In the first part of this post, we looked at magical carbon and where to find it. We now continue to look at how much CO2 volcanoes produce, and how it compares to our own emissions. Who wins the battle?
The results of the polls are: A small majority believes that volcanoes produce less CO2 than humanity (39% versus 37% for the opposite), with 4% going for similar CO2 levels and 19% is open minded (numbers updated 22 August). The second poll has a clear winner, with 66% expecting that volcanoes outperform human breath in CO2 production. Read below to find the facts!
Let’s first look at humanity’s CO2 emissions from fossil fuels and other sources. The picture shows the amounts we produce per year; the numbers are an average over the period 2006 to 2015.
We produce 34 gigatons of CO2 per year through the burning of fossil fuels. A further 3.5 gigatons is added from changes in the land use, for instance by the growth of cities replacing farm land or forest. On the other hand, 11.5 gigatons is taken up by land elsewhere, including through increased plant growth. About 10 gigatons of the CO2 is taken up by the oceans.
This leaves 16 gigatons of CO2 per year as year-on-year increase in the atmosphere. Since the industrial revolution, we have produced 2000 gigatons of CO2 in total, of which 1200 gigatons remains in the atmosphere, 400 gigatons has gone into the oceans and another 400 into the soil. About a third of the CO2 currently in the atmosphere is man-made.
The increase in the atmosphere fluctuates a bit from year to year. For instance, in 2015 CO2 in the atmosphere increased more than was expected because plants took up less than normal. The high temperatures associated with the strong El Nino decreased global plant growth that year.
Breathing is excluded from the numbers: each person produces about 1 kilogram of CO2 per day, which means all of humanity produces 2.7 gigatons per year. This number adds about 8 per cent to our total CO2 emission, but it is in balance with plant and algae growth – if it weren’t, we would see the amount of oxygen in the atmosphere change! (The claim that rain forests are the main lungs of the world is not correct. They are more or less carbon neutral: the trees grow very fast, but the dead wood also decays very fast.)
Now let’s go back to the topic of this post – volcanoes as source of CO2. Firstly, where do they get their CO2 from? It is a mixture of three different reservoirs. Some of it comes from the mantle, some from subducted ocean floor, and some from the upper levels of the continental crust through which the volcano erupts. A little (very little) carbon comes up as diamond, formed at high pressure and enriching the soil. But most of the carbon comes out as CO2 and is added to the atmosphere.
Through their gas recycling, volcanoes help to keep our atmosphere in balance. It is a slow dance: volcanoes may be a few hundred kilometer downstream from the subduction zone, and at a typical rate of 5 cm per year, this distance may take 10 million years to cover. Recycling of subducted carbon takes time. But it is an essential process. Without the volcanic replenishment, the CO2 in the sea and atmosphere would be gone within half a million years. Life depends on volcanoes, and it could not cope with a volcanic hiatus.
There are rather few direct measurements for CO2 output from volcanoes. The erupted CO2 mixes with the air and becomes heavily diluted. For instance, Etna produces around 15 kiloton of CO2 per day, which you might think is easily measured. But at 1 kilometer from the summit, this only adds 4 ppm to the CO2 already in the atmosphere. So you are measuring a very small increase. Nowadays, CO2 can be measured from space, but it is still a difficult experiment to separate volcanic emission from the normal air. Instead of CO2, SO2 may be measured as this doesn’t suffer this dilution. The amount of volcanic CO2 is than calculated from the SO2.
In this way, we know that the famous 2010 eruption of Eyjafjallajokull produced between 150 and 300 kilotons of CO2 per day. If we assume that the main eruption lasted two weeks (the sputterings went on for longer), the total CO2 output was around 2 million tons. This eruption is best known for its effect on air travel, closing much of European air space. These grounded flights resulted in significant CO2 savings, estimated at 2.8 million tons. This exceeds the volcanic emission, and the offsetting made Eyjafjallajokull the first carbon-neutral eruption!
Bardarbunga did rather better, befitting its size, and emitted 6.5 million tons of CO2 during its six month eruption. This is a bit less than the 11 million tons of SO2 it produced. And this is a bit strange, since in general there is rather more carbon than sulfur. The probable answer to the CO2 deficiency is that the erupted basalt was already partially degassed before it erupted. (CO2 degasses before SO2 does.) The degassing would have occurred while the magma was at around 9 kilometre depth.
Pinatubo was more productive, with some 42 million tons of CO2. It was so high because the magma was completely saturated with volatiles. Mt. St. Helens erupted between 4.8 and 22 million tons. Bardarbunga was indeed low in CO2 for it size, compared to the other eruptions.
The quiet ones
But in fact, most of the volcanic CO2 does not come during eruptions. It comes from quiet degassing of volcanoes, in fumaroles and hydrothermal activity. When you see a plume above a dormant volcano, you are looking at these emissions. Volcanoes don’t need to erupt in order to breath: the magma chambers are continuously exhaling CO2, and the gas can make its way to the surface through tiny cracks, while the magma stays below. The CO2 can come out at the summit, but it may also be spread out over the flanks. Kilauea is an example where the emission is mainly from the summit, while Etna emits a lot from the upper flanks. Only the most volatile gases escape this way, and the quiescent emissions are therefore mainly CO2 and helium. There is little or no acidic component, such as SO2, HCl or HF.
For Japanese volcanoes, measurements by Hiroshi Shinohara in 2013 confirm that volcanic CO2 mainly comes from persistent degassing, and less than 15% comes during explosions. Hydrothermal activity with CO2 emission is found in El Hierro, White Island, and many other places. El Hierro produces around 200 kilotons of CO2 per year, while busily not erupting. Sometimes the CO2 comes out in a sudden burst, as happened in Dieng volcano. In that particular case the burst was triggered by a phreatic eruption, but the CO2 had collected under ground over a long time, and did not come from the eruption.
An example of these diffuse emissions is found at Mammoth Lake in the US. After a magma intrusion, CO2 emissions began to increase in 1989, and by 1990 trees began to die near Horseshoe lake. Measurements showed that the gas underground in places was almost pure CO2. That sounds like paradise for a plant, but in fact tree roots do need oxygen – they don’t get it from the tree above it and they can’t photosynthesize below ground. The degassing of new magma thus killed the vegetation. The mountain calmed down later and the CO2 went back to normal. A failed eruption can still be a CO2 success.
Doing the volcanic sums
To get the total volcanic production of CO2 per year, all these contributions need to be evaluated and added up. The estimates range from 120-200 million tons (USGS) to 300 million tons (the British Geological Survey). But inevitably, these numbers were estimated from little data and they are quite uncertain. Mike Burton and collaborators published a study in 2013 where they tried to account for all volcanic sources of CO2. They combined data from 33 volcanic regions, from which they extrapolated to the rest of the world. And this gave some surprising results.
The table below comes from their paper. It lists the main volcanic polluters in order of CO2 production rates, from among the 33 measured volcanoes (their paper has the full list). The numbers refer to the gas measured in the persistent plumes, while the volcano is otherwise quiescent. Top of the list is Nyiragongo (where it should be noted that this is based on only two, widely different, measurements). Kilauea is high on the list, as is Etna. In fact, the main polluters are those volcanoes that are most frequently active, suggesting that the CO2 production depends on magma supply rate. The only Icelandic volcano on their list is Grimsvotn, at 200 kilotons of CO2 per year. More recent measurements at Hekla find that it emits around 16 kilotons of CO2 per year, of which 5 kilotons come from a small area at the volcano’s summit. This is less than a tenth of Grimsvotn: It is clear who is the carbon ruler of Iceland.
Ambrym is remarkably high on the list. It has a very high magma supply rate, which may be close to 1 km3 per year according to one (high) estimate. The magma reaches the surface in long-lived lava lakes, but does not go further: Ambrym erupts rather little. However, the magma manages to degas very effectively, perhaps because of the lake.
Ol Doinyo Lengai, in Tanzania, 6th on the list of 33, erupts a carbonatite lava, which perhaps gives a hint to the cause of its high CO2 emissions. It has frequent, minor eruptions, and at the time of writing appears to be heading for a new eruption, the first since 2007. There are ancient human artefacts in the area, which in some cases were preserved by the ash from the volcano.
The 33 measured volcanoes together account for 60 megatons per year. Extrapolating from this to all such volcanoes, the authors find a total CO2 production in volcanic plumes of 270 megatons per year.
A second series of measurements concerns the diffuse CO2 emissions, rising from the ground in volcanic and tectonically active regions, including deeply dormant volcanoes. The measured amounts are dominated by Yellowstone, which emits 8.5 megatons per year, and the volcanic belt in Italy which emits 10 megatons per year. For comparison, the Costa Rica–Nicaragua volcanic segment is estimated at 1 megaton per year, while Indonesia and the Philippines together account for 2 megatons per year. Extrapolating to the rest of the world yields an estimate of 200 megatons per year for these diffuse emissions. This could be an underestimate.
Volcanic lakes are a third source of CO2, and they add another 95 megatons per year to the global emissions. This number is rather uncertain, as only 32 such lakes have been measured, whilst 138 such lakes are known to exist and the total number may exceed 700. Very few of those lakes are stratified, and they do not pose as much danger as Lake Nyos. But their CO2 emission still matters.
Finally, the mid-oceanic ridges which extend around the world contribute CO2. The amount of new oceanic crust that they produce is staggering, and far exceeds any volcanic produce on land. But the amount of CO2 is rather less than you might expect. This is because the magma comes directly from the mantle. It doesn’t pick up any CO2 from subduction slabs, or from crustal melt, and therefore the material is quite CO2 poor. The measurements, which were made around black smokers, yield a total of 97 megatons per year for the entire oceanic spreading ridge system.
There is a peculiarity here. The hot rock that surfaces in the mid-oceanic ridge immediately begins to absorb CO2 from the water. The amount that it takes up is estimated as 150 megatons per year, and this more than cancels out the CO2 production.
Sometimes a rift can develop within a continental plate. The East Africa Rift is such a place, and it is in the process of applying for mid-oceanic spreading ridge status. It produces much more CO2 than an equivalent oceanic spreading ridge because it mines the continental crust. A recent paper by Lee et al. (Nature, 2016) finds that the East African rift emits more than 20 megatons per year, a number that is extrapolated from 4 megatons per year measured in the Magadi–Natron Basin.
Adding all the various contributions together, Burton and collaborators calculate that the Earth’s volcanoes emit a total of 637 megatons of CO2 per year. This is more than double the older estimates.
Volcanoes versus humanity – the verdict
So how do our emissions compare to those of the volcanic eco system? Was Plimer right in discounting humanity? Let’s start with Mike Burton’s number, which seems the best established value: volcanoes, in all forms, sizes and states of activity, add over 600 megatons of CO2 to the air (and water) each year. There are some uncertainties and perhaps some major emitters were missed, so we can increase their estimate to 1 gigaton per year and still be within the bounds of uncertainty. Of course, the Earth is used to these emissions, and it manages to remove the same amount by its various mechanisms. Because it is dominant by quiescent volcanoes, the number will not vary much from year to year.
Between 2006 and 2015, humanity added on average 34 gigaton of CO2 per year through fossil fuel burning. Some of this went into the ground and the sea, leaving 16 gigaton of CO2 per year added to the atmosphere as humanity’s net imbalance.
So humanity produces 34 times as much as all volcanoes combined through its emissions. It even produces 3 times as much CO2 as volcanoes do just through breathing! This answers the two poll questions from part 1 of this post. The factors become as high as 50 and 4 when using the numbers from Mike Burton. And volcanic eruptions contribute even less. Pinatubo, the largest eruption over the last 100 years, contributed less than 0.3 per cent of the annual increase in CO2 content of the atmosphere.
In conclusion, Mike Huckabee and Ian Plimer were not just a little wrong: they were was massively wrong. Mike Huckabee claimed that Eyjafjallajökull produced as much CO2 as humanity in 100 years. The correct number is 30 minutes. Huckabee was therefore wrong by a factor of 1.7 million, a sad indictment of the level of numeracy required for government. Ian Plimer stated that one volcanic cough equalled 250 years of human CO2 emissions. He doesn’t define what a ‘cough’ is, but presumably it is less than one Eyjafjallajökull in which case he was wrong by a factor of at least 4 million, and possibly much more. ‘Astronomical’ doesn’t begin to describe it. Such extreme arithmetic errors can be career-defining for a politician but would be career-ending for an engineer. We can safely discount their numbers as silly.
But how about some very large eruption? Could a one-off event compete with us? Take Long Valley Caldera, which went off 700,000 years ago and generated 100 km3 of tephra. This was (very) roughly 1000 times larger than Eyjafjallajökull, so we can guess that its CO2 emissions would have been around 2 gigatons. We could get even bigger eruptions, but even a ten times larger one, which might happen say once in ten million years, would produce less CO2 than we do each year. So the answer is no.
(Of course, if there were such an eruption we would have other things to worry about. Even sulphate would be a more urgent killer.)
Origin of the myth
So where did these outrageously wrong numbers come from? Both people wouldn’t have done the calculations themselves: they quoted from somewhere else. It turns out, there were several steps on their path from science to silliness.
The story began with a paper by David Johnson, published just after he died in the St Helens eruption. The paper discussed the chlorine emission from Augustine volcano, Alaska, 1976 and argued that of the 525 million kg of chlorine it produced (it erupted 0.19 km3 DRE), roughly 100 million kg could have reached the stratosphere in the form of HCl: such a stratospheric contribution would be equivalent to a quarter of the 1975 world industrial production of chlorine in fluorocarbons. Read that sentence twice to see what he actually said, and remember that at this time, fluorocarbons were just about to become a major concern because of the destruction of the ozone layer. Johnson argued that explosive eruptions could inject enough chlorine into the stratosphere to be a significant factor.
He also wrote that if the Bishop Tuff of Long Valley Caldera emitted the same fraction of chlorine, its stratospheric contribution would account for 570 times the 1975 world industrial production of chlorine in fluorocarbons. This number was about a 100 km3 eruption 700,000 years ago. In hindsight, the paper was correct about the total amount of chlorine, but overestimated how much of it entered the stratosphere. The explosions also eject water, and the HCl dissolves into this water and rains out before it gets to the stratosphere. But that is a different story.
Dixie Lee Ray, in 1990, wrote that Augustine volcano put out 570 times the total world production of chlorine and fluorocarbon compounds in the year 1975. He quoted Johnson but mixed up the two volcanoes and ended up overstating the Johnson number by a factor of 2000. These things happen. Worse, he also used the number for the stratosphere as if it were for the entire atmosphere.
The next year, Pinatubo exploded and the discussions re-ignited. This was just at a time when industries were trying to avoid restrictions on CFC production. In their defence, Rush Limbaugh stated in 1992: “Mount Pinatubo has put 570 times the amount of chlorine into the atmosphere in one eruption than all of man-made chlorofluorocarbons in one year”, again mixing up volcanoes, and also confusing the atmosphere and the stratosphere. (In fact, measurements did not show any significant increase of stratospheric chlorine after the eruption.) In Limbaugh’s book of 1993, he amended the statement to: “Mount Pinatubo in the Philippines spewed forth more than a thousand times the amount of ozone-depleting chemical in one eruption than all the fluorocarbons manufactured by wicked, diabolical, and insensitive corporations in history.” He changed a single year (1975) to all of history, and increased ‘570’ to ‘more than a thousand’. Note that he wrote in the context of defending the chemical industry, and note his use of hyperbole.
After this, it took on a life of its own. In 2006, Christopher Monckton said “In a good year for eruptions, Erebus can put out as much CFCs as Man used to.” This is another change of volcano, and again there is confusion with Long Valley, but in addition HCl has now become CFCs, something volcanoes do not produce. This confusion was clearly based on using Limbaugh as his primary source, as Limbaugh hadn’t specified which ozone-depleting chemical was meant – Monckton substituted the only one he knew about.
Remember the mentions of fluorocarbons? Jude Wanski, in 2004, wrote “The eruption of Mt. St. Helens in 1980 dumped more greenhouse gases into the atmosphere than all that has been released since the industrial revolution”. Long Valley has now morphed into St Helens, HCl has become ‘greenhouse gases’, and the stratosphere is again equated with the atmosphere. The greenhouse confusion came about because Monckton had changed HCl to CFCs, and these do act as powerful greenhouse gases.
And this led directly to Ian Plimer, who changed ‘greenhouse gases’ to ‘CO2’ (unaware there were others), and elsewhere in his book writes that Mt. Pinatubo released “very large quantities of chlorofluorocarbons”. He gives a reference to a 1992 paper, but this paper contradicts his writings and he cannot have read it. His information clearly came from Wanski. Mike Huckabee used Plimer as source, but substituted Eyjafjallajökull for Pinatubo (which through several previous substitutions was actually Long Valley) , CO2 for HCl, and 100 years for one year. (An interesting aside is that Eyjafjallajökull and Augustine had the same eruption volume in DRE, so replacing one with the other could have been defended, had it been that simple.)
Regardless how you feel about global heating, you probably do not want to use a chain of chinese whispers as your primary source of information. It is important to go back to the original source, preferably peer reviewed. The strange story here shows how important it is to use authorative, reliable sources for your factual information.
The final score
And the final CO2 score is
Humanity: 34; Volcanoes: 1
Albert Zijlstra, August 2017
Sources used (these may be paywalled and inaccessible to the public, in which case apologies! Scientists would like people to read their work – but some publishers seem to prefer if people didn’t.)
Massive and prolonged deep carbon emissions associated with continental rifting. Lee et al. http://www.nature.com/ngeo/journal/v9/n2/abs/ngeo2622.html
The Volcano Gambit, Gavin Schmidt http://www.realclimate.org/index.php/archives/2016/04/the-volcano-gambit/
Deep Carbon Emissions from Volcanoes. Michael Burton et al. http://rimg.geoscienceworld.org/content/75/1/323
Massing life, Philip Hunter https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2897122/
Environmental pressure from the 2014–15 eruption of Bárðarbunga volcano, Iceland. Gíslason et al. 2015, https://www.geochemicalperspectivesletters.org/documents/GPL1509_noSI.pdf
Staged storage and magma convection at Ambrym volcano, Vanuatu, Fionnuala Sheehan and Jenni Barclay http://www.sciencedirect.com/science/article/pii/S0377027316300117
Volcanic Contributions to the Global Carbon Cycle. Vicky Hards, British Geological Survey. https://www.bgs.ac.uk/downloads/start.cfm?id=432
Significant discharge of CO2 from hydrothermalism associated with the submarine volcano of El Hierro Island. J. M. Santana-Casiano et al. https://www.nature.com/articles/srep25686
Degassing regime of Hekla volcano 2012-2013. Evgenia Ilyinskaya et al. http://www.sciencedirect.com/science/article/pii/S0016703715000241
The carbon cycle
CO2 in the atmosphere can dissolve into (rain) water, and react with silicate rock:
CaSiO3(s) (wollastonite) + 2CO2(g) + H2O(l) → Ca2+(aq) + 2HCO−3 (aq) (bicarbonate) + SiO2(aq)
where (s) means solid, (aq) means dissolved into liquid water, and (l) is that liquid. This is followed by
Ca2+ (aq) + 2HCO3−(aq) → CaCO3(s) + CO2(g) + H2O(l)
These two reactions remove one molecule of CO2, and produce one molecule of CaCO3 which precipitates on to the ocean floor. Subduction carries the calcium-carbonate down to the mantle, safely out of the way. If it came down on the continental shelf, it will escape subduction but may eventually be lifted above sea level. Thus the white cliffs of Dover came to be. Remember the dusty white chalk at school by which knowledge was transferred from the teacher to the black board?
This carbon cycle goes back a long way. Originally, the Earth had an atmosphere similar to Venus, with a pressure close to 100 times the current, almost entirely consisting of CO2. As soon as the oceans formed, the CO2 began to disappear into it, and all but a minor fraction of the original CO2 ended up in the mantle. Going through the numbers, there may be 700 million gigatons of CO2 (equivalent, as it isn’t in the form of CO2) safely sequestered into the Earth’s mantle. (If this sounds like a lot, it is less than 0.02 per cent of the mass of the mantle. For comparison, CO2 accounts for 0.06 per cent of the mass of the atmosphere.)
Mantle material comes back to the surface at mid-oceanic spreading rifts and at hot spots. A little of the carbon in the mantle is thus returned each year.
The carbon cycle is not always in equilibrium. An example is the Siberian Trap eruptions, which enormously increased CO2 (and possibly methane) in the air, much faster than it could be removed. It caused 99% of life to die of heat stroke. The equilibrium of the atmosphere can be punctuated, and it takes a while (thousands to millions of years) to regain stability. The present year-on-year increase of the CO2 in the atmosphere shows that the cycle is not in equilibrium at the moment.
The human CO2 budget
Now let’s look at all the incomings and outgoings which affected the atmosphere since the industrial revolution. The plot below shows the emissions from different sources, versus the amounts taken up by the land and ocean.
The ocean has taken up about half of the CO2 produced by burning fossil fuels. And about a quarter of our CO2 emissions have come not from fossil fuels, but from land use. However, plant growth has more than compensated for that. After subtracting the various sinks, the net effect is that CO2 in the atmosphere has increased from 288 ppm to 400 ppm.
Update: Conveniently, a new paper on CO2 from mantle plumes has just been published, giving a 40% higher CO2 emission estimate but in fair agreement with my estimates above. Their number for the carbon content of the mantle seems close to my number.