Differences in Atmospheric Circulation between the Development of Weak and Strong Warm Events in the Southern Oscillation

Abstract

In this study temporal and spatial aspects of El Niño (warm event) development are explored by comparing composite sequences of sea level pressure (SLP), surface wind, and sea surface temperature (SST) anomalies leading into strong and weak events. El Niño strength is found to be related to the magnitude and spatial extent of large-scale SLP anomalies that move in a low-frequency mode. In association with this, it is also intricately linked to the amplitude and wavelength of the Rossby waves in the southern midlatitudes. The primary signature of the Southern Oscillation is a more pronounced standing wave of pressure anomalies between southeastern Australia and the central South Pacific leading into stronger events. A strong reversal in the strength of the annual cycle between these two regions causes a stronger (weaker) SLP gradient that drives southwesterly (northwesterly) wind stress forcing toward (away from) the western equatorial Pacific in austral winter?spring of year 0 (?1). Thus, pressure variations in the southwest Pacific preconditions the equatorial environment to a particular phase of ENSO and establishes the setting for greater tropical?extratropical interactions to occur in stronger events. Maximum warming in the Niño-3 region occurs between April and July (0) when a strong South Pacific trough most influences the trade winds at both ends of the Pacific. Cool SST anomalies that form to the east of high pressure anomalies over Indo?Australia assist an eastward propogation of high pressure into the Pacific midlatitudes and the demise of El Niño. Strong events have a more pronounced eastward propogation of SST and SLP anomalies and a much more noticeable enhancement of winter hemisphere Rossby waves from May?July (?1) to November?January (+1). Weak events require an enhanced South Pacific trough to develop but have much less support from the North Pacific. They also appear more variable in their development and more difficult to predict with lead time.