EGU General Assembly 2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Persistence and frequency of drought-relevant circulation types during temperature extremes in southern Central Europe

Selina Thanheiser, Markus Homann, Andreas Philipp, Christoph Beck, and Jucundus Jacobeit
Selina Thanheiser et al.
  • Augsburg, Physical Geography and Climate Science, Geography, Augsburg, Germany (

The German weather service reports a new record mean June temperature for Germany and intensive heat waves during 2018 and 2019. Between January 2018 and June 2019, three new monthly top extremes were recorded (April 2018, May 2018 and June 2019).

In this study the relationships between the persistence and frequency of atmospheric circulation patterns related to drought and surface air temperature anomalies are investigated. The study area is in southern Central Europe, including parts of Germany and Switzerland as well as Austria and Czech Republic.

Large-scale atmospheric circulation types (relevant to drought) have been derived by using the COST733 classification software. Atmospheric variables from gridded daily JRA55 reanalysis data (Japanese Meteorological Agency 2018) and gridded precipitation data for the study area (6x6km, based on timeseries of 1756 weather stations from Zentralanstalt für Meteorologie und Geodynamik 2018) were used for the classification. All input variables were specifically weighted in the classification process. Daily maximum temperature data from ECA&D (2019) for different stations within the study area are used to evaluate the relationship between a circulation type and heat (cold) waves.

The drought-relevant circulation types are determined according to relative frequencies of circulation type days under a particular percentile of precipitation: If at least 20 percent of the circulation type days are below the 20th percentile of precipitation, the circulation type is defined as drought relevant.

For the derived drought-relevant circulation types, the mean seasonal frequencies [in %] (April-September, October-March) and the mean persistence [in days] (1961-2017) are calculated. To evaluate the relationship between a circulation type and heat (cold) waves, an efficiency coefficient is calculated. The efficiency coefficient is defined as ratio between the frequency of the circulation type in heat (cold) waves and its mean seasonal frequency.

For the study area, those circulation types relevant to drought with a high proportion of seasonal temperature anomalies could be identified. The circulation type with a dominant Azores high with ridges of high-pressure towards Central/Eastern Europe has the highest proportion of positive temperature anomalies in summer.

How to cite: Thanheiser, S., Homann, M., Philipp, A., Beck, C., and Jacobeit, J.: Persistence and frequency of drought-relevant circulation types during temperature extremes in southern Central Europe, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-21274,, 2020

Comments on the presentation

AC: Author Comment | CC: Community Comment | Report abuse

Presentation version 2 – uploaded on 06 May 2020 , no comments
Revised and extended version!
Presentation version 1 – uploaded on 30 Apr 2020
  • CC1: Comment on EGU2020-21274, Chiem van Straaten, 05 May 2020

    Dear Authors,

    Thank you for this interesting work. I like how you present timeseries of the relative frequencies and relative persistence of the drought-relevant circulation types

    I have a question regarding the order in your clustering methodology. Do I understand it correctly that your precipitation timeseries are used within the SANDRA method to get drought relevant clusters or is the conditioning done afterwards by looking at the subset of your general clusters that is most frequent during drought? In case you used the former I would like to learn more about that method. Could you point me to relevant publications?

    Kind regards,

    Chiem van Straaten (KNMI / VU Amsterdam)

    • AC1: Reply to CC1, Selina Thanheiser, 05 May 2020

      Dear Chiem van Straaten,

      thank you for your interest. You have understood correctly. The precipitation time series are used within the SANDRA method to get drought relevant clusters.

      To find the relevant circulation types we use non-hierarchical cluster analysis. SANDRA is a clustering scheme combining the concepts of simulated annealing and diversified randomization. As we know from the previous project, this method provides an optimal division of classes. We use a medium number of 18 classes. This prevents that by too small classes, special cases dominate a class that cannot be represented in the climate models.

      The classification is conditioned to the precipitation time series as a target variable using the below average percentile of precipitation. Because we define drought as precipitation under the long-term average. The reason is that we do not investigate extreme dry days, but periods of drought. We use six precipitation time series. One for each region of similar precipitation variability within the study area. These time series result from an s-mod PCA that was applied to our data.

      Here you can find more information: 

      Jacobeit, J., M. Homann, Philipp, A., Beck, C. (2017). Atmospheric circulation types and extreme areal precipitation in southern central Europe. Adv. Sci. Res., 14, 71-75.

      Hofstätter M., J. Jacobeit, M. Homann, A. Lexer, B. Chimani, A. Philipp, C. Beck and M. Ganekind (2015): WETRAX - WEather Patterns, Cyclone TRAcks and related precipitation EXtremes. Geographica Augustana, Manuskripte Band 19.

      Beck C., A. Philipp and J. Jacobeit (2007): An intercomparison of selected circulation type classifications for the European region.  Geophysical Research Abstracts, Vol. 9, 10659, 2007

      Beck C. (2008): Quantitative evaluation and comparison of circulation classifications – examples from the EU COST Action 733. COST 733 mid-term conference: "Advances in weather and circulation type classifications and applicatons", Cracow, 2008. (invited)

      Philipp A., C. Beck, R. Huth and J. Jacobeit (2016): Development and comparison of circulation type classifications using the COST 733 dataset and software. Int. J. Climatol. 36: 2673–2691.

      Best Regards,

      Selina Thanheiser

      If you have further questions you can send an email as well: