Resumen
A homogenized database of minimum temperatures from 135 observatories distributed around Europe was analyzed in order to estimate trends and to determine the main synoptic weather patterns that contribute to an extremely cold day (ECD). An ECD is defined as a day whose minimum temperature is within the lowest 5th centile of the daily temperature series for each observatory; these values represent temperatures below which public health concerns may be expected. The period from 1 January 1955 to 31 December 1998 was chosen for this analysis because it represents the period for which the greatest number of measuring stations with complete time series was available. The relationship between the occurrence of an ECD and the general circulation of the atmosphere was based on a statistical analysis in which a coefficient of effectiveness for each synoptic pattern was obtained. In order to simplify the manipulation of the data, a rotated principal components (RPC) analysis was applied and the synoptic patterns with the greatest contribution to the ECD were obtained. Trends for each synoptic pattern were studied in order to examine the relationship among the ECD.