FORECASTERS look to our oceans to predict the coming season – it is well-known that interactions between the air and oceans have profound influences on Australia’s climate. The formation of El Niño and La Niña in the tropical Pacific Ocean, for example, has global reach, affecting society and the environment. But in a paper published in Science this month, climate scientists say we need to look beyond our neighbouring oceans to understand changes several seasons ahead – to the Atlantic Ocean on the other side of the world.
Dr Wenju Cai, Director of CSIRO’s Centre for Southern Hemisphere Oceans Research (with the oceanographically apt acronym C-SHOR, pronounced “sea shore”), says that understanding the influence of the Atlantic on the Pacific and Indian Oceans can provide more accurate seasonal predictions, and provide a clearer picture of climate change over the coming decades.
"The Atlantic can influence the Pacific in terms of the El Niño – Southern Oscillation (ENSO),” says Cai.
"The little brother can dominate – even though it’s the smallest ocean, the Atlantic can play a big part. This realisation can be crucial for increasing predictability and awareness of ENSO.”
Cai and colleagues looked at how the Atlantic has increasingly influenced the Pacific, and how the Atlantic has also influenced the Indian Ocean, which in turn influences the Pacific.
"We have these three oceans that are tightly connected across the whole of the tropics. People used to think that the Indian Ocean was a slave to the Pacific. But that’s not true – the Indian Ocean can also influence El Niño. Cooling in the eastern Indian Ocean leads to high pressures that drive westerly winds, which is conducive to El Niño formation in the Pacific.”
To illustrate this influence, he explains that a positive phase of the Indian Ocean Dipole in 1997, the strongest in the 20th century, enhanced the 1997 El Niño that also turned out to be the strongest in the 20th century. Warming of the Indian Ocean subsequently consolidated a transition from El Niño to the 1998 La Niña – which also became the strongest in the 20th century.
"Extreme El Niños of the 20th century were preceded by Atlantic cooling. And when you have Atlantic warming, you tend to have La Niña the following year. Knowledge of this Atlantic influence gives a six- to nine-month lead-time and can help predict El Niño and La Niña. So, assimilating this information into climate models could improve seasonal prediction.”
The difficult next step to progress seasonal prediction is to develop climate models that capture the El Niño and La Niña precursors identified in the Indian and Atlantic Oceans.
"Seasonal prediction has been stagnant for 20 years, partly because the models focus on the Pacific, and the Pacific is not everything. So maybe we need to incorporate all these other potential precursors; that could provide a good avenue to progress. That’s the upshot of what we’re saying in the Science paper.”
Seasonal prediction has been stagnant for 20 years, partly because the models focus on the Pacific...
An understanding of the interconnection between the three oceans has implications on projections of climate over coming decades.
"Most climate models are not able to simulate the recent hiatus,” explains Cai, referring to the stagnation of land surface temperatures in the early 2000s despite the continuing increase in heating of the Earth system.
"Interactions between the oceans were largely responsible. Of 34 global climate models, only three or four properly simulated the hiatus. It turns out these were the ones that modelled the Atlantic’s influence on the Pacific. If you can’t simulate the influence of the Atlantic on the Pacific, and if the Atlantic is warming faster (which leads to a cooling in the Pacific), it means that the Pacific is warming too fast in the models.”
Dr Cai says if the modelled warming in the oceans is happening too fast, this could affect projections of extreme weather.
"The biggest source of uncertainty in climate projections is how the Atlantic’s influence on the Pacific will affect climate extremes.
"On the one hand, overestimated ocean warming could suggest more extremes; on the other, the effect on the atmosphere makes the weather more stable, leading to an underestimate of extremes. The balance between these two factors are unclear without an improved climate model.”
The biggest source of uncertainty in climate projections is how the Atlantic’s influence on the Pacific will affect climate extremes.
Other work by Dr Cai, published in Nature in December 2018, showed that rising global temperatures will lead to a greater number of extreme El Niños. The results, which Nature called a 'breakthrough' and 'a milestone in climate research', solved the decades-old mystery of how ENSO will change in a warming world.
"Climate models had given mixed results about strong El Niños,” says Cai. “For 30 years, people had tried to look at the influence of climate change on ENSO. We looked at where the modelled centre of anomalously warmer sea surface temperatures was located in each of the models, rather than its location in the real world. This showed that most models agreed on an increase in extreme El Niños.”
Dr Cai says the availability of ocean observations is important to validate models and show the strength of the Atlantic influence.
"With only 20 years of data, we can’t be sure a climate model is capturing these interactions properly. Any other records, such as palaeoclimate proxies, such as ancient coral records, will be important. We have little to go on in terms of learning about the Atlantic’s role in the climate system.”
Read more: "Pantropical climate interactions", by W. Cai et al., appears in the March 2019 (363) issue of Science, DOI: 10.1126/science.aav4236.