Volcanoes And The ‘Pause’
Another week and another explanation for the ‘pause’ in global surface/lower atmosphere temperatures. This time it’s the return of the ‘small volcanoes add up to big effect’ explanation in the form of a paper by Santer et al 2014 in Nature Geoscience. Could the cumulative effect from small volcanoes be causing a reduction in sunlight reaching the Earth’s surface, and hence a reduction in the rate of global surface warming?
Now you would have thought after reading the Guardian for example, that this was a clear result and they quote co-author of the study Carl Mears, “We were able to show that part of the cause of the recent lack of temperature increase is the large number of minor volcanic eruptions during the last 15 years.” A closer look at the paper shows it to be fascinating, though not as clear as the Guardian and others report. This is after all messy climate science.
First of all consider this graph from a recent presentation. It shows the optical depth or transparency of the atmosphere. The effects of El Chichon and Pinatubo volcanic aerosols are clear in reflecting sunlight and reducing the Earth’s temperature. But look at the post-2000 period. It is remarkably flat. So flat in fact that many scientists noticing its lack of interference from volcanic effects are describing it as a “background” event that shows a base-level of atmospheric opacity – a reference level from which to measure other variations. Click on image to enlarge.
Santer et al 2014 point out that there have been 17 small eruptions since 1999, over 50% in the tropics (where their global influence is thought to be greater than high-latitude ones). To look at their influence on lower atmospheric temperature they take the lower atmospheric temperature and remove ENSO and the effects of the El Chichon and Pinatubo eruptions. They say this reduces the internal noise allowing the small volcanic effects to be better studied.
Their Fig 1 shows raw lower temperature data (a), that with the El Nino removed (b) and that with El Nino and El Chichon and Pinatubo removed (c). Looking at 1c one sees that the lower atmosphere shows a standstill since 1993, that is 20 years! This is in itself a remarkable graph extending the ‘pause’ into the start of its third decade. Click on image to enlarge.
They say the volcanic effects correlate with optical depth and top of atmosphere. Consider their Fig 2a, changes in observed Stratospheric Aerosol Optical Depth (SAOD) between 2000 and 2012 along with the various volcanic eruptions seen in that decade. Now I know one can draw a straight line through almost any set of data but I wonder if a straight line is a good representation of the data? Likewise their Fig 2b the clear-sky short-wave radiation at the top of the atmosphere. Sure there was a lower region around 2002 but does its presence really justify a decadal rising trend? I am also doubtful about any correlation between the SAOD and the top of atmosphere data, with the El Nino removed atmospheric temperature data.
Looking at one of the paper’s references for the stated change in atmospheric optical depth one finds this. Does the globally measured SAOD really show a 4-7% annual increase between 2000-2009? Sure it is changing, but a linear trend? I think that the trend in global stratospheric aerosols must be speculative.
Overall this paper shows that volcanoes are having a minor to marginal effect, in many scenarios about the same effect as the noise in the global temperature dataset for the past 15 years. It concludes that the ratio of simulations of global surface temperatures to the actual observations is about 15% smaller, even though there are large uncertainties in the magnitude of the effect!
It is formally true that including volcanoes means the models are more able to match the observed temperature but not that usefully. There are so many other factors that have to be taken into account that have uncertain effects and that are adding to the inaccuracy of the models in explaining the ‘pause’. Despite this the Guardian says the models work, period.