Revisiting El Niño Modokis

Abstract

The suggestion that there exist two types of El Nino in the tropical Pacific has generated a debate in the community. Applying various linear and non-linear approaches and composite analysis technique on observed and reanalyzed climate datasets primarily for the 1950–2010 period, we revisit the variability of the tropical Pacific in the light of this debate. Our objective is to examine whether the proposed El Nino Modokis need a classification distinct from canonical El Ninos. Even if the distinction is subject to short data records, we demonstrate that the El Nino Modoki events indeed display a seasonal evolution and teleconnections different from the canonical El Ninos, and that the distinction is not subject to inclusion of the two extreme El Ninos 1982 and 1997 as canonical El Ninos. We show that the El Nino Modoki events are not an artifact associated with the orthogonality constraint associated with the EOF technique. Our cluster analysis shows that evolutions of the canonical El Nino and El Nino Modokis through various seasons differ from one another. Importantly, the dynamic and thermodynamic air–sea coupling strength is distinctly different between the El Nino Modoki and the canonical El Nino events. We find that, dynamic feedback intensity is stronger for El Nino Modoki (canonical El Nino) during boreal summer (winter); though the air–sea coupling strength, a major contributor to Bjerknes feedback, is maximum for Modokis during the developing stages, it decreases thereafter. In case of thermodynamic feedback intensity, SST-wind-evaporation feedback is dominant for El Ninos while SST-SHF feedback is important during El Nino Modokis. However, we find that the thermodynamic feedback values significantly differ across the flux datasets.