The Biggest Control Knob

Introduction
This post is based on a talk given by Professor Richard Alley of Penn State University.

The talk was given during the Fall 2009 meeting of the American Geophysical Union (AGU)

Alley's presentation is lively, interesting and informative.

I have written this long post as a means of improving my understanding of the underlying issues.

My comments will be in square [] brackets. Any unattributed quotes ("") are from Alley during the talk.

The talk can be found at this link.

Below is a photo of Alley during his presentation and his opening slide.

As always click on any of the images to see a larger (and clearer) version opened in a new window or tab.



Alley started by noting that he is not an atmospheric scientist but that he teaches a course on the history of atmospheric science, and that he planned in the talk to take the audience on a history tour.

In his first slide shown below he made the point that many people are interested in Climate Science, many so that they can understand the issues, though sadly some so that they can distort the science. He emphasised the point that although it is clear that CO₂ produces atmospheric warming there are some important issues that CO₂ warming tells us little about. These include regional climate, ENSO and a number of others. "So there are more things that CO₂ that matter, studying them is wise, they should be in what we do as we go forward. As I focus on CO₂, remember that I am looking at this through the lens of a historian."



He then displayed part of an email that had been sent to the administration of the University and copied to him.



Alley made the point that there was a logical fallacy in the email, [and from the laughter from the audience I expect that most if not all of the people in the audience recognised the silliness of the claim and the ignorance of the author of the email. Alley debunks the claim towards the end of the talk when he discusses the warming after the last glacial maximum.]

Alley briefly discussed his course and made the point that it is best to teach the subject from the perspective of the broad history of the globe's climate, and that when you try to teach that, CO₂ keeps inserting itself everywhere we look. "Really if you leave CO₂ out nothing makes sense, if you put CO₂ in a whole lot of it makes sense and then you can put the other pieces into the puzzle and make it work. And what we are going to do is take a wander through the history of the Earth' s climate ... we are going to start at the beginning and see where we end up."

Faint young sun paradox

The output from the sun was only about 70% of current levels when the Earth formed 4.6 billion years ago. This would mean that there would be no liquid water on the surface early in the Earth's history, but water related sediments have been found that date to as early as 3.8 Billion years ago. The most likely explanation is a stronger greenhouse. Scientists are trying to find more greenhouses, an issue that is briefly discussed at this link.

The reason for the increase in the output of the sun is that Helium is produced by the fusion of Hydrogen, Helium is heavier than Hydrogen, the sun therefore contracts to counter the increased gravity and the contraction causes more energy to be output. This conclusion is an outcome of astrophysics theory, there is no direct evidence for it.

[An alternative, though minority view, is that Cosmic Rays warm the young Earth. Alley did not mention this probably because there is evidence against the Cosmic Ray Hypothersis as he explains later in the talk. Details can be found at the link above.]



The best explanation for the ability of the Earth to maintain temperatures in a relatively limited range, while the Sun increases its output is the Rock-Weathering Thermostat.

Rock-weathering thermostat

The image below shows the Rock-Weathering Thermostat.



Volcanos put out CO₂ and rock (CaSiO₃) and the rate at which volcanoes put out CO₂ and rock has very little to do with the surface temperature. The recombination of the CO₂ and rock (called rock weathering) is a chemical reaction that is thermally activated. Turn up the temperature and the recombination of the CO₂ with the rock goes faster and draws down CO₂ from the atmosphere (because the CO₂ is combined with the rock). The weathered rock eventually goes down into the ocean and enters a subduction zone and comes back out of a volcano. So we have a thermostat; turn up the temperature CO₂ is drawn down and that cools; turn down the temperature and CO₂ builds up, which warms. If you put this physics into models it is sufficient to explain what happened and so far we can't find anything else that is. This is not totally tied down yet but is clearly the best explanation for why the planet's temperature has been kept in a range to allow life and liquid water for 4 billion years.

Is there any evidence, other than the compelling nature of the chemistry, to demonstrate that the rock weathering thermostat is regulating the long term behaviour of the plantet? It turns out that there is. The CO₂ thermostat takes a half a million years, or so, to work, which means that it is possible to get fairly large fluctuations in temperature before this thermostat brings it back. The Snowball Earth periods of the Earth's history, demonstrate that the thermostat is working with moderately high confidence.

Snowball Earth

There were three occasions when there was low latitude glaciation [ie glaciation in or near the tropics and maybe even on the equator]. The image below shows these periods:


Note that the continents were quite different way back then. The stars show locations where sea level glaciation has been detected. Note that the glaciations were quite wide spread and some is on the equator.

[This discussion is my attempt to make sense of Alley's brief explanation of the maps above and to add some more detail to satisfy myself that I have a reasonable understanding of the issues.

1. How were the locations of the sediments determined.

This can be explained with reference to the diagram below:

When sediments form they take on the magnetism of the Earth at the location of their formation. As you can see from the diagram above, the angle of the magnetic field will vary from 90⁰ at the poles to 0⁰ at the equator. If we assume that the location of the magnetic pole averages to the geographic pole over a sufficiently long period, the , then a formula can be developed to relate the angle of incidence of the magnetic field (imprinted on the rocks at their formation) to the latitude at which the rocks were laid down.

This formula is:
tan I = 2tan L,
where I is angle of incidence of magnetic field and L is latitude.

This can be solved for latitude (L) as follows:
tan I = 2tan L
tan L = (tan I) / 2
L = tan^-1(tan I / 2)

Here is an example assuming the angle of incidence = 30⁰, ie I = 30.
L = tan^-1(tan I / 2)
L = tan^-1(tan 30 / 2)
   = tan^-1(0.5774 / 2)
   = tan^-1(0.2887)
   = 16.1⁰

So an magnetic field incidence of 30⁰ corresponds to a latitude of 16.1⁰. Sadly I haven't worked out how they decide that the location is north or south of the equator.

2. How are the ages determined?
Not all rocks can be dated, so often approximate dates are assigned to a formation, between datable rocks above and below it.

This link gives information about dating some of the Snowball Earth periods.

3. How does the planet escape a Snowball

At first sight it seems unlikely that the planet could escape from a Snowball because of the Ice Albedo Effect. Snow and ice are highly reflective of sunlight, whereas soil and water (ie the sea) absorb rather than reflect sunlight. This fact makes a Snowball Earth possible. As the ice builds up more of the sun's energy is reflected into space (and less is absorbed). This provides a positive feedback, increasing the extent of the ice and resulting in less sunlight being absorbed by the Earth system which, increases the ice ... . Given the positive feedback, once a snowball develops, why isn't the Earth kept permanently in this icy condition? Or to put it another way, how does the planet break out of a Snowball?

As we know from our discussion of the Rock Weathering Thermostat, volcanos are constantly emitting CO₂. When the planet is in a snowball condition, there isn't any rock or ocean to take up the CO₂ (as the ocean and land are covered by ice). Consequently the CO₂ builds up in the atmosphere over a few million years and eventually reaches a such high level that the Greenhouse effect overcomes the Ice Albedo feedback and the planet warms, the ice starts to melt. Less ice and more soil, rock and sea means that the Earth starts absorbing more energy from the sun which reinforces the greenhouse warming.

When the ice has melted there is a great deal of rock and a very high level of CO₂ in the atmosphere, so the rock weathering process begins and gets into high gear. The CO₂ breaks down the rocks, the material for carbonate rock enters the ocean, and then precipitates. The resulting rock is called Cap Carbonate, because it sits on top of the glacial sediments. The cap carbonates can be seen in the photograph in the following section.

4. How do we know that glaciations occurred.

Glaciations leave deposits that can be recognised by geologists. The photo below was used by Alley in the talk. It shows Paul Hoffman, the chief proponent of the Snowball Earth hypothesis, pointing to the transition between the glacial sediments and the cap carbonate. The boulders (drop stones) are thought to have been dropped by melting icebergs. The lighter rocks above the transition are the cap carbonates (formed by the extreme greenhouse conditions that brought about the end of the snowball.



The photograph can be found at this link.

The photo graphically shows the rock weathering process. ]

Alley notes that "the Snowball Earths demonstrate that (the Rock Weathering Thermostat) really is working, with moderately high confidence", but not complete certainty as we don't have really good methods (Paleobarometers) for determining CO₂ as old as the Snowball Earths.

CO₂ paleobarometers

The gold standard for determining CO₂ level in the past is the ice core record. Drill the ice core, suck the CO₂ out and measure the concentration. There is one core that goes out to 800,000 years and many up to 450,000, which duplicate each other very well. We know this process works well for a number of reasons. If you take the youngest samples in places where it "snows like crazy", they match the instrumental record. If you compare the gasses in the firn (ice that is at an intermediate stage between snow and glacial ice) with the instrumental record you find a beautiful match. Go to different places in Antarctica, different snowfall rate, different temperatures, different impurity loadings, you get the same record. So we have high confidence that the ice core record really is good.

Furthermore the "failures" are "sensible" ie well explained. If you go to Greenland there are places were you get melt. And when the melt water trickles into the snow and refreezes it traps excess CO₂. And you find that excess CO₂ in the refrozen melt water, but "it isn't smeared out by diffusion, it isn't trickling down into other ices, it is right were it was". So the good ice is really good and the good CO₂ record from an ice core is the CO₂ record.

Older than the ice core record, it is much more difficult to determine the CO₂ record as there isn't anything that is directly writing down the CO₂ concentration. "The CO₂ concentration is controlling something and that something is being recorded in the geologic record." So there is some uncertainty, and it is unlikely that any one method will give a result that can be completely trusted. However, there are a variety of techniques, which have assumptions that are completely independent of each other. When the different techniques agree, we can have confidence that we actually are watching the CO₂ history of the planet.

The next slide shows the main techniques:



There are two stable isotopes of carbon: carbon - 12, often written as ¹²C, for details see this link, and carbon - 13 (¹³C). Carbon 12 has 6 protons and 6 neutrons in its nucleus, carbon 13 has 6 protons and 7 neutrons. ¹²C makes up about 98.89% of carbon in atmosphere and ¹³C about 1.1%. There are trace amounts of the radioactive iotope ¹⁴C.

The ratio of ¹³C to ¹²C is used in many ways in climate studies. As Alley explains plants will preferentially use the lighter isotope ¹²C rather than ¹³C, if there is plenty of CO₂. If there is not much CO₂ then plants will have to use ¹³C, "the heavy stuff" as well, "and so the ratio of heavy to light carbon in a plant is a measure of the availability of the CO₂ for that plant". Alley mentions three places where traces of plants that are useful for reconstructing ¹³C to ¹²C ratio: alkenones from the cell walls of plants, soil carbonates and liverworts. For an interesting video on alkenones see this link. (It turns out that alkenones can also be used to measure temperature as well as CO₂.)

The second method of measuring CO₂ in the past is by measuring the element Boron in the shells of foraminifera (usually abbreviated to forams) which are tiny creatures (usually less than 1 mm in size) that live in the oceans. The boron measurements relate to the Ph (acidity or basicity) of the ocean which in turn "listens to CO₂".

The third method involves stomata on the undersides of plant leaves. Plants "breath" through the stomata, taking in the CO₂ that they require for photosynthesis, but losing water out of the stomata. If there is a lot of CO₂ in the air plants reduce the number of stomata, but if there is limited CO₂ we would expect that plants increase stomata. These relationships are confirmed in fossil plant leaves. So the "number of stomata in fossil leaves is a tracer of CO₂".

The fourth method involves using models of geologic record and our understandings of the processes by which the Earth responds to different changes, such as increased vulcanism, mountain building and continental moving, to predict the level of CO₂. The leader of this effort was Robert Berner who described the development of :

a model for the evolution of the carbon cycle and of atmospheric CO2
over Phanerozoic time was presented based on inputs of geological, geochemical,
biological, and climatological data

Source document here.

Processes that changed CO₂ in the past

More CO₂ from volcanoes; if more shells of marine animals are fed into the volcanoes, through shallower subduction zones, then more CO₂ is output.

Slower rock weathering can mean that less CO₂ is taken from the atmosphere. Slower rock weathering can be caused by flat terrain with thick layers of soil making it difficult for CO₂ to react with the rocks. Faster rock weathering will take more CO₂ from the atmosphere. Faster rock weathering can occur if mountains are being built, which makes it easier for CO₂ to get at the rocks.

Changes in the biosphere can effect CO₂ level. Land plants make it possible to sequester CO₂ as coal.

These processes are long term, taking place over millions of years.

Atmospheric CO₂ and continental glaciation over the last 400 million years

The diagram below shows atmospheric CO₂ and continental glaciation over the last 400 million years.


The lines on the bottom half of the diagram are various measures of CO₂. The green shaded area is the CO₂ prediction from the Geocarb III model. "A lot of uncertainty but you will see certain patterns which are nicely preserved ... and a fair amount of agreement."

The blue areas hanging down from the top is how close to the equator ice was getting.

Note that the ice expands in the low areas of CO₂. "When CO₂ is low we have ice, when CO₂ is high we don't have ice."

Events - the great dying

The greatest mass extinction occurred 251 million years ago at the end of the Permian Era. About 95% of all species disappeared at this time. As a species can survive with relatively low numbers, it is likely that nearly all organisms on the planet died at this time. It appears that the photic zone of the oceans were full of hydrogen sulfide, H₂S, as there are wide spread biomarkers of green sulfur bacteria, which use H₂S for photosynthesis. H₂S is very toxic to oxygen breathing organisms. This occurs in a warm time caused by extensive vulcanism, the Siberian Traps, which was one of the largest volcanic episodes in the last 500 million years.

"And so you make it hot, and when it is hot it's hard to get much oxygen into the ocean, and you crank up CO₂ like crazy and you crank out rocks that the CO₂ can beat up, but the CO₂ is coming out so fast that you can't pull it down and you just fertilize the crap out of the ocean and the bugs grow, and the bugs die and they sink, and they use up the oxygen and then all heck goes. And this is far more than a CO₂ story and is far more than a warm story but when the ocean's cold its really hard to run it out of oxygen and when the ocean's warm it's a lot easier." There is a warmth that is attributed to CO₂ and that combined with some other things gives you a very interesting event.

For more details on the end Permian mass extinction, see this link.

Events - the saurian sauna

The saurian sauna refers to the mid cretaceous period. The cretaceous began 145.5 million years age and finished 65.5 million years ago, so the mid cretaceous is around 100 million years ago. It is hot, there is no ice near sea level near the poles, with forests crowding up to the edge of the Arctic Ocean. The continents are not that different to now. The only explanation for the heat at this time is that CO₂ is really high again, with the reason again high volcanism.

The diagram below shows some data on this period of time.



On the left is a temperature scale in the tropics, at 9 degrees north. Note tropical ocean temperatures in mid cretaceous were around 37 degrees. [Alley does not directly indicate current temperatures, but I think they are at most in the high 20s to low 30s.] Certainly 37⁰ C is a very high tropical ocean temperature. The right hand side shows CO₂ levels which were just below 1500 parts per million (ppm). [Current CO₂ is at 385 ppm and rising at a rate of about 1.9 ppm per year.] "High CO₂ high temperatures."

Events - Paleocene / Eocene thermal maximum (PETM)

The PETM can be seen in the diagram below, as the thin black line extending up from the main trend line, near the arrow. Temperatures increased at this time by about 5⁰C.



As can be seen from the diagram, temperatures were already very warm at this time - compare PETM temperatures with current temperatures at the far right.

The figure below provides evidence for the cause of the PETM.



The top line shows ¹³C level, and the bottom line shows temperature. Note that the temperature spiked just at the time that the 13C dropped dramatically. [This can be a little confusing if you forget that a significant decline in ¹³C means a very large increase in ¹²C.] The author of the article that the diagram is taken from notes that the drop in ¹³C is "thought to represent the influx of up to 2600 Giga tons (Gt - billion tons) of methane from dissociation of seafloor clathrate [which would have converted to CO₂ fairly quickly.] As Alley notes, "the CO₂ went up and it got warm."

Methane clathrates are described at this link.

The isotope excursion (CO₂ release) can also be seen on land as shown below from two different sources - the red line from soil, the blue line from tooth enamel.


The stress on the biosphere at this time is illustrated below. The diagram shows the extensive leaf damage to plants at the time, as Alley says "the leaf damage goes through the roof at this time... you have sort of messed with the ecosystem and something weird is happening". Other biosphere changes included plant and animal migrations, animal dwarfing and perhaps evolutionary innovations following "bottlenecks".



The extra CO₂ acidified the ocean as shown below, in cores drilled in ocean sediments.


[More acidic oceans make it difficult for shelly animals (eg forams) to build their carbonate shells.] The sea sediment cores grow younger from bottom to top. Note at the bottom the cores are white, indicating normal levels of carbonate [so a not particularly acidic ocean]. Then the cores go brown indicating a large decline in carbonate [and a more acidic ocean], and finally the carbonate returns at the top as the ocean returns to normal Ph levels and the forams can again build their shells effectively.

The time scale of this event is really interesting. The CO₂ and tempeature rose quickly in just a few thousand years, but took about 140,000 years to recover. This pattern is confirmed by Carbon Cycle models. [This is has frightening implications for our current situation as it shows that it takes a long time for the climate system to recover from an injection of a significant amount of CO₂ into the atmosphere.]

Changes in the understanding of data

Alley noted that when he started learning about these issues, as a student and young professor, "there were a bunch of places where we could point at and say 'Oh but there, there was a big global change and CO₂ didn't go with it'. And in the time I've been in this field watching, almost all of those disappeared."

One example is the Ordovician Glaciation. The Ordovician period lasted from 488.3 to 443.7 million years ago. Most of the time during the Ordovician temperatures were hot and CO₂ was high, but towards the end of the Ordovician there was a short glaciation, lasting about 1 million years. Early studies seemed to show that CO₂ levels remained high during the glaciation, "and then you refine the sampling and there is a drop in CO₂ at the glaciation."

[Here is a video that discusses the developments in understanding of the Ordovician Glaciation in much more detail than Alley does.



This video is taken from a series found at this link. The series of videos are very well worth watching - they provide an introduction to climate science and explode some of the more egregious myths put out by the so called "skeptics".]

For a long time there was a sense that during the Cretaceous Saurian Sauna the tropics weren't very warm, "the poles were but the tropics weren't." The tropical temperatures were determined by measurements of foram shells. Alley says that the reason why the tropics appeared to be cooler at this time is that "the shells had been changed and they weren't really recording the temperature" and that if you look carefully and find unchanged shells then they give a warm temperature. This type of change is well known and is called diagenetic alteration.

"Two years ago I would have said that something goes wrong in the Miocene; that 15 million years ago this doesn't work, and now I say 'well maybe it does'". Temperatures have been declining for the last 50 million years - see first diagram in PETM section.

The diagram below shows temperature (green line) and CO₂ measurements from 25 million years ago to 10 million years ago. Over the 15 million years ocean bottom temperature declined from about 5⁰C to 2.5⁰C and CO₂ declined from about 750 parts per million (ppm) to 320 ppm.



The open circles show ¹³C measurements of CO₂ and the triangles show ¹¹B measurements. Note that they do not match the temperature record well. Until recently these were the only measurements of CO₂ concentration at this time. The circles filled with yellow are fairly recent (2008) measurements and much more closely match the temperature measurement.

In October 2009, a group led by Aradhna Tripati released a new study. The diagram below is taken from this study, with CO₂ at the top and temperature at the bottom. Note that temperature closely matches CO₂ level.



[This diagram is of concern in relation to our current climate. The current (November 2009) CO₂ level is 386 parts per million - for details see this link. The average for the year so far is 387.8 (there is a regular yearly cycle.) The average annual increase over the last ten years is 1.9 ppm. It is clear that the planet will reach 400 ppm CO₂ during the next decade.

If you look at the CO₂ levels in the diagram above, you will notice that there is only one measurement above current levels (at about 15 million years), and it is only slightly higher.

As Tripati comments at this link:

The last time carbon dioxide levels were apparently as high as they are today — and were sustained at those levels — global temperatures were 5 to 10 degrees Fahrenheit higher than they are today, the sea level was approximately 75 to 120 feet higher than today, there was no permanent sea ice cap in the Arctic and very little ice on Antarctica and Greenland.

Aradhna Tripati,is a UCLA assistant professor in the department of Earth and space sciences and the department of atmospheric and oceanic sciences.]

Alley finishes this section by noting that all of the measurements from the two previous diagrams are good science, "so we're not sure; there are still some uncertainties floating around in there. But two years ago I would have said 'Wow, we've got a big global temperature change and it's not CO₂', but with more data it sort of tracks the CO₂. And so the anamolies are disappearing fairly rapidly. It is very difficult now to point to a big temperature change without a CO₂ change accompanying it. The Tripati group's indicator tracks the ice cores very well, a point also made in the link above.

The ice record in the last 800,000 years

There are very good records of temperature and CO₂ in the ice cores. The diagram below shows the record out to 400,000 years. This is well duplicated. There is one record out to 800,000.



Blue is temperature, red is CO₂. They match pretty well, and we do know that sometimes temperature leads CO₂ by a few centuries. Remember the email at the beginning of the presentation that claimed that CO₂ lagging temperature in some of the ice core record "proved" that the current warming was not caused by CO₂. Alley now explains why this argument is false.

The timing of glaciations is determined by changes in the Earth's orbit. These are called Milankovitch cycles, for details see link here. The cycles change the distribution of heat from the sun. The crucial location is about 65⁰ N in summer. Cooler summers in the far north mean that ice and snow survive over the summer and ice sheets expand. Warmer northern summers melt the ice and snow and the ice sheets retreat. Ice is highly reflective so the smaller amount of ice reflects less of the sun's energy and land and sea uncovered by the retreating ice absorb more solar energy, so the planet takes in more of the sun's energy. Consequently the planet warms, and the ice retreats further, leading to more energy intake and more warming and so on.

As the Earth warms, CO₂ builds up in the atmosphere, largely from the warming oceans, thought it seems that it takes 6 to 8 centuries for significant amounts of CO₂ to appear. The CO₂ then amplifies the warming by greenhouse forcing. Calculations and model runs demonstrate that about half of the warming in a deglaciation comes from the CO₂.

When the science is actually understood it is not at all a surprise that CO₂ appears after the warming during a deglaciation, but that does not mean that it does not significantly contribute to the warming after the first 6 to 8 centuries. The total warming takes at least 5,000 years.

Unlike during a deglaciation the current warming is initiated by increased CO₂. A proper understanding of the deglaciation process is a cause of concern as it indicates that the current warming, that was initiated by increased CO₂ is likely to be made worse by releases of CO₂ from the natural environment, caused by the warming.

Here is a video discussing these issues. The video is taken from this site, which is well worth visiting as there are many more interesting videos there.




Other possible climate forcings

CO₂ is "necessary and sufficient to explain most or all of the big temperature changes of the globe."

Is there anything else that we should be worried about?

A whole lot of the competing hypotheses don't work.

* Meteorites - a meteorite did kill the dinasaurs but there aren't many big meteorites in the record

* Volcanos - a big volcano does cause cooling, "and if they could get organised they would rule the world ... but they are not very organised ... there is no good way today for a volcano in Indonesia to tell a volcano Alaska, its time to erupt."

* Sun - "as far back as we can see well, the sun does not change much, if the sun changed a lot it would really control things hugely, but it only changes really slowly or really small as far as we can tell. Records are not as good as we would like, there is work to be done here, but it doesn't seem to be doing much."

* Cosmic rays - some people say "the sun doesn't change much, but, the sun modulates the cosmic rays, the cosmic rays modulate the clouds, the clouds modulate the temperature, so the sun is amplified hugely. Its a really interesting hypothesis, there is really good science to be done on this, but we have reason to think its a fine tuning knob." as shown by the diagram below.

Cosmic rays make ¹⁰Be (Berillium). The top graph is for temperature the bottom one for ¹⁰Be (the cosmic ray proxy). The sun modulates cosmic rays, so does the Earth's magnetic field. Forty thousand years age the magnetic field basically zeroed out for a millenium or so (the Laschamp Anamoly). This resulted in a large increase in the cosmic ray flux, which can be seen in the diagram below, indicated by the orange arrow. "We had a big cosmic ray signal, and the climate ignored it." So cosmic rays are a "fine tuning knob at best."



[There are other good reasons to doubt the cosmic ray theory as indicated at this post.]

Climate sensitivity to CO₂ from the paleo-history

CO₂ sensitivity is measured as the amount of temperature increase for a doubling of CO₂.

[The IPCC gives a range from 2⁰C to 4.5⁰C, with 3⁰C being the most likely value.

The following is taken from the Summary for Policy Makers AR4 (from the 2007 report):

The equilibrium climate sensitivity is a measure of the climate system response to sustained radiative forcing. It is not a projection but is defined as the global average surface warming following a doubling of carbon dioxide concentrations. It is likely to be in the range 2°C to 4.5°C with a best estimate of about 3°C, and is very unlikely to be less than 1.5°C. Values substantially higher than 4.5°C cannot be excluded, but agreement of models with observations is not as good for those values. Water vapour changes represent the largest feedback affecting climate sensitivity.

]

Glacial periods, PETM and Cretaceous warm poles indicate that climate sensitivity might be higher than the accepted value (3⁰C for doubling of CO₂).

"The models do pretty well when you compare them to the past, which means that ... the mid range isn't bad. There is cheating on the ice age cycle as you have to get half the cooling from the ice sheets being reflective, you have to put the ice sheets in, but you really need the CO₂ to get the ice sheets. And so the sensitivity of the CO₂ on the glacial / intergalcial cycling is probably a bit higher than the models have in."

Using models with the usual sensitivity (3⁰C for doubling CO₂) it is difficult to get the PETM warm enough and the poles warm enough during the Cretaceous, indicating that acutal climate sensitivity might be higher than the accepted value.

"We have this themostat of rock weathering, if CO₂ didn't make it warmer then when there's more volcanoes the CO₂ would just stay in the air and we'd have indications of huge amounts of CO₂." If CO₂ did make it significantly warmer then the rock weathering process would start up quickly and draw down the CO₂, so the level would not change much.


The diagram is the history of CO₂. The dashed line shows the actual levels of CO₂ over time, as reconstructed by proxies. The green curve is the history of CO₂ if climate sensitivity was huge, 6⁰C for double CO₂. The actual levels were much more variable. The red curve is the CO₂ history if sensitivity was only 1.5⁰C. The very high levels indicated just did not occur. The blue curve is the best fit for the proxies, and represents 2.8⁰C, which is very close to the IPCC figure.

"You could find somewhere in the literature people saying you get less than 1⁰C for doubling CO₂ and [others saying] you might get up to 12⁰C, paleoclimate says 'no, cut off the ends'."

So where does that leave us?

* If CO₂ warms Earth's climate then history makes sense, with CO₂ having caused or amplified the main changes and at present there is no plausible alternative to this.

* If CO₂ does not warm we have to explain how radiation physicists are so wrong and how a lot of (otherwise) inexplicable climate events happened over Earth's history.

* Higher CO₂ might be a forcing or a feedback, a CO₂ molecule in the air is radiatively active regardless of how it got there. [A forcing is something outside of the climate system that pushes (forces) it in a particular direction. These include, changes in the sun's luminosity, a large volcanic eruption, and of cause a large increase in CO₂ like our current one or the PETM. A feedback is a change within the climate system in response to a forcing, that effects the change produced by the forcing - it can increase the effect, positive feedback, or decrease the effect, negative feedback. The ice albedo effect and the increase in CO₂) as a result to the initial warming in a deglaciation are examples. Another interesting example is the water vapour feedback. An increase in temperature (whether caused by the sun or increased CO₂ increases the amount of water vapour in the atmosphere, and as water vapour is a greenhouse gas this amplifies the effect of the initial warming. An doubling of CO₂ in the atmosphere will only lead to an increase of 1.2⁰C, the water vapour feedback amplifies this up to about 3⁰C.]

* Paleoclimatic data show show climate sensitivity similar values in modern models (3⁰C warming for doubling CO₂) perhaps with somewhat higher values over centuries or millenia especially in polar regions.

* Lots of knobs control the Earth's climate system.

* The "Sun" knob isn't twiddled very much over short times, and hasn't done very much over long times because of CO₂.

* If cosmic rays, space dust, magnetic field and other "space" knobs matter, available evidence indicates that they do no more that fine-tune, and even that is not demonstrated.

* Lots of things on Earth matter regionally - moving a continent from equator to poles cools it - but evidence is weak for major control of climate, except for CO₂. "... there are lots of things in regional climate that don't do much to the globe ... in terms of the things that people care about CO₂ is just a start not the end." There is a lot more work to be done on regional influences.

"Where we really stand now we are not quite yet at the pound on the table, this is nailed and this is our confidence level. The paleoclimate data are coming in really fast, they're real good ... but these latest advances have not had time to percolate through to the IPCC yet, ... we are going to see more discusion on this; this story is very clearly not done, but it is fairly clear where we stand now which would be that an increasing body of science indicates that CO₂ has been the most important controller of Earth's climate.