Interestingly, Volcanoes can Trigger El Ninos

Once again, I’m starting this post as I’m halfway through another one that may or may not see the light of day. I was trying to put together a rant following the Royal Society’s (RS’s) panel discussion on geo-engineering (available on, which I attended on Tuesday evening. The meeting followed a report issued by the RS last September.

Rather than rule out most of the possibilities, the RS boffins recommend further research. A cynic might suggest this was self-serving; I couldn’t possibly comment.

There are numerous problems with many of the geo-engineering approaches. But I wanted to be original and see if I could find evidence to support my hypothesis (noted in a previous post) that trying to cool the planet by injecting sulphur dioxide (SO2) into the stratosphere where it would produce reflective particles would block a disproportionate amount of sunlight striking the atmosphere at a shallow angle (i.e. more tangentially). More sunlight would therefore be blocked at the ends of the day, high latitudes and in winter.

In fact, the boffins noted on Tuesday that a disproportionate effect in the Arctic could be “beneficial”. This doesn’t stop them apparently relying on a computer modelling study that simply plugs in “a reduction in global mean insolation of 1.84%”.

The RS study repeatedly discusses recent volcanic events in order to assess possible effects of the geo-engineering plan.

Having looked into the matter, I can say this is bordering on a waste of time.

Drs Strangelove want to fire enough SO2 into the sky to block out around 2% (on average) of the sunlight, their sums suggesting this would counterbalance a doubling of CO2 levels.

But this interesting graph (courtesy of WIkipedia) shows what volcanoes do:

Mauno Loa observations of atmospheric transmission of sunlight

Wow! They don’t block 2% of solar radiation, rather 10 to 20% on a regular basis, and presumably even more when a real biggie goes off.

And this is enough to cause real disruption.

First off, the boffins worry about affecting the monsoon and other aspects of the hydrological cycle, citing the effects of the Pinatubo eruption in 1991. But Pinatubo caused massive short-term cooling. Monsoons rely on the land becoming warmer than the oceans, leading to rising air, drawing moist air towards the landmass. Obviously, if you reduce sunlight by 10% or so, the land will warm much slower and could remain too cold, relative to the ocean (which is kept warm by stored heat), for a healthy monsoon.

Second, I noted in a comment on a previous post that El Chichon was followed by a strong El Nino. As can be seen from the graph I gave at the time, there was a weaker one after Pinatubo. “Could they possibly fit together?”, I found myself wondering. Yesterday, via Realclimate, I came across a paper suggesting that yes, indeed they could (pdf).

The point, of course, is that El Ninos occur when warm surface water flows (unusually) east across the Pacific (see also Wikipedia). The warm water builds up in the first place because the initial flow (ultimately due to the rotation of the Earth) creates, in turn, an atmospheric warm zone to the west (around Indonesia), and a cooler region near South America. Lower pressure maintains a significant difference in the surface level between the west and the east of the ocean (maybe 60cm!). But the feedback relies on maintenance of the temperature (and hence atmospheric pressure) differential and eventually breaks down, typically in December (when the sun is not overhead at the Equator) and the whole thing collapses like a lop-sided souffle in an unevenly heated oven. Warm waters temporarily flow east with significant effects on the global climate for a year or two.

Fairly obviously, a general cooling event, such as a volcanic eruption, is likely to trigger an El Nino.

As an aside, it might be worth noting that a warming period is likely to lead to a strong El Nino, as observed in 1997-8, for example. The warming will reinforce the feedback creating the original imbalance. 1982-3 was also a strong El Nino event:

El Chichon anomaly (1983-4 temperatures compared to 1980s)

Maybe the cooling caused by the 1982 eruption of El Chichon enhanced an El Nino that was anyway ready to take place. Or maybe it was always going to be a big one.

At the risk of trying to read too much into limited evidence, it might be possible to surmise that the 1997-8 El Nino was so strong because the global warming trend leading up to it was reinforced by recovery from the Pinatubo cooling event. Similarly, to stretch the point even further, 1972-3 is listed as a strong El Nino, and followed the recovery after the Agung eruption, though that was 10 years before and not so large as El Chichon and Pinatubo (though a graph over at SkepticalScience gives a different impression). Fascinating stuff – no wonder climate scientists can’t wait for another major eruption!

Incidentally, because we mainly measure the temperature at the surface of the planet, El Ninos show as spikes in the data, because warm surface water covers cooler layers over a large area of ocean (and in turn affects temperatures on land). When a volcano triggers an El Nino, the cooling caused by the volcano is therefore partially obscured by the El Nino. An eruption when we were already in an El Nino state would consequently likely appear to have a greater effect on global temperatures than one that triggered an El Nino.

The geo-engineering plan is entirely different to the case of intermittent volcanic forcings. The plan involves a semi-permanent sunscreen to block 2% of sunlight. The problems will be entirely different. Relying on the historical record of the effects of volcanic eruptions won’t allow us to predict all the effects of the geo-engineering proposal.

Logic tells me that an SO2 sunscreen will disproportionately affect high latitudes, where sunlight is a highly valued commodity. Politically, it would of course be next to impossible to achieve broad agreement to go ahead with the geo-engineering plan. Furthermore, in a warmer world, with increased tropical desertification, we may be relying on food production in more northerly areas. Blocking sunlight might not be a bright idea.

Nevertheless, I carried on surfing for a bit for evidence that volcanic forcings could affect high latitudes more. The best I could come up with was the Russian famine of 1601-3, likely triggered by an eruption in Peru.