A study of wintertime circulation anomalies during past El Niño events using a high resolution general circulation model. II: Variability of the seasonal mean response

Abstract

AbstractIn this paper the response of a (relatively) high‐resolution 11‐layer general circulation model (GCM) to a number of different sea surface temperature (s.s.t.) anomalies in the tropical East and West Pacific is studied. The integrations were all run for 540 days in a perpetual January mode and had either mean or envelope orography.The s.s.t. anomalies included an enhanced composite El Niño anomaly, the observed anomalies during December 1982, and December‐February 1982–3, and a weak (∼1 K) positive s.s.t. anomaly in the tropical West Pacific.The objective of this study is to assess whether the observed variability in wintertime circulation anomalies between past El Niño events can be modelled by considering the 90‐day mean (‘interannual’) variability of the integration with the composite El Niño s.s.t. anomaly. It is concluded that such observed variability cannot be modelled in this way and that variations in tropical Pacific s.s.t. anomaly from one event to another are crucial in accounting for observed extratropical variability over the Pacific and North America. In particular it is shown that the model's response to the 1982/3 anomaly and the weak West Pacific anomaly are significantly different from the response to the composite El Niño anomaly over the Pacific and North America, and agree well with the appropriate observations for 1982/3 and 1976/7 respectively. It is also concluded that the relatively weak extratropical response during the winter of 1972/3 was related to the fact the s.s.t. anomalies were confined to the East Pacific. By contrast the response to the weak West Pacific anomaly was very strong and statistically significant over the Pacific and North America.