by Jonathan Gregory

Explosive volcanic eruptions can have a cooling effect on the climate for a few years, and this reduces the warming and sea level rise caused by increasing concentrations of greenhouse gases. Climate model results appear to show that the cooling caused by the eruption of Krakatoa in 1883 could have had an effect which lingered through much of the twentieth century. Recent work that I've done suggests a partial reinterpretation of these results.

Explosive volcanoes can produce a haze of particles which lasts for a few years in the stratosphere. These particles reflect sunlight and have a cooling effect on the climate. Such an effect tends to cool down the ocean too, and to make sea level fall, because sea water (like most substances) contracts when it cools. Computer climate models have been widely used to simulate the past 150 years, including the cooling effect of historical volcanic eruptions, like those of Krakatoa in 1883 and Pinatubo in 1991. The models also include the warming effect of increasing concentrations of carbon dioxide and other greenhouse gases in the atmosphere during the same period. In the climate models, the greenhouse warming outweighs the volcanic cooling, so there is a rise in global temperature and sea level, as has occurred in the real world.

A number of papers published around 2006 pointed out that the climate models appear to show that the ocean warms up only very slowly after each large volcanic cooling, so that the eruption of Krakatoa, for instance, might have had an effect which persisted through much of the twentieth century. I've recently published a paper which suggests a partial reinterpretation of the earlier model results. Since I was an author of some of the previous papers, I am partly disagreeing with what I wrote before, but that is OK; it's normal in scientific research to refine and revise our understanding as our work proceeds.

There's no doubt that volcanoes have mitigated climatic warming and sea level rise, but I think now that the ocean warms up faster after each eruption than we thought before. If my new paper is correct - which we will find out if other researchers try similar experiments - the important point is whether there have been more than the average number of large volcanic eruptions, or fewer than average, in the decades before the time of interest. The paper explains how and why this atters. As it happens, there have been eruptions more frequently than usual in recent decades, and sea level has probably risen less during this period than it would have done if there had been only the average number of large eruptions. But the signature of Krakatoa may not have lasted far into the twentieth century after all.

If another volcano of the magnitude of Krakatoa were to erupt today, we would be far better equipped than ever before to observe what changes take place in the ocean and the climate system generally, and would be able to get answers to several important scientific questions.

pseudovol xs blogx

This diagram shows global-mean temperature change in the ocean at various depths (marked on the vertical axis) beneath the surface. This is the result of a simulation I did for my recent paper using a climate model called FAMOUS, which we maintain in NCAS-Climate for use by the UK climate science community. Although the diagram shows historical dates (on the horizontal axis) and the four volcanoes marked erupted at the dates indicated, this isn't a realistic simulation because the magnitudes of the volcanic eruptions and the changes in greenhouse gas concentrations have been modified in order to investigate their effects more straightforwardly. The diagram shows warming spreading downward from the surface and increasing as time passes, due to the growing greenhouse effect. Superimposed on this are short-lived but strong coolings due to large volcanic eruptions. These cool the surface and spread downwards, and their effects persist for a decade or two following the eruption.