremely unlikely, even when the observed warming is taken into account. We propose that a regime with an increased variability of temperatures (in addition to increases in mean temperature) may be able to account for summer 2003. To test this proposal, we simulate possible future European climate with a regional climate model in a scenario with increased atmospheric greenhouse-gas concentrations, and find that temperature variability increases by up to 100%, with maximum changes in central and eastern Europe.

Multimodel combination is a pragmatic approach to estimating model uncertainties and to making climate projections more reliable. The simplest way of constructing a multimodel is to give one vote to each model (''equal weighting''), while more sophisticated approaches suggest applying model weights according to some measure of performance (''optimum weighting''). In this study, a simple conceptual model of climate change projections is introduced and applied to discuss the effects of model weighting in more generic terms. The results confirm that equally weighted multimodels on average outperform the single models, and that projection errors can in principle be further reduced by optimum weighting. However, this not only requires accurate knowledge of the single model skill, but the relative contributions of the joint model error and unpredictable noise also need to be known to avoid biased weights. If weights are applied that do not appropriately represent the true underlying uncertainties, weighted multimodels perform on average worse than equally weighted ones, which is a scenario that is not unlikely, given that at present there is no consensus on how skill-based weights can be obtained. Particularly when internal variability is large, more information may be lost by inappropriate weighting than could potentially be gained by optimum weighting. These results indicate that for many applications equal weighting may be the safer and more transparent way to combine models. However, also within the presented framework eliminating models from an ensemble can be justified if they are known to lack key mechanisms that are indispensable for meaningful climate projections.

ABSTRACT:The success of multi-model ensemble combination has been demonstrated in many studies. Given that a multi-model contains information from all participating models, including the less skilful ones, the question remains as to why, and under what conditions, a multi-model can outperform the best participating single model. It is the aim of this paper to resolve this apparent paradox.The study is based on a synthetic forecast generator, allowing the generation of perfectly-calibrated single-model ensembles of any size and skill. Additionally, the degree of ensemble under-dispersion (or overconfidence) can be prescribed. Multi-model ensembles are then constructed from both weighted and unweighted averages of these single-model ensembles.Applying this toy model, we carry out systematic model-combination experiments. We evaluate how multi-model performance depends on the skill and overconfidence of the participating single models. It turns out that multi-model ensembles can indeed locally outperform a 'best-model' approach, but only if the single-model ensembles are overconfident. The reason is that multi-model combination reduces overconfidence, i.e. ensemble spread is widened while average ensemble-mean error is reduced. This implies a net gain in prediction skill, because probabilistic skill scores penalize overconfidence. Under these conditions, even the addition of an objectively-poor model can improve multi-model skill. It seems that simple ensemble inflation methods cannot yield the same skill improvement.Using seasonal near-surface temperature forecasts from the DEMETER dataset, we show that the conclusions drawn from the toy-model experiments hold equally in a real multi-model ensemble prediction system.

SUMMARYDynamical aspects of the life cycle of the winter storm 'Lothar' (24-26 December 1999) are investigated with the aid of the European Centre for Medium-Range Weather Forecasts analysis data and mesoscale model simulations. Neither of these datasets capture the full amplitude of the observed extreme pressure fall and surface wind speeds, but they do help identify a range of key dynamical and physical features that characterize the development of this unusual event. The analysis and interpretation is primarily based upon the evolution of the lower-and upper-level potential vorticity (PV) eld complemented by three-dimensional trajectory calculations.'Lothar' originated in the western Atlantic and travelled as a shallow low-level cyclone of moderate intensity towards Europe. This translation took place below and slightly to the south of a very intense upper-level jet and was accompanied by continuous and intense condensational heating that sustained a pronounced positive low-level PV anomaly (not unlike the concept of a 'diabatic Rossby wave'). No signi cant PV anomalies were evident at the tropopause level during this early phase of the life cycle. The surface cyclone intensi ed rapidly when the shallow cyclone approached the jet-stream axis. The circulation induced by the diabatically produced low-tropospheric PV anomaly on steeply sloping isentropic surfaces that transect the intense upper-level jet contributed signi cantly to the rapid formation of a narrow and deep tropopause fold. This stratospheric PV anomaly virtually merged with the diabatically produced ephemeral PV feature to form a vertically aligned tower of positive PV at the time of maximum storm intensity. A sensitivity study with a dry adiabatic hindcast simulation shows no PV-tower con guration (and only a very weak surface development) and con rms the primary importance of the cloud diabatic heating for the tropopause fold formation and the rapid 'bottom-up' intensi cation of 'Lothar'.A comparison of the anomalously high Atlantic sea surface temperatures in December 1999 with the watervapour source regions for the latent-heat release that accompanied the rapid intensi cation phase of 'Lothar' shows a close relationship. This is of importance when discussing the possible implications of climate variability and change on the development of North Atlantic winter storms.

Updated European averaged autumn and winter surface air temperature (SAT) timeseries indicate that the autumn 2006 and winter 2007 were extremely likely (>95%) the warmest for more than 500 years. In both seasons, the European SAT anomaly is widespread with anomalies up to three standard deviations from normal. The anomalous warmth is associated with strong anticyclonic conditions and warm air advection from south west. Phenological impacts related to this warmth included some plant species having a partial second flowering or extended flowering till the beginning of winter. Species that typically flower in early spring were found to have a distinct earlier flowering after winter 2007.

The Brier skill score (BSS) and the ranked probability skill score (RPSS) are widely used measures to describe the quality of categorical probabilistic forecasts. They quantify the extent to which a forecast strategy improves predictions with respect to a (usually climatological) reference forecast. The BSS can thereby be regarded as the special case of an RPSS with two forecast categories. From the work of Müller et al., it is known that the RPSS is negatively biased for ensemble prediction systems with small ensemble sizes, and that a debiased version, the RPSSD, can be obtained quasi empirically by random resampling from the reference forecast. In this paper, an analytical formula is derived to directly calculate the RPSS bias correction for any ensemble size and combination of probability categories, thus allowing an easy implementation of the RPSSD. The correction term itself is identified as the “intrinsic unreliability” of the ensemble prediction system. The performance of this new formulation of the RPSSD is illustrated in two examples. First, it is applied to a synthetic random white noise climate, and then, using the ECMWF Seasonal Forecast System 2, to seasonal predictions of near-surface temperature in several regions of different predictability. In both examples, the skill score is independent of ensemble size while the associated confidence thresholds decrease as the number of ensemble members and forecast/observation pairs increase.

Consideration of the time scales and vertical extent of stratosphere-troposphere exchange events reveals new insight into this phenomenon and its impact on surface ozone. However, STE is still poorly understood and inadequately quantified, due to the involvement of physical and dynamical processes on local to global scales (Holton et al. 1995), and conceptual problems.On a long-term and global scale, and in the zonally averaged sense (Brewer 1949), there is slow ascent from the troposphere to the stratosphere in the Tropics (Plumb 1996;Mote et al. 1996), quasi-isentropic transport to the extratropics in the stratosphere (Waugh 1996), and downward flow from the stratosphere to the troposphere in middle and higher latitudes (Fig. 1). This circulation is related to the dissi-

The influence of uncertainties in gridded observational reference data on regional climate model (RCM) evaluation is quantified on a pan‐European scale. Three different reference data sets are considered: the coarse‐resolved E‐OBS data set, a compilation of regional high‐resolution gridded products (HR) and the European‐scale MESAN reanalysis. Five high‐resolution ERA‐Interim‐driven RCM experiments of the EURO‐CORDEX initiative are evaluated against each of these references over eight European sub‐regions and considering a range of performance metrics for mean daily temperature and daily precipitation. The spatial scale of the evaluation is 0.22°, that is, the grid spacing of the coarsest data set in the exercise (E‐OBS). While the three reference grids agree on the overall mean climatology, differences can be pronounced over individual regions. These differences partly translate into RCM evaluation uncertainty. For most cases observational uncertainty is smaller than RCM uncertainty. Nevertheless, for individual sub‐regions and performance metrics observational uncertainty can dominate. This is especially true for precipitation and for metrics targeting the wet‐day frequency, the pattern correlation and the distributional similarity. In some cases the spatially averaged mean bias can also be considerably affected. An illustrative ranking exercise highlights the overall effect of observational uncertainty on RCM ranking. Over individual sub‐domains, the choice of a specific reference can modify RCM ranks by up to four levels (out of five RCMs). For most cases, however, RCM ranks are stable irrespective of the reference. These results provide a twofold picture: model uncertainty dominates for most regions and for most performance metrics considered, and observational uncertainty plays a minor role. For individual cases, however, observational uncertainty can be pronounced and needs to be definitely taken into account. Results can, to some extent, also depend on the treatment of precipitation undercatch in the observational reference.