New study explains unusual 2015/16 El Niño heat budget

Michael Mayer, Magdalena Alonso Balmaseda, Leopold Haimberger (University of Vienna)


A new study uses ECMWF’s ocean reanalysis ORAS5 to show that the unusual lack of cooling of the tropical Pacific upper ocean during the 2015/16 El Niño event can to a large extent be attributed to changes in heat transport from the Pacific into the Indian Ocean, which were in turn caused by changes in the Indian Ocean state.

An unusual El Niño

The 2015/16 El Niño was a strong event with peak sea-surface temperatures in the equatorial Pacific similar to those of the exceptional 1997/98 El Niño. Usually tropical Pacific Ocean heat content decreases during El Niño events, mainly due to enhanced evaporation associated with the anomalously warm sea-surface temperatures. This process usually leaves an anomalously cold tropical Pacific, which prepares the ground for a swing towards relatively cool La Niña conditions and the start of a new cycle.

A large fraction of the heat lost from the tropical Pacific Ocean during an El Niño event is transported out of the tropical Pacific region by the atmosphere, fuelling changes in weather patterns worldwide (teleconnections). The heat is eventually reabsorbed by the ocean in other regions of the world or radiated into space. Only a relatively small fraction of the exchanged heat is immediately lost to space via radiation in the tropical Pacific region. The net effect of typical El Niño events is thus to cool the tropical Pacific Ocean and the Earth system as a whole. For example, during the 1997/98 El Niño, the upper 300 m of the tropical Pacific (averaged between 30°S and 30°N) lost about 11.5 zettajoules (1 zettajoule = 1021 joule) to other parts of the Earth system, which is equivalent to about 20 years of world energy consumption.

Upper ocean heat content changes. The changes are shown for the periods January 1997 to December 1998 and January 2015 to December 2016, based on ORAS5. Arrows and associated numbers (in zettajoules) indicate the direction and strength of two-year accumulated heat transport and flux anomalies across the boundaries of the tropical Pacific. For example, the positive numbers associated with the ITF represent reductions in westward ITF heat transports during the El Niño events. Integrated ocean heat content changes are given for 0–300 m and for the full-depth ocean (in brackets). Signs are chosen such that positive (negative) values indicate a contribution to a warming (cooling) of the Pacific.

However, the energetics of the 2015/16 event deviated substantially from this canonical behaviour. While there was sizeable cooling along the equator, the tropical Pacific upper ocean as a whole, instead of losing heat, gained about 9.6 zettajoules of heat during the 2015/16 El Niño. The absence of tropical Pacific cooling during the 2015/16 El Niño may explain why subsequent cool La Niña conditions were only weak and short-lived. Several unexpected weather extremes, such as the strong hurricane season in the Eastern Pacific during 2016 and the ’coastal El Niño’ in 2017 that brought torrential rain to Peru and Colombia, can be attributed to the near-absence of La Niña conditions in 2016/17.

Use of ocean reanalysis

The study used various atmospheric and oceanic datasets to investigate the reasons for the differences between 2015/16 and 1997/98 in the coupled atmosphere–ocean energy budget. Satellite-based radiation data from CERES instruments and atmospheric reanalyses (e.g. ERA-Interim) were used for atmospheric diagnostics, while the oceanic part of the study strongly relies on ORAS5. Ocean reanalysis ingests various types of ocean observations (e.g. in-situ temperature/salinity profiles and satellite altimeter data) as well as boundary fluxes derived from atmospheric reanalysis into a dynamical ocean model to obtain a physically consistent estimate of the evolution of the ocean state. One particularly relevant output from ocean reanalysis is oceanic transports, which are relevant for regional oceanic heat budgets but difficult to measure directly.

Role of Indian Ocean state

An analysis of the heat budgets of the 2015/16 and 1997/98 El Niños reveals that differences in ocean heat export from the tropical Pacific can explain about 74% of the difference in the tropical Pacific Ocean heat content evolution between the two events. The main cause for this difference in oceanic heat export was the different behaviour of the Indonesian Throughflow (ITF), which usually transports warm waters from the Pacific into the Indian Ocean. Heat transports associated with the ITF during 2015/16 were weaker than at any other time in the ORAS5 record (which begins in 1975). Further exploration shows that this weakening was linked to the relatively large warming and associated sea level rise in the Indian Ocean compared to other tropical ocean basins over the last decade, which has acted to decrease westward ITF volume and heat transports. The reduction of ITF heat transport in 2015/16 was strong enough to outweigh the surface heat loss of the tropical Pacific, leading to a net warming.

The results of the study point to the need for a realistic representation of Indo-Pacific energy transfers in ocean reanalyses and numerical models to further improve seasonal-to-decadal predictions. In particular, the correct initialisation of the sea level and associated temperature gradients between the Indo-Pacific basins plays a role in ENSO (El Niño–Southern Oscillation) predictions. Seasonal prediction experiments are currently under way at ECMWF to better characterise the impact of the Indian Ocean state on the evolution of El Niño events.

More details on the study can be found in an article published by the journal Geophysical Research Letters (doi:10.1002/2018GL077106).