Réf. Beniston 2009a - A

Référence bibliographique complète

BENISTON, M., 2009: Decadal-scale changes in the tails of probability distribution functions of climate variables in Switzerland . International Journal of Climatology, 29, 1362-1368. Réf. Beniston 2009a [Etude en ligne]

Abstract: An analysis of several Swiss climatological sites reveals that a substantial change in the behaviour of pressure, minimum and maximum temperature extremes has occurred in the past two decades. Extreme cold tails defined by the 10% quantiles of temperature drop by a factor of 2 or 3, while the upper tails (beyond the 90% quantile) exhibit a four- or five-fold increase in all seasons. Pressure shows contrasting behaviour, with increases in wintertime highs and summertime lows, while precipitation shows little change. On the basis of the observed datasets, temperature biases related to extremes of pressure or precipitation have been computed, as well as for joint combinations of precipitation and pressure extremes. The most dominant bias is associated with periods without rainfall, during which temperatures are at least 1 °C warmer than otherwise. Changes in the behaviour of joint combinations of extreme pressure and precipitation regimes also have a discernible influence on temperatures.


Climatic change – Extremes - Probability density functions


Organismes / Contact

• Chair for Climate Research, University of Geneva, Geneva, Switzerland (Martin.Beniston@unige.ch)

This work was conducted in part in the context of the Swiss NCCR-Climate networked project and the EU/FP6 ENSEMBLES Project.


(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



Période(s) d'observation








(1) - Modifications des paramètres atmosphériques




The analysis has shown that in the last 20 years, there have been significant shifts in the behaviour of temperature and pressure extremes, as recorded in the tails of the distributions, while precipitation shows little change. Extremes of pressure and precipitation, or a combination of modes, are associated with strong biases in temperatures and their long-term trends despite the fact that these extreme conditions account for only a small fraction of each season. Since the 1980s, the strong shifts in pressure extremes, as well as the low precipitation/high pressure (LH) mode have clearly marked the response of minimum and maximum temperatures at all the low and high-elevation sites investigated.

The objective of this paper has been to report on the observed changes in quantiles of temperature, precipitation and pressure, and not to explore the physical mechanisms that account for these changes. The fact that the marked PDF shifts that have occurred in the past 20 years coincides with the warmest part of the temperature record should not be considered, at this stage of the analysis, as necessarily constituting a direct link with the enhanced greenhouse effect. While this may indeed be one of the underlying causes, there are many other possible explanations that could explain the observations, e.g. changes in circulation patterns; behaviour of the North Atlantic Oscillation; changes in cloudiness and soil moisture characteristics, to name but a few. This type of analysis should ideally be undertaken in many other regions of the globe to assess whether what is reported here is a common feature […].


In addition, because of the links between extremes of precipitation and pressure, it would be of interest to assess in a future investigation how these relationships may function in a warming climate. Regional climate model (RCM) projections suggest that the upper quantiles of current climate will become the norm under greenhouse-gas scenario climates (IPCC, 2007). According to many RCMs, summertime precipitation in a scenario climate may diminish by up to 30% or more compared to today (e.g. Beniston et al., 2007), such that the temperature biases linked to more extensive dry periods can be expected to be even greater than today, estimated at roughly 1 °C for each additional 10 days without rain.




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

The data employed in the present paper stem from the digital climatological database of the Swiss weather service, MeteoSwiss. This investigation has compiled data from both low (Basel, 369 m asl; Neuchâtel, 487 m; and Zurich, 569 m) and high-elevation sites (Engelberg, 1018 m; Davos, 1590 m; and Saentis, 2500 m) […].The data sets used have been quality checked by MeteoSwiss in terms of homogeneity in the records and continuity in the geographical location of the measurement stations (Begert et al., 2005). As such, Switzerland is an ideal locale for diurnal to century scale climatological time-series analyses, as has been shown in earlier publications (e.g. Beniston and Diaz, 2004; Schaer et al., 2004; Beniston, 2005). In the discussions, sets of daily data are used to compute seasonal means, i.e. according to the commonly used three-monthly periods December-January-February (DJF), March-April-May (MAM), June-July-August (JJA), and September-October-November (SON).

The reference 10 and 90% quantile thresholds are calculated using the daily temperature, precipitation and pressure data for each season for the 30-year reference period 1961–1990, i.e. 30 years × 90, 91 or 92 days, or 2700–2760 data points, according to the season. The thresholds calculated in this manner then serve to define the exceedances for all the time periods considered in this paper.

The computation of joint tails of the PDFs of two climate variables V1 and V 2 involves a fairly simple procedure whereby simultaneous exceedances of V1 and V 2 for combinations of the 10 and 90% quantiles are computed, i.e. V 110/V 210, V 110/V 290, V 190/V 210, and V 190/V 290; subscripts 10 and 90 refer to the respective quantile level for variable V1 and V 2. Frequencies are computed by counting the number of occurrences above or below a particular quantile threshold for each season and each year.


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










Sensibilité du milieu à des paramètres climatiques

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




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










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 :

Begert M, Schlegel T, Kirchofer W. 2005. Homogeneous temperature and precipitation series of Switzerland from 1864 to 2000. International Journal of Climatology 25: 65–80.

Beniston M. 2005. Warm winter spells in the Swiss Alps: Strong heat waves in a cold season? Geophysical Research Letters 32: L01812.

Beniston M, Diaz HF. 2004. The 2003 heat wave as an example of summers in a greenhouse climate? Observations and climate model simulations for Basel, Switzerland. Global and Planetary Change 44: 73–81.

Beniston M, et al. 2007. Future extreme events in European climate: An exploration of Regional Climate Model projections. Climatic Change 81: 71–95.

IPCC. 2007. Climate Change, the IPCC 4th Assessment Report. Cambridge University Press: Cambridge, UK.

Sch¨ar C, et al. 2004. The role of increasing temperature variability in European summer heat waves. Nature 427: 332–336.