Development and comparison of circulation type classifications using the COST 733 dataset and software

In order to examine correspondence between different methods for circulation type classification, a dataset of classification catalogs for 12 different European regions has been created using a specially developed software package. Twenty-seven basic automatic classification methods have been applied in several variants to different input datasets describing atmospheric circulation. Together with six manual classifications a total of 33 methods are available for inter-comparison. Pattern correlation, frequency time-series correlation and the adjusted Rand index have been used for comparison. Highly significant correspondence has been detected only for two clustering techniques while the remaining classification methods show surprisingly low similarity. A Monte-Carlo test with 1000 classifications of randomly defined types even shows that most of the methods are not more similar among each other than any arbitrarily chosen types. The predominant dissimilarity between the methods is interpreted to be a result of a lack of inherent structures of the input data. Only simulated annealing clustering and self-organizing maps get nearly identical results because they can optimally fit the partitioning to the outer shape of the data cloud in the phase space. Also methods based on pre-defined types come to very different results because small changes in the definition of thresholds may lead to large differences in the partitioning. It is concluded that because of the missing inner structure of the data there is no clear statistical reason to prefer any of the examined methods. For practice in synoptic climatology this means that finding a suited classification for a certain purpose may require a broad comparison of methods. The software package cost733class for development, comparison and evaluation of classifications which was developed and used in this study is available at http://cost733.geo.uni-augsburg.de to facilitate this task.

[1]  Bo Christiansen.  On the physical nature of the Arctic Oscillation , 2002 .

[2]  L. Kalkstein,et al.  An Evaluation of Three Clustering Procedures for Use in Synoptic Climatological Classification , 1987 .

[3]  C. Beck.  Are there weekly cycles in occurrence frequencies of large‐scale circulation types? , 2012 .

[4]  Ian T. Jolliffe,et al.  Some recent developments in cluster analysis , 2010 .

[5]  Andreas Philipp,et al.  Classifications of Atmospheric Circulation Patterns , 2008, Annals of the New York Academy of Sciences.

[6]  Jucundus Jacobeit,et al.  Frequency and within‐type variations of large‐scale circulation types and their effects on low‐frequency climate variability in central europe since 1780 , 2007 .

[7]  Bo Christiansen,et al.  Atmospheric Circulation Regimes: Can Cluster Analysis Provide the Number? , 2007 .

[8]  William M. Rand,et al.  Objective Criteria for the Evaluation of Clustering Methods , 1971 .

[9]  Mike Hulme,et al.  A COMPARISON OF LAMB CIRCULATION TYPES WITH AN OBJECTIVE CLASSIFICATION SCHEME , 1993 .

[10]  Andreas Philipp,et al.  Comparison of principal component and cluster analysis for classifying circulation pattern sequences for the European domain , 2009 .

[11]  Iver A. Lund.  Map-Pattern Classification by Statistical Methods , 1963 .

[12]  L. Mcquitty.  Similarity Analysis by Reciprocal Pairs for Discrete and Continuous Data , 1966 .

[13]  Radan Huth,et al.  AN INTERCOMPARISON OF COMPUTER‐ASSISTED CIRCULATION CLASSIFICATION METHODS , 1996 .

[14]  David B. Stephenson,et al.  On the existence of multiple climate regimes , 2004 .

[15]  P. Jones,et al.  Long-Term Variability of Daily North Atlantic–European Pressure Patterns since 1850 Classified by Simulated Annealing Clustering , 2007 .

[16]  Camille Roth,et al.  Natural Scales in Geographical Patterns , 2017, Scientific Reports.

[17]  D. Steinley.  Properties of the Hubert-Arabie adjusted Rand index. , 2004, Psychological methods.

[18]  R. Vautard,et al.  Weather Regimes: Recurrence and Quasi Stationarity , 1995 .

[19]  Danny Blair,et al.  The Kirchhofer technique of synoptic typing revisited , 1998 .

[20]  Peter J. Rousseeuw,et al.  Finding Groups in Data: An Introduction to Cluster Analysis , 1990 .

[21]  John A. Hartigan,et al.  Clustering Algorithms , 1975 .

[22]  J. H. Ward.  Hierarchical Grouping to Optimize an Objective Function , 1963 .

[23]  Pere Esteban,et al.  Atmospheric circulation patterns related to heavy snowfall days in Andorra, Pyrenees , 2005 .

[24]  Andreas Philipp,et al.  Cluster Analysis of North Atlantic-European Circulation Types and Links with Tropical Pacific Sea Surface Temperatures , 2008 .

[25]  Radan Huth,et al.  Ein Beispiel für die Anwendung der Hauptkomponentenanalyse zur Auffindung von Zirkulationstypen über Europa , 1993 .

[26]  Christos N. Schizas,et al.  Synoptic Classification and Establishment of Analogues with Artificial Neural Networks , 2007 .

[27]  W. Enke,et al.  Downscaling climate model outputs into local and regional weather elements by classification and regression , 1997 .

[28]  András Bárdossy,et al.  Statistical comparison of European circulation patterns and development of a continental scale classification , 2003 .

[29]  R. Barry,et al.  Synoptic and dynamic climatology , 1980 .

[30]  Judit Bartholy,et al.  Cost733cat - A database of weather and circulation type classifications , 2010 .

[31]  Brent Yarnal,et al.  Developments and prospects in synoptic climatology , 2001 .

[32]  P. Legendre.  Spatial Autocorrelation: Trouble or New Paradigm? , 1993 .

[33]  A. Sterl,et al.  The ERA‐40 re‐analysis , 2005 .

[34]  Roger G. Barry,et al.  Comparison of the MSL synoptic pressure patterns of the Arctic as observed and simulated by the GISS General Circulation Model , 1988 .

[35]  P. Clifford,et al.  Modifying the t test for assessing the correlation between two spatial processes , 1993 .

[36]  G. Plaut,et al.  Large-scale circulation classification, weather regimes, and local climate over France, the Alps and Western Europe , 2001 .