Figure 1: Comparison of the oxygen isotope record from ODP Site 677 (proxy for global ice volume) with timing of large volcanic eruptions.
Volcanoes and Climate, part 2.
Guest Lecturer: Ellen
Thomas (Office 459, extension 2238;
ethomas@wesleyan.edu)
Date: 17 November 2004
Note: This is the second of two lectures on ‘Volcanoes
and Climate’. This lecture will deal will with speculations on
very large-scale climate and environmental effects of volcanic
eruptions in the geological past (severe climate change, ice ages,
mass extinctions), after the first
lecture (Monday November 15) dealt
with historical eruptions and their observed global climate
effects.
Readings for this lecture:
Additional on-line information:
Main Points of today’s lecture:
How catastrophic could large eruptions have been?
Volcanoes emit abundant CO2 and SO2. CO2 is a greenhouse gas, thus could cause global warming, whereas SO2 reacts to an aerosol and induces tropospheric cooling. Because CO2 is a gas and does not react to form aerosols, it has a longer atmospheric residence time than SO2, and possible greenhouse effects thus should be occurring on a longer time scale than SO2-induced cooling. Atmospheric residence times of CO2 are difficult to estimate because of the complexity of the carbon cycle, but values given are in the decades to centuries range.
What evidence do we have for either catastrophic cooling or catastrophic warming induced by very large volcanic eruptions? We’ll look for a linkage between large volcanic eruptions and major global climate change (i.e., beginning of ice ages of major episodes of glaciation) or mass extinctions.
Mass Extinctions:
During a mass extinction many unrelated species living in many different habitats (land and ocean) become extinct within a geologically short period (practically, within about 10 kyr or so). For a mass extinction to occur, environmental change must have been so severe that organisms cannot survive it, so global and/or rapid that many (very different) organisms can not migrate or evolve, and there are no refugia. This calls for severe conditions: there were no very high rates of extinction during the alternation between glacial and interglacial episodes during the Plio-Pleistocene ice ages.
Five mass extinctions are seen as most significant; note that estimates of % of families and species becoming extinct are not very reliable or precise.
Note that there is evidence for flood basalts as well as impacts during several of these extinctions.
Catastrophic global cooling:
In general, explosive volcanoes (and specifically, large ignimbritic eruptions) have been proposed to cause cooling (VEI = 7 or 8), because their plumes easily penetrate into the stratosphere and thus can deliver large amounts of SO2. Sulfate degassed from flood basalts, however, could possibly reach the stratosphere in convective plumes, reaching heights of 3-6 km above fire fountains, 8-11 km above fissures, and even higher during active eruptions. Thordarson and Self (1996) estimated that a total of 9,000 Mt of SO2, 1300 Mt of HF, and 400 Mt of HCl were released by the Roza flow in the Columbia River Plateau,which took about 10 years to flow out (compare to ~ 20 Mt SO2 from the Pinatubo eruption; ~120 Mt for the Laki eruption, 1883-1884). The Laki eruption was, however, only about 14km3 as compared to ~1300 km3 of the Roza flow. Note that the Laki eruption was followed by a rather unusual mix of cold and warm weather, and Laki’s climate effects are not well known or understood, mainly because of a lack of quantitative data collected at the time of eruption). There is considerable disagreement in various modeling efforts (click here for a website where you can download 'Atmospheric Impact of the 1783-1784 Laki eruption: Part 1 Chemistry Modeling').
Flood basalts, however, occur in very large igneous provinces, so we have evidence for many voluminous outflows over longer times (a few million years) than in the case of large ignimbritic explosions (e.g., large explosions in Yellowstone have been dated at 2.1, 1.3 and 0.64 Ma).
Ever larger eruptions, however, do not produce ever larger climatic effects in a simple linear fashion. If aerosols become very dense, the particles collide and will cluster together, thus reaching a particle size that will rapidly fall out. The size of an eruptions that is so large that an increase in size will no longer lead to increased effect on climate is not well known.
The Toba eruption (~74 kyr is the most recently given age), and the 3 eruptions of Yellowstone mentioned above have been said to have triggered ice ages (i.e., glacial episodes). The evidence is in my opinion rather poor (see figure 1): some eruptions (e.g., Toba) indeed occur at a transition into a glacial, but this glacial occurred at a time exactly predicted by orbital calculations. The three Yellowstone eruptions show an even less convincing pattern. Possibly, more precise ages for the eruptions would make the correlation more convincing.
On longer time scales, the initiation of the East Antarctic Ice sheet (~33.5 Ma, earliest Oligocene) and the later increase in size of that ice sheet (~14.5 Ma, middle Miocene) have been linked to volcanism. The first has been linked to explosive volcanism in Mexico and the SW USA: it has been estimated that about 2/3 of the ignimbrites in that region (total estimated at ~ 4x105 km3 of silicic ignimbrite) erupted during hundreds of individual flows (102-103 km3) at 38-28 Ma. There is no evidence for a large pulse of volcanism just before the major glaciation at 33.5 Ma. Alternatively, this glaciation has been tentatively linked to the Ethiopian Flood basalts, but this episode appears to be too young (29.5-31.0 Ma). The middle Miocene increase in Antarctic glaciation has sometimes been linked to the flood basalts of the Columbia River Plateau, but most flows formed between 17-15 Ma, a bit early for ice cap growth at 14.5 Ma.
Flood basalts, warming, mass extinctions:
Much of the debate about flood basalt eruptions as a cause of mass extinctions is derived from one version or another of a figure in which the age of larger and smaller mass extinctions (and Oceanic Anoxic Events in the Jurassic and Cretaceous) is plotted versus the age of LIPs (see figure). Note that the largest mass extinctions (Devonian, end Permian, end Triassic, end Cretaceous) all occurred during flood basalt activity. But some are also coeval with meteorite impacts, certainly the end Cretaceous extinction, and possibly the ones at the end Permian, end Triassic, and late Devonian. It has been argued that the combination of flood basalts and impacts can not be coincidence.
How could flood basalt eruption cause mass extinction? The duration of flood basalts episodes is usually estimated at a few millions of years, so it has been questioned whether any effects be fast enough to cause a mass extinction? Would the eruption of one flow, however large, be enough to cause a mass extinction? The massive emission of various gases (including SO2, HF, HCl) would cause major extinction on land (compare loss of lifestock on Iceland as a result of the Laki eruption), at least in the proximity of the lava flows, but how would oceanic biota be affected? Many indirect mechanisms have been proposed.
Figure 1: Comparison of the oxygen isotope
record from ODP Site 677 (proxy for global ice volume) with timing of
large volcanic eruptions.
Figure 2; from Courtillot & Renne, 2004 in press, On the
ages of flood basalts, C. R. Geoscience, doi:
10.1016/S1631-0713(03)00006-3