From the University of Colorado at Boulder a claim that computer modeling has solved the problem, with an atmosphere that is 20,000 parts per million of CO2 and 1,000 ppm of methane.
CU study shows how early Earth kept warm enough to support life
Scientists tackle faint young sun paradox with 3-D climate models
Solving the “faint young sun paradox” — explaining how early Earth was warm and habitable for life beginning more than 3 billion years ago even though the sun was 20 percent dimmer than today — may not be as difficult as believed, says a new University of Colorado Boulder study.
In fact, two CU-Boulder researchers say all that may have been required to sustain liquid water and primitive life on Earth during the Archean eon 2.8 billion years ago were reasonable atmospheric carbon dioxide amounts believed to be present at the time and perhaps a dash of methane. The key to the solution was the use of sophisticated three-dimensional climate models that were run for thousands of hours on CU’s Janus supercomputer, rather than crude, one-dimensional models used by almost all scientists attempting to solve the paradox, said doctoral student Eric Wolf, lead study author.
“It’s really not that hard in a three-dimensional climate model to get average surface temperatures during the Archean that are in fact moderate,” said Wolf, a doctoral student in CU-Boulder’s atmospheric and oceanic sciences department. “Our models indicate the Archean climate may have been similar to our present climate, perhaps a little cooler. Even if Earth was sliding in and out of glacial periods back then, there still would have been a large amount of liquid water in equatorial regions, just like today.”
Evolutionary biologists believe life arose on Earth as simple cells roughly 3.5 billion years ago, about a billion years after the planet is thought to have formed. Scientists have speculated the first life may have evolved in shallow tide pools, freshwater ponds, freshwater or deep-sea hydrothermal vents, or even arrived on objects from space.
A cover article by Wolf and Professor Brian Toon on the topic appears in the July issue of Astrobiology. The study was funded by two NASA grants and by the National Science Foundation, which supports CU-Boulder’s Janus supercomputer used for the study.
Scientists have been trying to solve the faint young sun paradox since 1972, when Cornell University scientist Carl Sagan — Toon’s doctoral adviser at the time — and colleague George Mullen broached the subject. Since then there have been many studies using 1-D climate models to try to solve the faint young sun paradox — with results ranging from a hot, tropical Earth to a “snowball Earth” with runaway glaciation — none of which have conclusively resolved the problem.
“In our opinion, the one-dimensional models of early Earth created by scientists to solve this paradox are too simple — they are essentially taking the early Earth and reducing it to a single column atmospheric profile,” said Toon. “One-dimensional models are simply too crude to give an accurate picture.”
Wolf and Toon used a general circulation model known as the Community Atmospheric Model version 3.0 developed by the National Center for Atmospheric Research in Boulder and which contains 3-D atmosphere, ocean, land, cloud and sea ice components. The two researchers also “tuned up” the model with a sophisticated radiative transfer component that allowed for the absorption, emission and scattering of solar energy and an accurate calculation of the greenhouse effect for the unusual atmosphere of early Earth, where there was no oxygen and no ozone, but lots of CO2 and possibly methane.
The simplest solution to the faint sun paradox, which duplicates Earth’s present climate, involves maintaining roughly 20,000 parts per million of the greenhouse gas CO2 and 1,000 ppm of methane in the ancient atmosphere some 2.8 billion years ago, said Wolf. While that may seem like a lot compared to today’s 400 ppm of CO2 in the atmosphere, geological studies of ancient soil samples support the idea that CO2 likely could have been that high during that time period. Methane is considered to be at least 20 times more powerful as a greenhouse gas than CO2 and could have played a significant role in warming the early Earth as well, said the CU researchers.
There are other reasons to believe that CO2 was much higher in the Archean, said Toon, who along with Wolf is associated with CU’s Laboratory for Atmospheric and Space Physics. The continental area of Earth was smaller back then so there was less weathering of the land and a lower release of minerals to the oceans. As a result there was a smaller conversion of CO2 to limestone in the ocean. Likewise, there were no “rooted” land plants in the Archean, which could have accelerated the weathering of the soils and indirectly lowered the atmospheric abundance of CO2, Toon said.
Another solution to achieving a habitable but slightly cooler climate under the faint sun conditions is for the Archean atmosphere to have contained roughly 15,000 to 20,000 ppm of CO2 and no methane, said Wolf. “Our results indicate that a weak version of the faint young sun paradox, requiring only that some portion of the planet’s surface maintain liquid water, may be resolved with moderate greenhouse gas inventories,” the authors wrote in Astrobiology.
“Even if half of Earth’s surface was below freezing back in the Archean and half was above freezing, it still would have constituted a habitable planet since at least 50 percent of the ocean would have remained open,” said Wolf. “Most scientists have not considered that there might have been a middle ground for the climate of the Archean.
“The leap from one-dimensional to three-dimensional models is an important step,” said Wolf. “Clouds and sea ice are critical factors in determining climate, but the one-dimensional models completely ignore them.”
Has the faint young sun paradox finally been solved? “I don’t want to be presumptuous here,” said Wolf. “But we show that the paradox is definitely not as challenging as was believed over the past 40 years. While we can’t say definitively what the atmosphere looked like back then without more geological evidence, it is certainly not a stretch at all with our model to get a warm early Earth that would have been hospitable to life.”
“The Janus supercomputer has been a tremendous addition to the campus, and this early Earth climate modeling project would have impossible without it,” said Toon. The researchers estimated the project required roughly 6,000 hours of supercomputer computation time, an effort equal to about 10 years on a home computer.
h/t to Dr. Leif Svalgaard
Janice Moore says:
July 9, 2013 at 8:22 pm
how do we know that the sun was 20% (or so) dimmer back then?
“Since its birth 4.5 billion years ago, the Sun’s luminosity has very gently increased by about 30%. This is an inevitable evolution which comes about because, as the billions of years roll by, the Sun is burning up the hydrogen in its core. The helium “ashes” left behind are denser than hydrogen, so the hydrogen/helium mix in the Sun’s core is very slowly becoming denser, thus raising the pressure. This causes the nuclear reactions to run a little hotter. The Sun brightens.”
I skimmed the paper. Under “Methods” I found this:
“CAM3 is a widely used GCM and has been extensively validated against modern climate. Regional and seasonal biases are present, but global and annual mean climatological statistics agree well with modern climate (Collins et al., 2006; Hurrell et al., 2006).”
Which I interpreted as meaning CAM3 gets the “right” global parameters but for the “wrong” reasons. Given recent WUWT articles on GCMs, I suspect the “agree well” part is open to interpretation. 🙂
Also under “Methods”:
“Continental configurations, topography, planetary rotation rate, ocean heat transport, cloud droplet sizes, land-based glacial ice, and surface vegetation are assumed to be those of the present day. Thus here we isolate the effects of reduced solar insolation and increased greenhouse forcing.”
BTW, the Faint Young Sun Paradox has been “explained” before at WUWT:
http://wattsupwiththat.com/2010/04/07/faint-sun-paradox-explained-by-stanford-greenhouse-effect-not-involved/
http://wattsupwiththat.com/2012/10/16/a-simple-resolution-to-the-faint-young-sun-paradox/
http://wattsupwiththat.com/2009/08/19/finding-an-answer-to-the-faint-sun-paradox-carbonyl-sulphide/
One last thing: The press release says “Methane is considered to be at least 20 times more powerful as a greenhouse gas than CO2…”. I’ve seen this claim before, but I see the IR absorption spectrum of methane partially overlaps water vapor’s spectrum. Can anyone explain to me why methane is so much more of a so-called “greenhouse gas” than carbon dioxide?
I remember when Galileo hit Jupiter’s atmosphere that it was so different from what various past theories assumed it would be that, I seem to recall, that at least one scientist said Jupiter may not have originated from our solar system.
@Janice Moore –
In re the Sun being less luminous at the beginning compared with the present, it is known that as main-sequence stars (which the Sun is, class G) age, their luminosity gradually increases over time. This has been demonstrated by spectrographic analyses which have been able to correlate luminosity of these stars to their ages. 20 percent is, as I understand it, probably a reasonable value for the increase in the Sun’s luminosity over the 4-1/2 billion years of the Earth’s existence.
The only 20,000 ppm of CO2 is suspicious, however, and even more so is the assumption that the density of the Earth’s atmosphere has not changed over the lifetime of the Earth. For one thing, the source of the oxygen in Earth’s atmosphere is from photosynthesis by blue-green algae that did not evolve until about 2 billion years ago. This process proceeded on a scale sufficient to convert nearly all of the CO2 into O2 – oxygen, by about 1 billion years ago, and for it to be possible for the atmosphere to be 20 percent O2 today, it originally had to be 20 percent CO2 (not 2 percent which is 20,000 ppm) just for that alone – and was probably even more due to carbonate deposition in coral-type organisms with calcareous skeletons that would have absorbed more CO2. And as for the constant density of the atmosphere, it is known that large asteroid impacts, such as the K-T event, blew off a substantial part of the Earth’s atmosphere – some estimates of as much as a third from K-T alone – which was never recovered.
I am sure they are very grateful for the 184 teraflop cluster featuring 16,416 total cores housed in 342 chassis. And the electrical power. And the cooling system. At least we got an artist’s impression.
Much denser atmosphere. Adiabatic lapse rate rules.
Chad Wozniak says:
July 9, 2013 at 8:57 pm
For one thing, the source of the oxygen in Earth’s atmosphere is from photosynthesis by blue-green algae that did not evolve until about 2 billion years ago. This process proceeded on a scale sufficient to convert nearly all of the CO2 into O2 – oxygen, by about 1 billion years ago, and for it to be possible for the atmosphere to be 20 percent O2 today, it originally had to be 20 percent CO2 (not 2 percent which is 20,000 ppm) just for that alone
Photosynthesis converts H2O into O2 not CO2!
Leif Svalgaard says:
July 9, 2013 at 6:33 pm
The researchers estimated the project required roughly 6,000 hours of supercomputer computation time, an effort equal to about 10 years on a home computer.
somebody got the numbers wrong here. Janus comprises more than 16,000 compute cores so the comparison is not credible. 6000 hours is 2/3 year, so Janus should only by 15 times as powerful as a ‘home computer’. Off by a factor of 1000.
++++++++++++++++++++++++++++++++++++++++++++++++++++
Leif,
As someone who works in IT for a living, I would read “roughly 6,000 hours of supercomputer computation time” as CPU time, not clock/calendar time. If that is the way it was meant, then the total number of cores in the Janus supercomputer is not relevant to the comparison. This would be saying that each individual core is 15 times faster than a typical home computer.
I would be very skeptical of any claim that a project like this would be allowed to burn 100% of the CPU time on this kind of computer for 9 months.
Hilarious – what do they plan to do about the normal sun paradox they’ve just created?
If CO2 levels started at 20,000ppm, and geological absorption of CO2 was weaker than today, then WTF brought down CO2 levels as the sun heated up?
And why can’t we infer from that that the world has no problem absorbing almost any amount of CO2 we throw into the atmosphere?
Its a bit like Deep Thought, in Hitch Hiker’s Guide to the Galaxy – the answer to life, the universe and everything is 42. Now you have to build an even bigger computer to discover what the question was.
Geologically, the young, Archaean Earth was a less-evolved, less compositionally-differentiated, planet than it is now, with a thinner crust, a higher geothermal gradient and widespread volcanism. It was hotter, with a higher heat flow to the surface, through a thinner crust.
The majority of the Archaean crustal rocks that you find preserved in the world’s greenstone belts are volcanic rocks, indicating a long Archaean history of widespread eruptive volcanism . This is also why so many of the worlds metalliferous mineral deposits occur in Archaean rocks: the processes that deposit these are driven by high heat flow and volcanic activity.
The Archaean Earth was a very different planet, within which the radioactive decay processes that generate heat in its interior were many half-lives younger and more energetic than they are now.
It would be hard quantify the contribution of higher heat flow up through the Archaean crust to atmospheric temperature at the time, but you only have to look at the rocks that were produced to see that the crust was generally thinner and the geothermal gradient significantly higher.
Then you only have to look at the prevalence of ocean floor eruptive rocks, sulphidic geothermal vent deposits, etc. during the Archaean to realise that the oceans were probably warmed from below as well as well as from above, at numerous localities if not always globally, to provide the temperatures at which life could develop, thrive and evolve.
Fainter young sun, but warmer young Earth! A rather nice balance, in fact.
CAM3 is an earlier version of the atmospheric model component of the land-ocean NCAR GCM Community Earth System Model (CESM). The newest version is CAM5, which was among the CMIP5 models recently evaluated by Jonathan Jiang, et al.
The CAM5 makes an annual average (+/-)9.4% error in total annual average cloud fraction, relative the the latest A-Train satellite observed cloud fraction. That’s equivalent to an annual uncertainty of (+/-)3.9 W/m^2 in atmospheric energy flux, under the fairly well-known conditions of the modern terrestrial climate.
The NCAR ccsm3 model, which includes the CAM3 as part of its internal set of four coupled models, is worse. It makes an average (+/-)12.5% error in total annual cloud fraction, relative to A-Train observations. That’s equivalent to an uncertainty of (+/-)5.2 W/m^2 in modeled atmospheric energy flux.
The published claim assumes the CAM3 model can accurately model the terrestrial cloud cover of 3 billion years ago, under nearly unknown conditions of climate. However, CAM3 can’t model current cloud cover under well known conditions of climate.
All they’ve done is model a previous atmosphere under the same set of assumptions of water-vapor-enhanced feedback that govern CAM 3 and CAM5 in calculating the current atmosphere.
Why, then, is it a surprise that they find several thousand ppm of CO2 will, with a dollop of methane, warm the archaic climate? That conclusion is built into the model as an a priori condition. All they needed to do was find the levels of GHGs they needed to get the answer they want.
CAM3 is unable to model the present atmosphere. There’s no reason to suppose it can model past atmospheres, super-Janus computer or no. I’m sure they got really pretty graphics of Earth, though, that look just like satellite photographs. Super computers are great for that.
So Janus is looking back hundreds of millennium for us mere mortals, and naturally has powers to look centuries ahead. Perhaps he will be flattered by the impersonation by the academics at CU.
This supercomputer series illustrates well the point that academics and Progressives have always interpreted all myths, legends, and religions as merely justifications for the power of the elite. It is written into every book on ancient history. And now the gods appear again, uniquely programmed, designed and constructed in a modular data center assembled in one day on campus, ready to justify the growing power of the elite through green ideology. Science, gods, who cares. Whatever comes to hand.
But Janus was guardian of doors and passageways, and sometimes of marriages and new beginnings. This supercomputer won’t be applied to tracking people and goods, will it Maestro? Because I think this 3D baby talk is not plausible at all for something of that size.
Dear Leif Svalgaard, Dear Chad Wozniak,
Thank you, so much, for your kindness in answering such a basic science question. Thank you, too, for not mocking my ignorance.
Gratefully,
Janice
P.S. If I am not mistaken, the science about the Sun’s warmth assumes that the Sun is not unique among stars. It just might be! #[:)]
Eric Worrall ,
The lesson to take from this is that knowing the right question is often more valuable than knowing the right answer.
Gary Hladik says:
July 9, 2013 at 8:45 pm
Can anyone explain to me why methane is so much more of a so-called “greenhouse gas” than carbon dioxide?
Roughly speaking the efficiency of a greenhouse gas depends on the number of different ways the molecules can vibrate, stretch, and shake. That is why N2 [with only two atoms] is not a GHG, H2O, O3, and CO2 [with three atoms] are GHGs, and Methane CH4 [with four atoms] is a very strong GHG, because it has many more ways to ‘shake and bake’, stretch and bend, etc.
MattS says:
July 9, 2013 at 9:22 pm
As someone who works in IT for a living, I would read “roughly 6,000 hours of supercomputer computation time” as CPU time
It most certainly is of CPU time. If they used only 100 cores all running at full speed, then in 60 hours they will have used 6000 hours of CPU time. But they don’t need a Janus to do that, 25 PCs each with 4 cores could do that as well in 60 hours, not 10 years. They should have told us how many wall-clock hours the calculation to [perhaps 1 hour] and compared that to the 10 years of ordinary PC wall-clock.
Mike McMillan says:
July 9, 2013 at 9:03 pm
Much denser atmosphere. Adiabatic lapse rate rules.
=========
Agreed. Any model that assumes the density of the atmosphere has remained unchanged over 3.5 billion years is nonsense. The early atmosphere was denser and the equilibrium point between outgoing and incoming radiation was higher in the atmosphere, raising the surface temperature due to the connection between the equilibrium point and the surface established by the lapse rate.
Once GHG reaches saturation it cannot raise the surface temperature above that allowed by the lapse rate, no matter how much GHG you add. Any extra back radiation simply increase the circulation. The only way to raise the surface temperature is to increase in the density of the atmosphere, allowing the lapse rate to operate over a greater height.
For this reason place of low elevation on earth are warmer than places of high elevation, for a given amount of incoming energy. If the atmosphere was thicker, today’s mountain tops would be much warmer, and the valleys warmer still.
“… like Deep Thought, in Hitch Hiker’s Guide to the Galaxy – the answer to life, the universe and everything is 42.” [Eric Worrall 9:25PM 7/9/13]
LOL.
That reminded me of, yes, I think this IS on topic (given the vague and dreamlike paper we are discussing)…. I was overwhelmed my first quarter of my freshman year of college. So much more reading than high school. A much quicker pace. To top it all off, my calculus professor, Dr. Stanley Luke, while he was extremely kind and brilliant, had a heavy East Indian accent. One night, I dreamed that I was solving a long, complicated, math problem. I struggled along and struggled along until…. I CAME UP WITH THE ANSWER. In the morning, I remembered it, too! It was…………. WD-40. Oh, man, I am laughing out loud, even now. While I was asleep, it seemed like a VERY IMPORTANT answer.
Climatologists need to wake up.
Janice Moore says:
July 9, 2013 at 9:36 pm
the science about the Sun’s warmth assumes that the Sun is not unique among stars. It just might be!
No it might not. We observe many stars just like the Sun, and the theory of stellar evolution explains the Sun just as well as it does those stars.
Tweedledee: ‘if it was so, it might be; and if it were so, it would be; but as it isn’t, it ain’t. That’s logic.’
lsvalgaard says:
July 9, 2013 at 8:31 pm
Janice Moore says:
July 9, 2013 at 8:22 pm
how do we know that the sun was 20% (or so) dimmer back then?
And, Dr. Svalgaard, PUUUH–leez do not utter some enigmatic oracle like, “That is the long-settled science.
Unfortunately, it is settled science. We have a very good understanding of stellar evolution. We know that the theory works because we observe millions of stars in all phases of their life: old stars, young stars, stars in middle age [like the sun], etc. so we can readily check that the theory works by simply looking around at many stars. http://en.wikipedia.org/wiki/Stellar_evolution or http://cosmos.phy.tufts.edu/~zirbel/ast21/handouts/StellarEvolution.PDF
Unfortunately, this too is based on models and assumptions. It is quite an assumption that the HR diagram represents evolution. It is equally valid to assume that the HR diagram looks as it does because they are the most stable configurations. Then the only reason we get such a nice diagram is because there are so many stars, all sitting at a different stable configuration. From there, the internal composition of the sun is based on models which are based on the assumption of nuclear reactions.
There are other explanations, perhaps electrical in nature. From there we can say that the faint young sun paradox might not even be a thing.
Larry Kirk,
Well said. Moving back even further to the Hadean, how do the Warmists explain the molten surface before the first crustal solidification?
Maybe the motley collection of self proclaimed “climatologists” should ask somebody with an understanding of something other than calculating averages, say a geophysicist specialising in the Hadean and Archaen, to explain in terms even a “climatologist” can understand, the process by which a molten blob cools to the point where we can live on it.
The blob is still cooling – slowly but inexorably. No amount of CO2 in the atmosphere can prevent this happening.
Live well and prosper.
Mike Flynn.
Given how close the moon was back then, wouldn’t it also have generated appreciable heat in the Archaen earth as it orbited? It has gradually spiraled outward in its orbit since then, reducing this effect as the sun brightened. Not sure just how big this effect was, but I recall seeing ice volcanoes on other planetary moons fed by gravitational interactions with their home world.
Pat Frank says (July 9, 2013 at 9:33 pm): “Why, then, is it a surprise that they find several thousand ppm of CO2 will, with a dollop of methane, warm the archaic climate? That conclusion is built into the model as an a priori condition. All they needed to do was find the levels of GHGs they needed to get the answer they want.”
Maybe they could have used Willis Eschenbach’s “model” of GCMs instead of CAM3. It would have saved the taxpayers many kilobucks’ worth of computer time. 🙂
Jarryd Beck says:
July 9, 2013 at 9:53 pm
It is quite an assumption that the HR diagram represents evolution.
We can compute the observed shape of the H-R diagram in fine detail from basic physics, that is usually taken as confirmation of the validity of the theory.
based on the assumption of nuclear reactions.
From basic physics we can calculate how much energy should be released by those reactions and how many neutrinos should be produced. The calculated quantities match the observations very nicely. You see, models are an encapsulation of our knowledge. When the predictions match the theory we believe the theory.
There are other explanations, perhaps electrical in nature.
No, there are no other explanations that predicts numerically what is observed.
Apparently if you work this Janus computer very hard, it requires over 500,000 watts, and they seem very proud of the fact that their simulation took 6,000 hours. Only 3M kwh, it was probably green electricity
lsvalgaard says (July 9, 2013 at 9:48 pm): “Roughly speaking the efficiency of a greenhouse gas depends on the number of different ways the molecules can vibrate, stretch, and shake.”
Thank you!