• northwest Iberian Peninsula;
  • rainfall;
  • sea surface temperatures;
  • North Atlantic;
  • linear regression;
  • predictability


This study assesses the relationship between monthly North Atlantic sea surface temperature anomalies (SSTAs) and a regional index of rainfall (NWIPR) in northwest Iberian Peninsula during the period 1951–2006 by means of the Pearson product-moment correlation. Results show a strong influence of SSTA on NWIPR in several months (February, April, May, October and December). The observed persistence of this influence up to 2-months in advance can be used for monthly predictions of rainfall. The most significant areas of the North Atlantic were clustered to be used as input variables in linear regression models. Correlations up to 0.59 between observed and predicted rainfall anomalies were obtained. Predictability ranged from 76 to 86% when considering rainfall as a discrete predictand. Copyright © 2009 Royal Meteorological Society