Accepted by Journal of Climate

• Main Authors: B. Dong, M. Collins
• Other Authors: The Pennsylvania State University CiteSeerX Archives
• Format: Text
• Language: English
• Published: 2005
• Subjects: Diaz; Mathieu; North Atlantic
• Online Access: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.156.2996; http://www.knmi.nl/publications/fulltexts/note_sutton.pdf

Linear teleconnections to winter climate in Europe are known to be very small (Kiladis and Diaz, 1989; van Oldenborgh et al., 2000), much smaller than in spring and fall. In Dec–Feb, of the 168 ECD (Klein Tank et al., 2002) precipitation stations (average record length 80 years), only 4 have a significant correlation with the Niño3 index (SST anomalies in 5 ◦ S–5 ◦ N, 150–90 ◦ W) at the 95 % level, even less than expected by chance alone. Of the 167 ECD temperature stations, only 2 show significant correlations. The correlations with 500mb geopotential height from the ERA-40 reanalysis (Gibson et al., 1997), which is less noisy, are also less than 0.2 over the whole area 30–80 ◦ N, 50 ◦ W–60 ◦ E except over northern Africa and southern Spain (Fig. 1a). These results suggest that the teleconnections noted in some of the earlier literature are mostly artifacts caused by lack of data. Mathieu et al. (2004) argue that ‘ENSO events have a significant influence on the climate of the North Atlantic European region ’ (in winter). They claim that the details of the SST anomalies determine the shape of the teleconnection, which can therefore not be described by linear correlations. This statement is