Linking precipitation variability to large-scale atmospheric circulation

Atmospheric circulation, particularly in mid-latitudes, is the main control behind regional changes in temperature, precipitation and other climatic variables ( Slonosky et al. , 2000 ). There is a growing interest in linking atmospheric circulation to local climate. Yarnal et al. ( 2001 ) and Tveito and Ustrul ( 2003 ) provided an extensive review of the development of classification systems and their application in synoptic climatology. Manual classification schemes have usually a long history and were developed for specific regions usually reflecting the dominating features of the atmospheric circulation influencing the climate in the targeted region. In Europe, the Grosswetterlagen system according to Hess and Brezowsky (see Gerstengarbe and Werner, 1999 ) for central Europe is probably the best known ( Tveito and Ustrul, 2003 ). Objective classification systems were developed in the second half of the last century. Popular classification techniques are correlation-based analyses, eigenvector-based analyses (mainly two related techniques, principal components analysis and empirical orthogonal functions), composing and indexing ( Yarnal et al. , 2001 ). The P–27 classification scheme developed at KNMI using the 500 hPa heights for nearly the whole of Europe ( Kruizinga, 1979 ) is an example of the eigenvector-based classification systems. The North Atlantic Oscillation (NAO) index is the most well-known circulation index for Europe ( Hurrell, 1995 ). Considering the geographical location of the Meuse basin in Europe and the long time period of concern, the Grosswetterlagen system and the NAO index are of most interest in this study. Brief descriptions of the Grosswetterlagen system and the NAO index were given in sections 2.6.1 and 2.6.2 , respectively.