Réf. Karagiannidis & al. 2012 - A

Référence bibliographique complète

KARAGIANNIDIS, A.F., KARACOSTAS, T., MAHERAS, P., MAKROGIANNIS, T. 2012. Climatological aspects of extreme precipitation in Europe, related to mid-latitude cyclonic systems. Theoretical and Applied Climatology, Vol. 107, 165–174.

Abstract: Climatic aspects of extreme European precipitation are studied. Daily pluviometric data from 280 stations across Europe, covering the period from 1958 to 2000, are used. First, the criteria for extreme precipitation cases and episodes are communicated using threshold and spatial definitions. The cases and episodes meeting these criteria are grouped according to their area of appearance. Most of them are located in three major areas: Greece, the Alps, and the Iberian Peninsula. The existence of trends in the annual and seasonal time series of these extreme events is examined. Decreasing trends are found in most of the cases, for Greece, the Iberian Peninsula, and Europe, as a whole. The Alps present a different behavior, with no trend at all in the southern part, and a possible increasing trend in the northern part. Finally, the positive impact of altitude in the frequency of occurrence of extreme precipitation episodes in Europe is discussed.




Organismes / Contact

• Department of Meteorologyand Climatology, School of Geology, Aristotle University, Thessaloniki, Greece (thankar@live.com)


(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




[Review]: Many researchers have investigated the extreme precipitation of the European continent for the existence of possible trends. The conclusions may vary, according to the scope, methodology, and database of each study [see details and references in the study]. Research was performed on the precipitation on specific European sub-regions. The conclusions of these studies are not always in complete agreement with each other. [Some papers] studied the relation between orography and precipitation. They have shown that the orography affects substantially the spatial characteristics and the amount of precipitation. A general increase of precipitation with altitude was reported, although the geomorphology of each different area can affect precipitation in a unique way.

[This study]:


Analysis of the time variation of the total number of spring extreme precipitation cases for Europe as a whole reveal a negative trend, which becomes statistically significant during the period 1980–1997. This conclusion is valid only for the continental Europe, the British Islands, and Ireland.

From the results of the application of linear regression analysis and Kendall-tau test on all the examined time series, and the results of the application of the non-parametric Mann–Kendall test, Europe presents statistically non-significant negative linear trends, when the full length of the series is examined. Non-significant negative trends are also evident for all the seasonal time series. However, the Mann–Kendall test shows short periods of statistically significant negative trends for all-months, winter, and spring and a non-significant negative trend for autumn.

The different areas and periods of examination, along with the various methodologies used in the studies [reviewed in the paper], make the comparison of the findings of the present study a difficult task. Moreover, the exclusion of the summer months is a justified parameter of discrepancy between the conclusions of this study and those of other ones that do not exclude summer months. Nevertheless, a general agreement, at least with the 4th IPCC Assessment Report, seems to exist.

Alpine region:

No significant trends are observed for the examined time series of the Alpine region. Only small fluctuations of extreme precipitation are noted, and they do not constitute abrupt climatic changes or statistically significant trends. Switzerland is considered to be representative of the northern part of the Alpine region, while northern Italy is considered to be representative of the southern part. Therefore, the Alpine region could also be examined as two separate parts. The northern part, represented by Switzerland, presents an upward trend for all-months time series. A statistically significant positive linear trend is found for the period 1958–2000, while the implementation of the Mann–Kendall test localizes the trend during the years 1989–2000. The investigation of the seasonal time series shows no linear trends but only statistically non-significant positive trends for winter and spring extreme precipitation. Northern Italy, on the other hand, presents no statistically significant trend in the seasonal and all-months time series.

Frei and Schär (2000) developed a methodology for the computation of the probability of rare events and studied the extreme precipitation trends in the Alps. The number of the stations presenting significant trends was very small for all seasons. They concluded that there is no significant trend in the Alpine precipitation for the period 1901–1994. Schmidli and Frei (2005) implemented two different methodologies for analyzing trends and investigating their statistical significance. They found a positive statistically significant trend in extreme winter precipitation in many stations of the northern and northwestern side of the Alps during the period 1901–2000. A negative trend was found in the southern and southeastern side, but it was not statistically significant. A statistically significant negative trend was found in the autumn in the southern side only. Brunetti et al. (2001), studying northeastern Italy time series, concluded that there is no trend in extreme precipitation cases. However, mean intensity, defined as mean precipitation amount per day, was found to be increasing, but not at a significant level. Using daily precipitation amount time series for a great number of Italian stations, Brunetti et al. (2002) found no significant trend in the extreme precipitation. In their paper, precipitation was considered as extreme, when it exceeded 10% of the mean annual amount in a series of five consecutive days. Most of the studies mentioned before agree with the conclusions of the present one, verifying, among others, the interesting differentiation between the northern and the southern Alpine region.

The effect of altitude on extreme precipitation:

The difference between the percentage of the number of cases and the percentage of the number of stations increases with the increase of the altitude. This difference is −6% for the 0–100 m class and −11% for the 100–200 m class. For the three next classes (200–300, 300–500, and 500–700 m), the difference is around zero (1%, 1%, and −1%, respectively), while for the last four classes is positive. For the 700–1,000 m class, the difference is 6%, for the class 1,000–1,500 m is 3%, for the 1,500–2,000 m is 5%, and finally for the class over 2,000 m is 3%. The increase of the station altitude results in an increase of the number of extreme precipitation cases. Up to 700 m, elevation seems to play a minor role, while a distinct increase of extreme precipitation is observed at altitudes over 700 m.

This effect of the orography to the volume and spatial characteristics of precipitation has already been examined by many researchers. […] All the studies mentioned [in the paper] concluded that there is a definite influence of the altitude to precipitation. However, due to the great variety of parameters that contribute to the final manifestation of this influence, most of the studies are spatially limited to a specific mountainous area. Nevertheless, the agreement between the general conclusions of the present study and those of the ones mentioned proves the clear positive influence of the altitude in extreme precipitation.






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

Although bibliography seems to be extensive on the subjects of extreme precipitation, very few studies examined the climatic characteristics of extreme precipitation that was produced by mid-latitude cyclonic systems only. In this study, some climatological aspects of extreme precipitation related to mid-latitude cyclonic systems are investigated.


In order to study the extreme precipitation related to cyclonic systems, thermally induced thunderstorms should be excluded. A criterion based on the different characteristics of these two processes of precipitation is set. This criterion also defines and distinguishes the “case” and “episode” of extreme precipitation. After the identification of all the cases and episodes of extreme precipitation produced by mid-latitude cyclonic systems, they are grouped according to their area of occurrence. Europe, as a whole, and specific sub-regions are investigated for possible trends, using the non-parametric Mann–Kendall test and the linear regression theory. The effect of orography on the frequency of appearance of extreme precipitation is also examined. The pluviometric stations are clustered into nine classes, according to their altitude. The number of stations of each class is compared to the number of cases of extreme precipitation recorded at the stations of the class, in order to reveal the positive relation between the increase of extreme precipitation and the altitude.

[Data]: Daily precipitation data, from 280 pluviometric stations across Europe, are used in the present study. The covered period extends from 1958 to 2000. These data were obtained from European Climate Assessment & Dataset and the Meteorological Services and Institutes, through the Statistical and Regional dynamical Downscaling of Extremes for European Regions (STARDEX) program, after its completion. [The stations] cover the continental Europe, up to 30° E longitude, and the British Islands and Ireland . Western Europe, the UK , Ireland , Czech Republic , Hungary , Romania , and Greece are covered quite satisfactorily by the pluviometric network. The Balkan area (except Greece and Romania), Poland, and central and southern Italy present a rather sparse network coverage, which may lead to partial “loss” of episodes of extreme precipitation in these areas. However, this problem does not affect the study substantially, since the objective is not to record every single extreme precipitation episode but to investigate specific climatological features of them. […]

Thermal thunderstorms, along with “weak cyclones of thermal origin” (Maheras et al. 2001), tend to develop mainly in the warm season of the year and the southern parts of Europe. They are associated with high intensity and significant amount of rainfall. The majority of them are local and limited area systems, in contrast to mid-latitude cyclonic systems which extend to much wider areas, and they usually take place during the passing of short wave disturbances of the middle and upper troposphere or even days of anticyclonic circulation. In order to avoid such events and not to incorporate within the sample, summer months (June, July, and August) were excluded from this study.

[Definitions of “case” and “episode” of extreme precipitation]: Although thermal thunderstorms tend to develop during summer, they can also occur during spring or autumn. In order to completely eliminate them from the study, a strict criterion is set and applied. This criterion defines the case and episode of extreme precipitation. The case of extreme precipitation applies to a single station and is defined as the recording of over 60 mm/day of precipitation at a single station. […] The precipitation threshold value of 60 mm/day and the maximum distance between neighboring stations (d max=500 km) are not set arbitrarily. Their setting is based on the different spatial and temporal characteristics of thermally induced and synoptic scale precipitation events. […] Three thousand two hundred thirty-three cases of extreme precipitation are found across the examined area, during the period 1958–2000.

In order to investigate the linear trends that may exist in extreme precipitation, linear regression analysis is used. The least squares line of each time series is calculated. The slope of this line represents the linear trend of the series. Then, the Kendall-tau statistic is applied to the linear trend, in order to determine its level of significance. [see methodology and references  in the study]

[Analysis of trends in extreme precipitation]: Two kinds of time series are examined. Those which consist of the number of extreme precipitation cases for the period from September to May, called “all-months” time series, and those which consist of the number of extreme precipitation cases for each of the seasons: autumn (September to November), winter (December to February), and spring (April to May). All-months and seasonal time series of extreme precipitation cases, for the whole European continent and three specific sub-regions of it, are subjected to linear regression analysis, the Kendall-tau test, and the non-parametric Mann–Kendall test, in order to reveal any possible trends. The sub-regions, which were selected because they accumulated most of the extreme precipitation episodes, are Greece , the Alpine Region, and the Iberian Peninsula .


Europe is a continent which, despite its relatively small size, has a remarkable number of mountains, some of them having peaks over 3,000 m. In the present paper, a study of the effect of altitude to the precipitation is attempted. The frequency of occurrence of extreme precipitation as a function of the altitude is examined. The pluviometric stations are divided in nine classes, according to the station’s altitude. […]


(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

Linear regression analysis and the non-parametric Mann–Kendall test are used to examine the possible trends of extreme precipitation in Europe and specific sub-regions of it. The effect of altitude increase in extreme precipitation is also examined.

·         Continental Europe as a whole, together with Great Britain and Ireland, do not present any statistically significant linear trends for the total of the examined period. However, short-term but statistically significant negative trends are revealed for the all-months, winter, and spring time series. The time period that hosts these trends varies between the different series. In the case of all-months time series, the negative trend starts in the late 1960s and reaches 2000. This trend could be associated to an analogous decrease of mid-latitude cyclonic systems in Europe.

·         The negative trend that is recorded when the continent was examined as a whole is not manifested in exactly the same form in the sub-regions which were examined. Therefore, the climatic change that could be associated to this trend is not manifested in the same way or severity in the whole continent.

·         Greece presents linear negative trends in extreme precipitation cases of all-months, autumn, and winter. The application of the Mann–Kendall test showed that these trends become statistically significant during the last decade, or the last two decades, if the winter time series is considered.

·         No trend is found in the extreme precipitation of the Alpine region as a whole. Lack of trend is also observed in the northern Italy, which is representative of the southern side of the Alps. On the contrary, in Switzerland, which is considered to represent the northern side of Alps, a positive trend was revealed in the all-months time series. With the application of Mann–Kendall test, this positive trend was found to be statistically significant, throughout the last decade.

·         An almost monotonous linear decrease in extreme precipitation was revealed for the Iberian Peninsula, when all-months, winter, and spring time series were examined. This decrease is statistically significant during the last two decades, with the exception of winter time series where the trend is constrained in the last 2 years of the examined period. Portugal, which is influenced by the Atlantic Ocean, presents analogous negative trends, which become significant during the last decade. Spain, mainly influenced by the Mediterranean, presents a similar negative trend. With the aid of the Mann–Kendall test, the trend is found to be statistically significant during the last two decades.

·         Extreme precipitation trends seem to vary according to the change of geographic latitude. Regions of northern geographic latitude lower than 43° (Iberian Peninsula and Greece) present a decrease of the extreme precipitation associated to mid-latitude cyclonic systems. In northern regions, specifically in the southern side of the Alps (northern latitude 43–45°), no significant trend was found. Finally, in the northern side of the Alps (northern latitude 45–47°), a statistically significant positive trend was revealed. These conclusions are similar to the ones presented by the STARDEX program (STARDEX 2005).

·         The comparison between the trends of total precipitation and those found in the present investigation shows that they are similar enough in sign and period of appearance. This fact confirms the hypothesis that extreme precipitation associated to mid-latitude cyclonic systems changes with time in patterns analogous to those of total precipitation.

·         Finally, it was shown that the increase of the altitude, especially over 700 m, increases the number of extreme precipitation cases recorded. This fact demonstrates the definite and positive in general effect of the altitude in extreme precipitation.

Références citées :

Brunetti M, Maugeri M, Nanni T (2001) Changes in total precipitation, rainy days and extreme events in northeastern Italy. Int J Climatol 21:861–871.

Brunetti M, Maugeri M, Nanni T, Navarra A (2002) Droughts and extreme events in regional daily Italian precipitation series. Int J Climatol 22:543–558.

Maheras P, Flocas HA, Patricas I, Chr A (2001) A 40 year objective climatology of surface cyclones in the Mediterranean region: spatial and temporal distribution. Int J Climatol 21:109–130.

Frei C, Schär C (2000) Detection probability of trends in rare events: theory and application to heavy precipitation in the Alpine region. J Climate 14:1568–1584.

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

STARDEX (2005) http://www.cru.uea.ac.uk/projects/stardex/deliverables/D9/. Accessed on 18 Nov 2010