News & Views | Published:

Global change

Eruptions linked to El Niño

Statistical validation of a relationship between explosive volcanic eruptions and the El Niño/Southern Oscillation is a step forward in understanding the effects of such eruptions on climate.

One of the strongest influences on global climate is the cycle of events in the eastern and central tropical Pacific Ocean known as the El Niño/Southern Oscillation (ENSO). The cycle consists of periodic warming and cooling phases, El Niño and La Niña respectively, and conventional wisdom is that they are driven by natural interactions between the atmosphere and ocean in the tropical Pacific1. On page 274 of this issue, however, Adams et al.2 present a convincing case that over the past 400 years the ENSO phenomenon has been influenced by the after-effects of explosive volcanic eruptions. Not only do previous hypotheses about the geophysical processes driving the volcano–ENSO relationship3 now have theoretical and empirical support, but there is new life in an old debate about the role of eruptions in the Earth system.

Although they arise regionally, ENSO events have a worldwide impact because the tropical Pacific is a major heat source that drives atmospheric circulation. Small changes in this heat source, through changes in sea surface temperatures for instance, can have widespread consequences — increased snowfall in the Andes, weak hurricane seasons in the Atlantic, drought and associated crop failures in southern Africa, and even wildfires in Indonesia have all been related to ENSO. The climatic effects have repercussions on almost every aspect of human life, in the form of disease outbreaks, low and high agricultural yields, fishery catches, natural disasters and so on.

Stratospheric volcanic aerosols (suspensions of fine liquid or solid particles dispersed in air) are known to influence the absorption, reflection and transmission of solar and terrestrial radiation in the atmosphere. So, given that ENSO is susceptible to small changes in sea surface temperatures, it is natural to consider whether the phenomena are connected. Volcanic aerosols, injected into the stratosphere from explosive tropical eruptions, result in a complex interplay between stratospheric heating and surface cooling that produces an overall cooling of a few tenths of a degree globally that may last for a few years after the eruption4. Regionally, the result may be a change in tropospheric circulation and sea surface temperatures that leads to an ENSO event3.

Support for such a mechanism would come from a demonstration of a statistical connection between the timing of explosive volcanic eruptions and ENSO events. Earlier attempts at identifying such a connection were fraught with problems, most notably in their reliance on data recorded by networks of instruments. These data extend back only into the mid-nineteenth century5 and the limited sample size made it impossible to identify a statistically significant correlation between the ENSO cycle and volcanic eruptions. This and other methodological inadequacies have meant that the case for a volcano–ENSO link has remained unsupported4.

In their analysis, Adams et al.2 overcome the deficiencies of the instrumental record by using indices of El Niño and Southern Oscillation activity, derived from palaeoclimate data. They correlate these widely accepted measures of El Niño activity against two well-established independent reconstructions of explosive volcanic activity — the Volcanic Explosivity Index, a measure of the size of the eruption, and the Ice Core Volcanic Index, a measure of potential climate impact. This provides them with a significant dataset of events going back to the mid-seventeenth century. The authors use a standard approach to time-series analysis, the Seasonal Epoch Analysis, and enhance this by ranking, normalizing and preselecting key parameters to guard against potential bias or undue influence from anomalously large events. Additionally, only the largest eruptions, those most likely to have had an impact on climate, were included in the analysis. This reduces another source of error, in that the timing of the large eruptions is well established and avoids the complication of years such as 1902, when at least five eruptions of varying size occurred. Only one of these is really significant, that of Santa Maria, Guatemala, and it is the only one considered.

The key finding is a recognizable El Niño-like response in years 1, 2 and 3 after large tropical explosive eruptions, followed by a weaker La Niña-like state in years 4, 5 and 6. Adams et al. conclude that volcanic eruptions do not trigger all El Niño events but that they help to drive the ocean and atmosphere towards a state in which El Niño-like conditions (warming) are favoured. This is a prudent message, because establishing statistical significance above the noise of the uncertainty in reconstructions of past ENSO behaviour, and the timing and impact of volcanic eruptions, is a tall order: a stronger conclusion is not warranted by their statistical case. Support may come from a better understanding of the evolution of volcanic aerosols by monitoring a large eruption in the future, and from improved models of the interaction between the radiative effects of aerosols and sea surface temperatures in the Pacific.

Nonetheless, Adams and colleagues' work implies that volcanic eruptions, such as that of Mount Pinatubo in 1991 (Fig. 1), may have a larger effect on Earth's climate than previously thought. If they influence the ENSO cycle as proposed, then explosive volcanism is a vital catalyst in global climatic interconnections, and a major player in Earth's climate system. This is grist to the scientific mill — let debate ensue.

Figure 1: Blown sky high — a view from the space shuttle Atlantis of volcanic dust, high in the Earth's atmosphere, following the 1991 eruption of Mount Pinatubo (below) in the Philippines.


Three distinct dust layers are visible.


  1. 1

    Philander, S. G. H. El Niño, La Niña and the Southern Oscillation (Academic, San Diego, 1990).

  2. 2

    Adams, J. B., Mann, M. E. & Ammann, C. M. Nature 426, 274–278 (2003).

  3. 3

    Hirono, M. J. Geophys. Res. 93, 5365–5384 (1988).

  4. 4

    Robock, A. Rev. Geophys. 38, 191–219 (2000).

  5. 5

    Handler, P. & Andsager, K. Int. J. Clim. 10, 413–424 (1988).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.