Arguments
The long hot tail of global warming - new thinking on the Eocene greenhouse climate
Posted on 9 October 2014 by howardlee
Past climate changes like the Eocene Hyperthermals left many traces in the geological record. These tell scientists a great deal about what the Earth looked like in these hothouse eras, the changes they made to rainfall, drought, landscape, oceans, ecosystems and life. Ultimately those records contain clues to the causes of the climate changes, and are signposts to the effects we can expect from modern climate change.
A trio of new studies show that the Eocene Hyperthermals were the result of, not the cause of, global warming in the Eocene. This refocuses attention on abrupt global warming episodes like the PETM, and their role in converting the cooling Paleocene climate into the long-lived Eocene hothouse.
Modern climate change is even more abrupt, and is likely to have a similarly long, hot tail.
Hot flashes
The Paleocene and Eocene Hyperthermals were numerous recurring periods of hot global climate (a bit like human “hot flashes”) between around 59 and 34 million years ago. They were variations on a climate that was generally some 15°C (27°F) warmer than today, when the poles were free of ice even in winter and sea levels were about as high as they have ever been. The land was largely covered in jungle and even polar areas were lushly vegetated.
A trio of recent studies together represent a tipping-point in our understanding of this fluctuating greenhouse world. Littler et al present a detailed 8 million year record from a borehole drilled the South Atlantic, and compare that with another borehole in the Pacific. Turner et al integrate a 4.25 million year record from the equatorial Atlantic with published data to create a 10 million year record in the Eocene. Smith et al studied 1.8 million years of terrestrial sediments in Wyoming, cross-referenced to the same Atlantic marine sediment record as used by Turner et al. The focus of all 3 studies was the links between the Paleocene-Eocene climate and variations in solar energy distribution (insolation) caused by Earth’s orbital wobbles.
Paleocene and Early Eocene Hyperthermals recorded in ocean sediments. Compilation of Atlantic and Pacific data from Littler et al EPSL 2014, © Elsevier, with annotation added for this article.
Cooling Paleocene
The data show that the Earth’s climate was on a long cooling trend in the Paleocene, associated with removal of CO2 from the atmosphere and burying it in sediments, deposits of peat, permafrost and methane hydrates. Those carbon reserves were like deposits in a bank account – available for withdrawal later. The first signs of warming began around 58.9 million years ago associated with a sharp jump in ocean bottom water temperatures (known as the “ELPE” - Early Late Paleocene Event), followed 1.2 million years later by a more steady warming trend that underpinned regular cycles of hotter and cooler climate (those hot flashes that we call Hyperthermals).
But what could cause those fluctuations?
Orbital wobbles
The clue is in the regularity of the fluctuations.
Scientists combined radiometric dates on volcanic ash layers with the record of Earth’s magnetic reversals and the known cycles in Earth’s orbit, to correlate the sedimentary records at the different locations. This also allowed a mathematical assessment of the correlation between the climate cycles recorded in sediments and Earth’s orbital cycles. There are 3 different kinds of orbital cycle that control the intensity of sunlight on Earth (solar insolation): eccentricity, precession, and obliquity – for more about them see this post.
The 3 papers show that the Hyperthermals and their effects were global, and were paced by orbital eccentricity on cycles of 405,000 and 100,000 years, and also by orbital precession (21,000 year cycles). The longer eccentricity cycles are associated with ocean bottom water temperature swings of up to 2-4°C (4-7°F), while the 100,000 year cycles correspond with 1.5°C (3°F) swings in bottom temperature. There’s very little contribution from the obliquity (41,000 year cycle) probably because there was very little ice at the poles at the time (unlike during the Pleistocene ice age, when obliquity was much more prevalent).
In the oceans these cycles show up as regular variations in carbon isotope ratios (reflecting carbon cycle changes) and oxygen isotope variations (reflecting ocean temperature) and the proportion of “Coarse Fraction” or Iron intensity in sediments (reflecting carbonate dissolution and sediment supply).
Fluctuating landscapes
On land, times of high rainfall with high lake levels and increased vegetation cover alternated with times when lakes diminished to salt pans and mudflats with loss of vegetation cover, on the 21,000 year precession pacing. In between those was a middle-ground climate with intermediate lake levels and vegetation cover, dominated by rivers. Every 100,000 and 405,000 years, orbital eccentricity suppressed the regular 21,000 year cycle, and the climate got “stuck” in the middle, river-dominated mode.
Orbital cycles of climate and landscape in Wyoming during the Eocene. Simplified from Smith et al EPSL 2014. Currently it isn’t clear if either A or C coincide with insolation maxima: salt pans may have coincided with the higher evaporation associated with summer insolation maxima, or alternatively enhancement of the monsoon over the western US due to insolation maxima could have led to high lake levels.
New thinking
Previous studies suggested that the strength of the Hyperthermals diminished over time and their frequency increased, implying that pulsed releases of carbon from methane hydrates or permafrost kept the climate hot throughout the Eocene. But Turner et al refute this, showing that the Hyperthermal beat continued long after reserves of methane hydrates and permafrost must have been emptied as the Eocene hothouse persisted through the extended “Early Eocene Climate Optimum” (53-50 million years ago). They also find that the:
“...carbon cycle processes behind these events, excluding the largest event, the Palaeocene–Eocene Thermal Maximum (about 56 million years ago), were not exceptional.”
They show that the regular orbitally-paced heartbeat of Hyperthermals continued over the whole 10 million years of their study, just like they did through many other eras of Earth’s past, including in the Devonian, Carboniferous to Permian, Cretaceous, Oligocene, Miocene and the Pleistocene ice age. In other words, orbitally-paced climate oscillations are not a uniquely Eocene phenomenon, so require an explanation not unique to the Eocene.
Oceans - a giant carbon reservoir
The jury is still out on the exact combination of changes in carbon storage drive these cycles (marine biological pump, carbon burial on land, terrestrial weathering), but it appears that the oceans are, once again, key to this.
The oscillations observed are faster than carbon can be exhumed or buried in the global sedimentary reservoir, so a faster-responding carbon source must have been in play that was able to equilibrate with the atmosphere on millennial timescales – the oceans. This reinforces earlier ideas which indicated that the Hyperthermals were accompanied by repeated, large-scale releases of dissolved organic carbon from the ocean by ventilation (strengthened oxidation) of the ocean interior. Orbital wobbles seem to modulate deep ocean acidity as well as the production and burial of global biomass, and the reason the 405,000-year eccentricity cycle is so prominent is because of the centuries-long residence time of carbon in the oceans. Temperature-enhanced metabolic processes and remineralization of organic carbon in deep ocean sediments, operating on timescales of tens of thousands of years, would also modulate the release of carbon into the oceans.
This new thinking is supported by Smith et al, who concluded that the regular carbon releases in the Eocene Hyperthermals:
“...were the effect rather than the cause of global paleoclimatic and geomorphic changes during the EECO.” (my emphasis)
Littler et al’s study would seem to concur. They found that ocean bottom temperatures lead the carbon cycle by around 6 millennia for the 405,000 year cycles (less in the Paleocene, more in the mid Eocene) and about 3,000 years for the 100,000 year cycles. This indicates that the ocean carbon signal in orbitally-controlled climate change is a feedback response to orbitally-driven temperatures.
In contrast, today’s climate change is not orbitally-driven – if it were we would be experiencing global cooling, not warming. It has also taken place in a time frame just 3% of even the fastest orbital cycle.
So if the Hyperthermals are the normal “heartbeat” of the planet and not the primary cause of the hot Eocene climate – what was?
Volcanic CO2 pushes
Superimposed on the background ‘hum’ of regular orbital cycles are several distinct spikes of strong climate change, too big or with too long a recovery time to be forced purely by orbital wobbles.
The earliest of these spikes is known as “ELPE” (Early Late Paleocene Event – 58.9 million years ago), which marks the end of the long term cooling trend in the Paleocene. It corresponds with a 4°C jump in ocean bottom water temperatures, swings in carbon isotope values, and ocean chemistry changes suggesting acidification. Intriguingly the ELPE coincides with a particularly violent phase of North Atlantic volcanic activity, resulting in the deposition of ash deposits like the Kettla Tuff in the Faroe-Shetland area, and contemporary volcanic ash deposits in western, and eastern Scotland, and igneous activity in Greenland and offshore Ireland.
Another sharp swing in ocean chemistry, carbon isotopes and abrupt ocean warming occurred at 57.7 Million years ago, known as the “Peak Paleocene Carbon Isotope Maximum” or “Peak PCIM.” Littler et al attribute the long term warming trend that began then and continued into the Eocene, to sustained carbon emissions resulting from North Atlantic Igneous Province activity.
The PETM's big push to warming
The PETM (Paleocene-Eocene Thermal Maximum) is chief among the exceptional, non-cyclical, dramatic warming events, as these papers agree. Even though it began coincident with the 100,000 year orbital pacing, it was not in phase with the 405,000 year cycle that dominates the other Hyperthermals, suggesting an “extra push” was provided, probably by the spectacular eruptions from Canada to Norway that accompanied the opening of the North Atlantic at that time.
Bank withdrawal
Rising temperatures beginning in the late Paleocene must have triggered the slow release of carbon reservoirs, (the peat, permafrost, methane hydrates that were ‘banked’ in the cooling part of the Paleocene) into the atmosphere, until they were exhausted by around the Early Eocene Climate Optimum. All that extra carbon was mobilized into the dynamic biosphere-ocean-atmosphere system, where it amplified and prolonged the warming.
These 3 discrete events, chiefly the PETM, seem to have provided the global warming “push,” that combined with feedbacks from carbon reservoirs on land and the oceans, which switched the cooling Paleocene over to the long, hot Eocene.
Long, hot tail – then and now
If those 3 episodes each represent major singular additions of CO2 to the atmosphere that caused the hot Eocene climate, then the duration of the Eocene hothouse is extraordinarily long. Mathematical modelling for a single pulse of carbon emissions like the PETM shows that, without additional carbon release, we expect a return to background temperatures after a few tens of thousands of years. The observed duration of the PETM is much longer than predicted by those models, requiring prolonged additional carbon input. It would appear that the finger is increasingly pointing at the oceans and their long term influence on the carbon cycle as the main sustaining factor.
uncomfortable parallels to today
As was the case during the Paleocene, our modern Earth has reserves of permafrost and methane hydrates that were ‘banked’ during the cool Pleistocene ice age all the way up to the pre-industrial era. And even though some interglacials may have been warmer than the preindustrial era, deep-ocean temperatures appear to have been only slightly elevated, keeping most deep sea methane locked away - until now. A Paleocene-to-Eocene-like methane release is expected from modern climate change because the deep oceans are now warming by a magnitude unprecedented in the past several million years!
Last year Zeebe and Zachos compared the long tail of the PETM with our present human-caused global warming. They concluded that our own long hot tail will last tens of thousands to hundreds of thousands of years. But they argued that future environmental effects will:
“more likely resemble the end-Permian and end-Cretaceous catastrophes, rather than the PETM,”
due to the far more abrupt nature of modern carbon emissions and warming.
In the second sentence after "New Thinking": " But Turner et al refute this, showing that the Hyperthermal beat continued long after reserves of methane hydrates and permafrost must have been emptied..."
"beat" should, presumably, be "heat."
Pretty sure "beat" is the right word, referring to the rhythmic recurrence of the warming even though the suspected sources for it had dried up. Hence the term "heartbeat" after the block quote.
Doh! Of course you're right, Wheels. I missed the metaphor. But they do use that throughout, now that I look again.
How does this fit with the Wright and Schaller study of last year that has the PETM push occuring within just 13 years. Though at first it seems unlikely, the study has not been refuted as far as I know.
Wili and WheelsOC - yes I meant beat as in heartbeat, sorry if that was confusing. The point is that this 100,000 year and 405,000 year rhythm is not recognized throughout most of the geological record, operating as an oscillation about a background climate state. In the Eocene the climate state was already hot so the orbital oscillations made hyperthermals.
Correction to the post above the ...rhythm IS recognized...
(not sure how 'not' got in there)
Ianperrin - I addressed Wright and Schaller's study in this post (under 'timing matters' about half way down, and see comments at the end)
I think we can say for now that their study was refuted in a series of replies (towards the end of this page) on the basis of:- (a) the heat capacity of the oceans required centuries to warm to the PETM extent, (b) an instant release of carbon in the atmosphere would produce a carbon isotopic shift far larger than observed, and (c) that microfossils ruled out the sedimentary rates claimed. There was also a claim that the apparent varves were drilling artefacts but Wright and Schaller pointed out varve-like layers in land exposures of the same clay unit - so they can't be just drilling artefacts.
The Marlboro Clay that they studied will I'm sure provide valuable insight to the PETM, but as things stand more work is needed to bridge the apparently-varved clay record with the longer term but coarser-resolution records.
Thanks for another illuminating and thought provoking post Howard.
You really know how to spoil somebody's day, don't you?
Cheers BIll F ;)
Wonderful summary, I think we are slowly coming around to a new perspective of the way the globe works. From my studies of atmospheric energy budgets, I come to a very unusual perspective, which this studies seem to support. I can't find the reference, but there is a paper showing that the acceleration of the Indian subcontinent into Asia was preceded by climate change.
Here is the unusual perspective: Tectonic plate motion is much easier to motivate with an energy source that is much larger than earth flux; climate. Solar flux onto earth is 3900 times larger. And the mechanics of GIA, glacial isostatic adjustment, transfer much greater, faster, and mechanically more viable energy impetus for tectonic plate movement.
In my mind I see Antarctica pumping with a potential of 60 petatons of energy, on the pulse of climate. In that model, waves from that action have piled the continental masses on the northern hemisphere.
Gravitation waves play nicely into the process too, but that is too detailed here. Although it may be farfetched, the mechanics are much more robust. If there is any truth to this at all, the seismicity implications within immediate human history may be substantial. Such volcanic reactions to climate with feedback mechanisms through carbon seem to be potentially valid. Even the 39 cubic kilometers of water in the Three Gorges Dam affected seismicity, how much seismicity can 72,000,000 km3, suddenly lifted cause?
Hi Pluvial,
If I understand you correctly, you are suggesting that climate drives tectonics via glacially-controlled isostatic ‘pumping’ of the crust and mantle. It’s an intriguing idea, but as you yourself say, this is an unusual perspective, and not mainstream science. I’d have to research that idea more, but here are my immediate thoughts:
I’m not an expert on mantle dynamics, but I can visualize your concept and I can imagine it having perhaps some effect around regions experiencing strong and repeated isostatic loading as demonstrated in Scandinavia. Repeated depression/elevation of the lithosphere might conceivably contribute to delamination – but I’m no expert there. I’m not sure how important that effect can be, though, as we had active tectonics in ice-cap-free greenhouse eras like the Mesozoic and the Eocene.
Tectonics is driven by the convecting mantle and the sinking of cold, dense plate slabs in subduction zones, with slow upwelling of mantle displaced by that sinking material. The energy driving tectonics is radioactive decay and the cooling of the planet, transferring heat energy from the core and mantle, very slowly, through the crust and atmosphere to space. The energy is immense, but only a tiny portion of that energy ‘leaks’ from the geosphere to the climate system. So the direct thermal energy transfer from the geosphere to the climate is small, much smaller than the energy powering tectonics within the mantle and core.
But tectonics does not drive climate by a direct thermal process. The main long-term effect of tectonics on climate is through the delicate balance between silicate weathering removing CO2 (via mountain building and erosion) and volcanic production adding CO2 (principally subduction over millions of years).
You refer to climate change associated with India’s motion northwards. I think you are referring to the Deccan Traps eruptions at the end of the Cretaceous. Those eruptions were a class of eruption called “Large Igneous Provinces” (it is a bit misleading to call them volcanic because they are so much bigger than any volcano the Earth has seen in the last 16 million years). They did indeed trigger abrupt climate change which triggered a strong extinction event just before the impact that wiped out the dinosaurs. The North Atlantic Igneous province associated with the Eocene hothouse is also a LIP.
You suggest that glacially-controlled isostatic pumping pushed continents away from Antarctica. For me the timing doesn’t match and the plate motions don’t fit. The separation of Africa from Antarctica began in the Jurassic, when there was negligible ice cover on Antarctica. Australia/Tasmania and South America finally broke with Australia at the Eocene Oligocene transition again at a time when Antarctica was essentially free of ice. While the motions of Australia and Africa are northward, the motion of South America is westward. Also, during the Messinian Salinity Crisis, where there’s evidence of strong isostatic motion due to salt accumulation, continents did not move away from the locus of isostatic load (the Mediterranean), they continued to converge.
Your last point linking dams to seismicity is a well-researched topic. Basically all the crust is under stress all the time. When that stress exceeds the strength of the rock it breaks, causing tremors or earthquakes (leaving aside ductile failure for now). Increasing water pressure in groundwater can counteract the pressure keeping 2 sides of a fault locked, making it easier for the fault to move and cause a tremor. It’s the same physics that is associated with fracking. Isostatic unloading has indeed induced seismicity, as you suggest, for example in Scandinavia.
You raise an interesting idea, though, and the interaction of isostatic loading and plate tectonics makes sense, though I’m not sure it can be a driver so much as a localized modifier.
Concerning the speed of the Indian plate, it is known to have travelled far faster that the other fragments of Gondwanaland and also to be thinner than the other fragments of Gondwanaland. My understanding is that the speed being the result of the thin crust is widely accepted and that the Traps are the result of that thinning process. From what I can gather, the inferred thinning process being due to the mantle plume that broke Gondwanaland apart remains hypothesis.
How come when the GW skeptics indicated that GW could be caused by natural cause e.g. global tilt or wobble and by volcanic release of CO2 it was ridiculed. But now these are used as new science supporting GW.
I still would like to know what caused the GW of 1100 years ago. During that period the earth was about as hot as it was in 2000. The startof the most recent leveling off of global warmimh. See NOAA chart on earth and sea heat
MA Rodger - the accelleration of the Indian Plate seems to have been due to the interaction with the Reunion-Deccan mantle plume which pushed the Indian plate faster for a while. The Indian lithosphere is indeed fairly thin probably due to it's encounter with the plume. The traps are the result of enormous flood basalt eruptions, in common with other traps like the Siberian Traps, which are also associated with abrupt climate change.
So the speed of motion is not due to the thin crust, but both are due to the interaction of the Indian Plate with the Reunion-Deccan Mantle Plume.
Fairoakien - it all about the timescales.
The shortest limb of the shortest orbital wobble is about 10,000 years. Human-caused GW is just a couple of centuries - just 2-3% of that. Also, we are on a cooling part of the orbital cycle, not warming.
Orbital cycles modifying climate is not at all new, what is new is the precision of dating which allows detailed sedimentary records across the globe to be compared with greater precision than ever before. The new thinking in the articles I revewed is that even in the Eocene the orbital cycles are not special, which refocuses attention on other, non-orbital causes to turn a cooling climate into a warming climate.
Globally every year volcanoes (including undersea volcanoes) only emit about as much CO2 as a single state like Ohio or Michigan does. The combined volcanic output of CO2 averages at around 260 million tonnes per year compared to about 35 billion tonnes of human-produced CO2 annually. Even the major volcanic eruption of Mount Pinatubo in 1991 only produced about 50 million tonnes of CO2. Human CO2 emissions are equivalent to an extra 11,200 Kīlauea volcanoes or about 360 more mid-ocean ridges!
Large Igneous Provinces are in a league of their own,- see this post, but there hasn't been one for 16 million years.
The "GW" about 1100 years ago is a bit of a myth from old data. Newer, more precise measurements show that since the industrial era we have reversed 6,000 years of that natural orbital cooling in just a century. Contrary to the myth, the planet is hotter now than the Medieval Warm Period (AD 800-1300AD) or the Roman Warm Period (500BC to 400AD), or any time in the last 6,000 years. Even 6,000 years ago, during the “Mid Holocene Warm Period,” when a warmer climate was caused by the same orbital/insolation change that ended the last ice age, global temperatures were about the same as the 2000s.
fairoakien
In addition to Howards points, there is fundamental observation of what is happening today that says AGW isn't natural variability. The warming of the oceans. If all we were looking at was warming of the atmosphere then certainly natural variability might be considered. But the warming of the oceans precludes this.
Heat build-up in the oceans is proceeding at a rate of around 250-300 trillion watts. If that heat were all going into the atmosphere then air temperatures - what we think of as climate - would be rising at around 15 deg C per decade.
And the rate of heat accumulation in the oceans says something very basic. It isn't coming from anywhere here on Earth. The largest heat source on the planet is geothermal heat. And geologists have been able to estimate the total rate of geothermal heat flowing out of the planet. 44.2 trillion watts. Simply too small to supply the energy we are seeing added to the oceans.
So the only conclusion that is possible is that this extra heat is coming from somewhere off the Earth.
In contrast the flow of heat from the Sun and then being radiated out to space by the Earth is around 120,000 trillion watts. Even a tiny disturbance in these two flows can easily supply all the heat we are observing. But the heat flow from the Sun hasn't increased - we have been monitoring the Sun for around 40 years now. If anything the energy flow from the Sun has declined ever so slightly.
Which leaves the heat flowing out from the planet as the source. If something is restricting that flow then the heating can easily be explained. And the something that modulates that flow is the Greenhouse Effect.
Key to this however is that this isn't natural variability! That can't explain what we are seeing.
So when skeptics suggest that natural variability is the explanation, they aren't looking at all the evidence.