Back to the future: Ancient climates give an integrated view of our possible climate future
The Pliocene and Eocene as analogs for near-future climates
Sometimes a single image can change the way you see the world. That happened when I came across a paper with Dr. Kevin Burke as lead author offering a perspective on climate change that I had not considered before. The concept was that we can look to past eras in the Earth’s history as analogs for our climate futures.
A picture is truly worth (much more than) a thousand words, and this one from the Burke paper is a brilliant presentation summarizing a telescoping view of changes in global temperature both recently and over geologic time.
Note the change in scale along the time axis, from years on the far right, through thousands of years (kyr) in the middle, and then millions of years (Myr) at the far left. This one graph captures the current rapid increase in temperature, the dynamics and range of temperatures during the ice ages (with a 100,000 year cycle – middle panel) and then looks farther back in time.
While this history and its causes are interesting stories on their own, the authors also use this exceptionally informative graph to pinpoint two periods in this long story that might offer analogs and insights into the climate future that awaits us.
A simple summary of their in-depth analysis comes from comparing the far right panel with the longer-term changes in temperature to the left. They propose that the lower projections of how much we will warm the planet by 2100 (~ 2 degrees Celsius) would bring the average global temperature to that of the middle Pliocene, ~5 million years ago. More extreme change predictions (~5 degree increase) would bring that global average back to that of the late Eocene, 40-50 million years ago.
Dr. Burke and colleagues talk of winding the climate clock back 5 million and 50 million years to catch a glimpse of our future. Back to the Future indeed. There will be some caveats on the use of these analogs at the end of this essay, but lets not break the narrative train just yet. Instead, lets ask:
What did the Earth look like in those eras?
For simplicity, we’ll look at just three major characteristics: Temperature and its distribution, vegetation types and shifts, and, the big one, average sea level.
The Pliocene
One of the most consistent predictions and most consistently measured characteristics of our changing planet is that warming has been and will be most extreme in the Arctic. Climate models projecting differences in ocean temperatures in the Pliocene (5 million years ago) relative to today display those same trends: up to 8 degrees warmer than current in the Arctic, with relatively little change in the tropics.
Note in particular the predicted increase in ocean temperatures off the west coast of South America. This is a classic “EL Niño” indicator, and some predict that El Niño was a permanent condition in the Pliocene. El Niño (or ENSO – the El Niño Southern Oscillation) is currently a cyclic occurrence with major implications for severe weather across the Americas and Australia, and especially for fisheries off Peru. Some see a return to a more frequent if not constant El Niño-like condition as part of our climate change future.
The extreme warming between Greenland and Scandinavia relates to the climate-critical Atlantic Meridional Overturning Circulation (AMOC) discussed in an earlier essay.
Predicted vegetation types for the Pliocene show some significant differences from what we see today. For the U.S., much of the eastern, central and southern regions currently in forest are projected in this study to have been more arid, supporting grasslands and savannas, while much of the desert southwest hosted those same vegetation types, reflecting a prediction of more humid conditions than is currently the case.
Across Canada and Scandinavia, Pliocene evergreen boreal forests would have been limited to the far north, replaced by temperate, mostly deciduous forests across much of what is the current boreal forest zone. The disappearance of boreal zone species has also been predicted for the northern U.S. as well as part of our climate future.
In addition to arctic warming, the most consistently predicted and consistently measured change in our current global climate dynamic is the inexorable rise in sea levels. This can be caused either by thermal expansion (water in the warmer ocean will expand and increase in volume, even without water being added by glacial melt in Greenland and Antarctica) and then also by addition through ice melt. Current rates of sea level rise are about half due to expansion and half to addition.
In this Pliocene world, permanent ice fields in Greenland were restricted to the northeastern quadrant of that large island, and sea levels were ~15 meters (~50 feet) higher than current. Impacts already ascribed to rising sea levels in our warming world have occurred with a rise of about 20-25 centimeters (8-10 inches). Note how widely sea levels have fluctuated (as much as 100 meters) in the last 800,000 years as the massive glaciers we now know were associated with the ice ages waxed and waned. It is clear that the amount of land-based ice in the world has been a dominant factor in determining sea level.
The Eocene
The Eocene scenario is even more extreme, but the ground rules have changed a bit as well. Five million years is not enough time for significant movement of the continents, but fifty million years is.
Fifty million years ago, the Indian subcontinent had not yet plowed north into Asia, causing the rise of the Himalayan mountains. Australia was moving northward into the climatic desert zone centered on 30 degrees south of the equator. Permanent ice fields were absent in both Greenland and Antarctica. All this is to say that the differences in basic geography and surface conditions that direct the movement of weather systems and set the boundary conditions for climate make a direct comparison with the present harder to make.
But there is another excellent recent paper with Dr. Nicholas Herold as first author that has described just how different that playing field was (with 55 million years ago as the time frame). The combination of land masses in motion and the absence of ice is projected to have given the Earth a very different look in the Eocene. Note the isolation of India from Asia, the proximity of Australia to Antarctica, and the extensive shallow seas that separated parts of what is now a united Eurasian land mass. In this world, there would have been no need for either the Panama or Suez canals!
The goal of the Herold et al. paper was to make available a consistent set of boundary conditions for modeling and predicting climate in the Eocene. The first figure at the top of this essay captures the average 5 degree increase in temperature that aligns this climate future with the Eocene epoch. Models using these boundary conditions predict a similar distribution of temperatures as currently exists, but warmer everywhere.
Part of this model-boundary-setting exercise was to estimate the rough distribution of vegetation types. There is some circularity here, as the climate estimates determine biome distribution, and vegetation plays a role in climate, but the main forcing function here is climate to vegetation. Using an existing model of this climate-biome dynamic, tropical forests are predicted to have been present over much of North America in the Eocene while evergreen conifer forests of the boreal zone were absent. Arid vegetation types were more extensive, all permanent ice fields were gone and Antarctica supported both tundra and forested systems!
Given the distribution of lands and seas in that map of the Eocene Earth, and the absence of the permanent ice fields we now have in Greenland and Antarctica, we would expect sea levels to be much higher. Describing that against the movement of continents is a little tricky, but one estimate puts the increase at about 70 meters, while another puts the value at about 60 meters at 35 million years ago.
Current estimates of sea level rise that would occur if all of Greenland and Antarctic ice melted are also around 70 meters. Those two massive accumulations of ice contain ~99% of the fresh water on Earth.
Caveats
The Pliocene and Eocene examples, driven by estimated changes in global temperature of 2 and 5 degrees Celsius, project stunningly different worlds than we now inhabit. Temperatures in the far northern hemisphere would be much higher than the current global average. Vegetation changes would be extreme with major biomes shifting towards the poles or disappearing altogether. Imagine an Antarctic continent devoid of ice and covered in grasses, trees and shrubs.
But the elephant in the room is sea level rise. To repeat, all of the current concerns and outcomes about sea level rise in our warming world have resulted from cumulative increases of less than half of a meter. These epochal Pliocene and Eocene scenarios suggest rises of from 20 to 70 meters, or at least 100 times what we have experienced to date.
The major caveat then for the relevance of these scenarios to our climate future is one of time scale. These scenarios represent Earths that have had millions of years to respond to changing temperatures. They are at equilibrium, if you like, in terms of ice and vegetation coverage. A fundamental truth about our current climate dynamic is that we are NOT in equilibrium. We are forcing change at a rate that is unprecedented in the geologic record.
Two major reservoirs in the Earth’s climate system are resisting our movement towards the Pliocene and Eocene worlds: Oceans and Ice.
The global ocean circulation system has, to date, absorbed about 90% of the excess heat produced by greenhouse gas increases, delaying global temperature increases in the atmosphere. The global ocean circulation system, also called the Thermohaline system, takes about 1,000 years to complete a turn and is sinking much of the heat increase into the deep oceans, also minimizing current increases in surface ocean temperature. Heat energy transferred to the deep oceans will resurface eventually, and impact atmospheric temperatures as well, but on a time-scale of hundreds to thousands of years.
And while the rate of ice melt is increasing in both Greenland and Antarctica, current projections call for sea levels to increase by less than 1 meter through 2100; far less than 20-70 meters! The Antarctic ice sheet is as much as 3,000 to 4,000 meters thick and at current rates of loss will take more than 100,000 years to disappear completely. There is uncertainty as to how fast that rate of ice loss will increase over time.
So, while we are moving toward a Pliocene/Eocene future as described by global temperatures, inertia in both the marine and ice systems of the Earth will delay that new equilibrium in terms of sea level rise for millennia.
This is good news.
Still, maybe it is time to start extending the timeline of our climate projections well beyond 2100 (the first figure at the top of this essay has done that). Children born today have a good chance of seeing 2100, and their children and grandchildren might see 2150 and beyond.
The Pliocene and Eocene are out there as signposts pointing the direction in which we are headed. Maybe with this perspective we can think about how to slow our progress back into the future.
Acknowledgement: Thanks to Dr. Matthew Huber and to Dr. Nicholas Herold for helpful comments on an earlier draft of this essay.
Sources
The paper by Burke et al, including the first figure in the essay, is here:
https://www.pnas.org/content/115/52/13288
A description of Pliocene climate is here:
https://en.wikipedia.org/wiki/Pliocene_climate
And the image of Pliocene ocean temperatures is here:
https://commons.wikimedia.org/wiki/File:Pliocene_sst_anomaly.png
The map of Pliocene vegetation is modified from:
https://upload.wikimedia.org/wikipedia/commons/9/9e/Pliocene_megabiome.png
https://en.wikipedia.org/wiki/Pliocene_climate
Projected distributions of forest species and forest types for the U.S. are derived by the Forest Inventory and Analysis program of the U.S. Forest service and can be found here:
https://www.nrs.fs.fed.us/atlas/tree/future_iv_318.html
Pliocene sea level rise estimate is from:
https://royalsocietypublishing.org/doi/10.1098/rsta.2012.0294
Measurements of modern sea level rise (since 1900) are presented here:
https://climate.nasa.gov/vital-signs/sea-level/
The Eocene paper with Herold as first author, and including both the map of the Earth’s surface and the distribution of vegetation types, is here:
A reference for consistent distribution of temperatures now and in the Eocene is here:
https://royalsocietypublishing.org/doi/10.1098/rsta.2013.0123
and some graphics are here:
References for a range of sea level rise in the Eocene include:
https://en.wikipedia.org/wiki/Past_sea_level
https://commons.wikimedia.org/wiki/File:Phanerozoic_Sea_Level.png
https://royalsocietypublishing.org/doi/10.1098/rsta.2012.0294
Estimates that 99% of fresh water on Earth is in the Greenland and Antarctic ice sheets, and that sea levels would rise 66 meters if both melted completely are from:
https://nsidc.org/cryosphere/quickfacts/icesheets.html
A reference for 90% of excess heat energy due to greenhouse gases having been absorbed by the ocean is here:
https://www.ncei.noaa.gov/news/ocean-heat-content-rises
A description and depiction of the Thermohaline circulation system is here:
https://en.wikipedia.org/wiki/Thermohaline_circulation
Estimates of how long it might take for all Antarctic ice to melt are calculated from data found here:
https://en.wikipedia.org/wiki/Antarctic_ice_sheet
The projection for actual sea level rise through 2100 is from the IPCC Sixth Assessment report: