Arguments
CO2 lags temperature - what does it mean?
What the science says...
| Select a level... |
Basic
|
Intermediate
| |||
|
CO2 didn't initiate warming from past ice ages but it did amplify the warming. In fact, about 90% of the global warming followed the CO2 increase. |
|||||
Climate Myth...
CO2 lags temperature
"An article in Science magazine illustrated that a rise in carbon dioxide did not precede a rise in temperatures, but actually lagged behind temperature rises by 200 to 1000 years. A rise in carbon dioxide levels could not have caused a rise in temperature if it followed the temperature." (Joe Barton)
Earth’s climate has varied widely over its history, from ice ages characterised by large ice sheets covering many land areas, to warm periods with no ice at the poles. Several factors have affected past climate change, including solar variability, volcanic activity and changes in the composition of the atmosphere. Data from Antarctic ice cores reveals an interesting story for the past 400,000 years. During this period, CO2 and temperatures are closely correlated, which means they rise and fall together. However, based on Antarctic ice core data, changes in CO2 follow changes in temperatures by about 600 to 1000 years, as illustrated in Figure 1 below. This has led some to conclude that CO2 simply cannot be responsible for current global warming.
Figure 1: Vostok ice core records for carbon dioxide concentration and temperature change.
This statement does not tell the whole story. The initial changes in temperature during this period are explained by changes in the Earth’s orbit around the sun, which affects the amount of seasonal sunlight reaching the Earth’s surface. In the case of warming, the lag between temperature and CO2 is explained as follows: as ocean temperatures rise, oceans release CO2 into the atmosphere. In turn, this release amplifies the warming trend, leading to yet more CO2 being released. In other words, increasing CO2 levels become both the cause and effect of further warming. This positive feedback is necessary to trigger the shifts between glacials and interglacials as the effect of orbital changes is too weak to cause such variation. Additional positive feedbacks which play an important role in this process include other greenhouse gases, and changes in ice sheet cover and vegetation patterns.
A 2012 study by Shakun et al. looked at temperature changes 20,000 years ago (the last glacial-interglacial transition) from around the world and added more detail to our understanding of the CO2-temperature change relationship. They found that:
- The Earth's orbital cycles triggered warming in the Arctic approximately 19,000 years ago, causing large amounts of ice to melt, flooding the oceans with fresh water.
- This influx of fresh water then disrupted ocean current circulation, in turn causing a seesawing of heat between the hemispheres.
- The Southern Hemisphere and its oceans warmed first, starting about 18,000 years ago. As the Southern Ocean warms, the solubility of CO2 in water falls. This causes the oceans to give up more CO2, releasing it into the atmosphere.
While the orbital cycles triggered the initial warming, overall, more than 90% of the glacial-interglacial warming occured after that atmospheric CO2 increase (Figure 2).
Figure 2: Average global temperature (blue), Antarctic temperature (red), and atmospheric CO2 concentration (yellow dots). Source.
Basic rebuttal written by dana1981
Update July 2015:
Here is a related lecture-video from Denial101x - Making Sense of Climate Science Denial
Last updated on 8 January 2016 by pattimer. View Archives
If you want to argue that the temperature requires CO2 to elevate which causes further temperature rise and that solar changes are NOT the primary cause of temperature changes, then you really need to explain convincingly what causes the temperature and CO2 levels to STOP their rises at the entry of interglacial periods. If you claim it is just a minor variation in solar changes, then it detracts from your overall argument for the forcing relationship between CO2 and temperature. If the start of the interglacials show anything, it is that once the CO2 reaches a high enough level that it no longer causes any additional rise in temperature. In fact, it would seem to indicate that once CO2 reaches a certain level it actually helps support the reduction of temperature over time.
Low level cloud formation is much more correlated to temperature rises than high level cloud formation. GCR is a much better predictor of low level clouds than CO2 levels are.
The researchers who have looked at past behavior of the climate have built a pretty strong case to show how those sequences emerged, how an initial rise in temperature due to solar variations triggered a feedback amplifying that rise. The exact details may still be in play, but at the end of the day it's important to remember, those events happened in the dim past and were triggered and driven by processes different than today.
The burp of carbon we're eructating today actually analogizes better with other episodes better characterized as catastrophic in nature, such as periodic eruptions of flood basalt in Eastern Washington.
John quoted two sentences where Barton is discussing the lag between temperature and CO2 in the ice core records. Barton suggests that the existence of this lag shows that CO2 did not cause the glacial/interglacial changes in temperature ("a rise in carbon dioxide levels could not have caused a rise in temperature if it followed the temperature").
The article at the top of this thread rightly explains that Barton's point is misleading, because in the period Barton is referring to CO2 functioned as a feedback whereas now it's acting as a forcing (more on this in the next comment).
You suggested that there was some context being left out, specifically the following:
* A preceding sentence and footnote (iv) discussing higher CO2 levels in the Eocene and Oligocene. This is tens of millions of years ago, i.e. two orders of magnitude further back in time than the glacial/interglacial cycles that are the subject of this thread. There are no ice cores that go back that far, and no evidence for any lag. Higher levels of CO2 in the Eocene and Oligocene don't somehow change the wrongness of Barton's discussion of a lag in Pleistocene ice core data. It's a completely different subject.
* A footnote (v) to a paper about the lag in the ice cores, and a concluding sentence with an appeal to authority (citing somebody from NAS who Barton says mentioned the lag in congressional testimony). These, likewise, are completely irrelevant -- all they do is provide support for Barton's claim that there was a lag in the ice core records. But nobody thinks that there wasn't a lag in the ice core records. If anybody is making a straw-man argument here, it's Barton himself!
So no, I don't see anything in the extended version of the Barton quote that even remotely suggests he is being misrepresented. You will need to be much more specific if you still think there's a problem somewhere there.
During the Pleistocene glacial/interglacial cycles, the primary forcing of temperature changes was variations in the seasonal and latitudinal distribution of insolation caused by variations in the Earth's orbital geometry (Milankovich cycles). Everyone here knows this. Without these spatial-temporal variations in insolation, there wouldn't have been swings in temperature.
However, those changes in temperature were amplified by various positive feedbacks (e.g., water vapor, CO2, and ice albedo). Without those feedbacks, the magnitude of the temperature swings would have been much smaller.
One could logically ask, if there are such positive feedbacks in the climate system, why did the temperature stop rising (or falling)? (This is the meaning of your "STOP" sentence, I believe.) There are at least two answers:
(1) A positive feedback does not imply unlimited increase, as long as the feedback coefficient is between 0 and 1. Many people (on both sides of this argument) don't understand this point, and assume that a positive feedback in the climate system has to lead to either runaway warming (like on Venus) or to a frozen snowball Earth. That's incorrect, though.
(2) Most importantly, the Milankovich cycles are cyclical. They alternately provide a warming forcing followed by a cooling. As soon as the direction of this forcing reversed (say, from warming to cooling), the CO2 and other feedbacks would likewise reverse (with the usual time lag, of course).
So there's nothing that needs to be explained about the fact that the temperature stopped rising during the interglacials (or why it stopped falling during the glacials, for that matter).
How is this relevant to the current situation? See the next comment ...
So, what's the relevance of the time-lag in ice cores to our current situation? Not much.
In the Pleistocene, CO2 was a feedback amplifying warming/cooling caused by the Milankovich cycle. Like many feedbacks, it took a while to kick in, so there was a time lag.
Today, in contrast, we're directly adding CO2 to the atmosphere. It's a forcing in its own right, not just a feedback (though there are additional CO2 feedbacks).
So the time lag in the ice cores is irrelevant to the current situation. Likewise, the fact that temperature stopped increasing during previous interglacials doesn't mean that temperature won't keep rising if we keep burning fossil fuels today. It stopped rising then because the M. cycle changed. But the M. cycle doesn't have any effect on the decade-to-century time scale we're dealing with now.
Hopefully this clears up some of the confusion.
You said yourself that you thought Barton's "first and last sentence are about something else entirely".
If you assume for a moment that Barton is not irrational and actually was capable of making a rational paragraph, then it needs to start with a reach if the sentences of a paragraph can be logically related and he is not just spewing random sentences.
You have inappropriately narrowed the interpretation of the scope of his second and third sentences to global warming of today- there is nothing to justify your assertion.
The CO2 produced after a warming period cannot be the cause of that particular warming period. If the argument is that additional CO2 produced necessary causes a warming period and that variability in solar influences are much less important, then the CO2 warming proponents need a better explanation of the dramatic trend change at the typical start of an interglacial.
Clearly, the CO2 levels 130 Kyrs ago were not high enough to prevent the dramatic temperature trend shift at the start of the preceding interglacial.
I fully understand that CO2 levels are higher today and are trending higher than they have been in millions of years. But the Earth and life on it, have survived with many times more CO2 than we currently have.
Science requires models that are accurate predictors of controlled experiments or future events.
What prediction of Global warming science are you most proud of? How many years into the future was that prediction?
CO2 scientists tend to very much underestimate the importance of solar emission variability changes to the Earth orbit.
If CO2 were a warming forcing function, then the model should predict a plateau at the start of an interglacial instead of the typical sharp peaks.
Is there any literature that demonstrates how the formulas from today's global warming computer models can be applied to the previous 3 interglacials?
Thanks.
You go on to say that science is at its best when it is accurately predictive and can be tested to complete satisfaction. True enough, but in this case we do not have a laboratory large enough and with the correct features to run controlled experiments on a global climate. More, there are so many variables at play here that I don't think the sort of controlled experiments you have in mind are even possible.
What we can instead look to is how the burgeoning collection of research results we have from multiple disciplines fit together. Do we see observations that appear to be mutually exclusive, that cannot be explained in the presence of other observations? No. The ultimate metric of validity, the requirement for self-consistency, is satisfied despite the vast gulf between many of the avenues of inquiry related to climate change.
What we have is a synthesis of multiple lines of inquiry, each yielding predictions and observations that taken together describe the gross features of our climate, features on the scale we're concerned with. For all the criticism leveled at it, that is what the IPCC report is, a synthesis, and the message of that synthesis is robust against the very most stringent criticism. Again, self-consistency-- a key metric of scientific validity-- is satisfied.
Now, policy makers also do not have the luxury of being able to run experiments of a global scale in laboratories. They must operate in the day-to-day world of human affairs. Multiple lines of mutually consistent scientific inquiry tell us our human affairs are modifying the climate. Human affairs are going to need some adjustments. At a certain point we need to come to grips with the self-consistent message multiple disciplines have delivered to us and make some changes in our habits.
Science, by definition, has to be predictive otherwise it is not science. The middle ages was full of charts that would be constantly updated to seem predictive but never took into account the theory of gravity. These tables that the "scientists" of the middle ages were made to exhibit "self-consistency". Self-consistency is a necessary but not a sufficient condition to determine true science. Prediction while measuring other potential factors is typically necessary to be considered a hard science.
Again I ask- what prediction of global warming science are you most proud of?
The graph presented with this article covers half a million years. Barton's paragraph referred to even greater time scales. His assertions were correct for the timeframes he covers.
Ned claims that previous points in history are not instructive to the present time at the present scale. And that new "science", that does require long term testable data, is not needed: all that is required is ever changing self-consistent models...
Just like what we had in the middle ages.
If this article, instead of showing half a million years, actually pointed to data sets from history that shows a substantial burst of CO2 (with similar levels of dust that are being produced today) and show that there was clearly a jump in temperatures that could not be explained by solar variation then I would take notice.
What is the best unchanged global warming model that is over ten years old? What did it predict correctly?
Climate is much more complicated than today's models describe and there is no historical data presented here that CO2 is a forcing function to global climate- The next 50 years will dramatically change our knowledge of climate and how humans can control it to make it more hospitable for any species humans choose. I expect the importance of ionization, solar variability, and magnetic fields will bring the actual models of decadal climate prediction to the level of today's weather prediction.
I'll leave my faith in the physicists - they do not seem to be as steeped in political biases.
An overemphasis on things that are observable for very long periods of time have fooled scientists in the past and are destined to continue to do so.
Your last paragraph was clearly regarding policy and so I will not respond to it as that it has been deemed inappropriate to discuss policy here.
The sort of large scale experimentation you insist is required to understand the phenomenon and make successful first order predictions is actually not necessary. Exactly what knock-on effects will transpire from the changes to the physical behavior of the ocean-atmosphere system we're launching by altering the radiation characteristics of the atmosphere are as yet emergent. But again, the gross effects of the work we're performing on our atmosphere are not actually controversial.
If I've not already done so, I suggest you spend some time reading an introduction to the topic of C02's characteristics as they pertain to radiative physics and our atmosphere. You can hardly do better for a primer on this subject than Weart:
The Carbon Dioxide Greenhouse Effect
And pardon me for mentioning policy. My concern is that on the one hand we do have an admirable desire for perfectionism when it comes to creating hermetically airtight science, yet on the other hand pragmatism suggests that at some point we must accept that multiple arrows of hypothesis leading to theory and then conclusion tell us we've got a need to take the conclusions we're reaching and act on them.
No, Barton's second and third sentences aren't about "today", they're about the EPICA and Vostok ice cores, covering the past half-million years prior to the start of the current interglacial. See where Barton refers to a "lag"? That's the data set and time period with the lag. That's what Barton is talking about, and that's what this thread is about.
Continuing, nhthinker writes: The CO2 produced after a warming period cannot be the cause of that particular warming period. If the argument is that additional CO2 produced necessary causes a warming period and that variability in solar influences are much less important, then the CO2 warming proponents need a better explanation of the dramatic trend change at the typical start of an interglacial.
I discussed this in detail in my comment above. There's not really any problem here, this is pretty well understood.
Then, nhthinker writes: But the Earth and life on it, have survived with many times more CO2 than we currently have.
Nobody's suggesting that the Earth will not survive, or that life will not survive. The question is what will be the economic and social impacts of rapid climate change, especially things like increased drought in continental interiors (a la 1930s Dust Bowl).
Science requires models that are accurate predictors of controlled experiments or future events.
What prediction of Global warming science are you most proud of? How many years into the future was that prediction?
This is actually a good question, and one that might make an interesting topic for a full post here.
There have been a number of good predictions. Since this thread is about the temperature/CO2 lag in Antarctic ice cores, here's a nifty example of a test of the predictive power of climate models.
The EPICA Challenge to the Earth System Modeling Community
The basic idea is that in June 2004 the EPICA team released the analysis of climate data from 740,000 years of the Dome C ice core, but they only released the greenhouse gas data for the most recent 430,000 years. They invited modeling groups to make predictions about what the unreleased 300,000+ years of CO2 data would look like, as a way of testing the models.
The model predictions are compared here:
Modeling Past Atmospheric CO2: Results of a Challenge.
Most or all of the groups produced pretty accurate predictions (see Figure 4 here for an example of a comparison of one group's predictions vs. the actual ice core CO2 data).
So yes, there are lots of predictions, and in fact we can even find some nice examples which are actually on topic for this thread!
Tricky- because there is a lack of natural causes that can suddenly release CO2 as a forcing not a feedback. I'm sorry but to me this like refusing to acknowledge that we could send a rocket to mars because there wasnt a "natural" trip by an object from earth to mars. We can safely predict the course within known uncertainties however because we just use physics. So do climate models.
What predictions from climate model do you think are poor? That they cant predict weather? They dont try to.
To Ned's list, I would actually add the Hansen 1988 for actual forcings. Sure it has some issues but good effort for such a primitive model! (For more on this and current model/data comparisons see
updates-to-model-data-comparisons
Climate science has also predicted pretty accurately both the change in OLR from AGW gases and the increased LR energy received at surface from these gases.
In recent years, perhaps. But some mass-extinction events involved sudden increases in CO2 (PETM, various flood basalt episodes).
The emphasis of this article is that the half million year data showed carbon following temperature. The point of the new climate models is to show CO2 as a forcing function.
Your reference to the preceding years 300 Kyrs data is only useful if it also shows a forcing function of CO2.
I would be much more interested in analysis of periods of history that show substantial emissions of CO2- but I'm afraid that these CO2 emissions are also correlated with major emissions of soot and other particulates from volcanoes. My minimal understanding of these events is that they quickly lead to a deep cooling periods as opposed to long warming periods. I think climate scientists need to search out and highlight these historic events that are most like today rather than just saying today is different and is producing different results that can be predicted just based on computer models, laboratory experiments and short term regional analysis.
How about looking at this from a slightly different perspective. What do you see happening now-- in the history of the past few years that is to say-- that controverts the underpinnings of our understanding of how C02 behaves as a component of our climate?
nhthinker - aerosols lead to cooling but only for brief periods of time because aerosols dont remain in the atmosphere. CO2 does.
"The point of the new climate models is to show CO2 as a forcing function."
Huh???? This would be news the modellers. CO2 IS a forcing. We can measure its forcing (3-4 W/m2) more or less directly even. See Evans 2006 The point of models is to see what happens with this forcing.
Perhaps some careful reading of the science in
its not us
You still seem to be not reading (or not grasping) the point. In the ice core record changes in CO2 may not be starting the warming/cooling swings, but changes in CO2 are most definitely amplifying the warming/cooling swings. In other words, adding CO2 to the Pleistocene atmosphere did produce more warming, and removing it from the atmosphere did produce more cooling.
Now, can you explain some reason why CO2 would act as a greenhouse gas in the laboratory, and in numerical models, and in the ice core records, and in many other times during Earth's long history, and on Venus ... but would mysteriously stop being a greenhouse gas once Homo sapiens began constructing coal-fired power plants?
If you add a forcing function with a time delay to a base cyclic pattern or even a pulse function, it produces a curve that is completely different than what the last 500 Kyrs of data demonstrates.
The current climate models are filled with fudge factors that are constantly being updated- And updated at a much faster rate than the planet position charts were prior to the invention of the mathematical model for gravity.
I believe in Occam's razor- the complexity of today's climate models are very much analogous to the planet charts of the middle ages.
Relatively simplified mathematical models of a reasonable number of variables should be all that is needed to explain the natural record without force fitting very odd shaped forcing functions.
That is what hard science strives for.
Water vapor is a dramatically more important GHG than CO2 is and it has only seriously been studied extremely recently- CO2 data has been around much longer. Much of what was thought as extensible to environments that have not been monitored will be proven wrong.
"ScienceDaily (Apr. 29, 2008) — MIT Professor Dara Entekhabi will lead the science team designing a NASA satellite mission to make global soil moisture and freeze/thaw measurements, data essential to the accuracy of weather forecasts and predictions of global carbon cycle and climate. NASA announced recently that the Soil Moisture Active-Passive mission (SMAP) is scheduled to launch December 2012."
Climate scientists are only now starting to get the tools that they need to make real progress. But in some ways, there is too much ego and passion involved in quarters on both sides of this debate.
Point me toward a solution that passes the Occam's razor test. I do have an open mind. If you point me to a solution that seems overly complex to explain the data, then I expect it needs to provide accurate prediction to be a useful model. Very complex models that were intentionally force fit to previous data are not very compelling.
But water vapor is a feedback, not a forcing, as explained in this Skeptical Science argument: Water vapor is the most powerful greenhouse gas. If you want to discuss that topic further, that thread is the appropriate place.
For more information, see the RealClimate post FAQ on Climate Models, Part I and Part II.
If you want to argue about that topic, those RealClimate posts would be appropriate places to do so, but unfortunately it looks like commenting is closed on those posts. I can't think of a Skeptical Science post that is appropriate for that topic, and off-topic comments often get deleted.
First some deep background:
Simple Models of Climate Change
Next, more recent developments leading more or less to the present:
General Circulation Models of Climate
Reading Tom's RC links as well as Weart's more pedagogical work will improve your ability to discuss this topic without committing embarrassing blunders. Elsewhere I've blithely mentioned spending "a few hours" reading Weart but in fact he's produced a synopsis of many decades and many careers of work just covering the effort and dedication devoted to models so in all honesty it does take a little determination to wade through the story. If you're interested in discussing climate models and climate change while staying within the boundaries of established history and uncontroversial facts Weart's narrative is well worth the time.
useful science requires predictive ability:
What significant prediction of current or past climate science with CO2 as a forcing function have you not been embarrassed by?
How far into the future was that prediction that came true? How long do we have to wait to know if your outstanding predictions are accurate?
As far as science goes, you guys are still very much in your infancy at predicting global climate changes or even trends.
What bets have you made on your predictions? What odds are you willing to give?
There are so many basic unanswered questions about climate that get papered over by climate scientists... Why are the climate patterns of the last 500 Kyrs significantly different from the 1500 Kyrs prior to it? Climate tends to have cyclic patterns to it but also gets major non-cyclic perturbations to it. What will be the next perturbations that will impact climate trends? I mean besides the unpredicted changes in solar emissions of the last few years.
What "embarrassing predictions?" Cites to these embarrassing papers please. Climate model predict 30 year trends as surely you are aware.
What climate science certainly does believe is that climate is a physical phenomenon governed by physical laws. Change requires forcing. If you are claiming some natural change, then lets see that natural forcing. Paleoclimate has countless riddles but perhaps you would like to consider how difficult it is to measure global solar, albedo, aerosol, and atmospheric composition from the geological record.
I just read this history of climate models:
http://www.aip.org/history/climate/GCM.htm
It seems that at least some effects are still not really based upon a fundamental understanding of underlying physics. The effects of clouds are still apparently used as fitting parameters to climate data. The fits to climate data are then used to predict climate over other periods. I don't really have a problem with this in principle, but it does seem that these are not really fully based on fundamental physics and this type of fitting leaves open the possibility of trying to use the fitted parameters outside the region of validity (extrapolation rather than interpolation). Apparently things like clouds are not really understood in enough detail to truly predict climate from fundamental physics.
Of course, this is just the impression I got from one source, perhaps it is not really accurate.
Models are unreliable
"The paucity of observations in the tropical Atlantic and Indian oceans have considerably retarded our understanding, modeling, and prediction of these coupled modes, which are extremely important not only because of their societal consequences but because it is through them that the ocean actually drives the atmosphere. These regions and coupled mechanisms should constitute a priority of observational and theoretical research. "
The confidence of CO2 forcing to the levels that climate models currently predict is tenuous, at best. Indeed, the understanding is still retarded. But it is improving all the time. Can you point to a researcher that has not put out predictions that did not turn out to be serious exaggerations of actually measured impacts?
During the glacial/interglacial cycles, CO2 acted as a feedback amplifying the warming or cooling caused by the Milankovich forcing. In other words, an external source of cooling would reduce the global mean temperature. The oceans would respond by absorbing more CO2 from the atmosphere, causing a feedback that would amplify the cooling a bit. Eventually, when the Milankovich cycle shifted towards warming, this process would reverse, and the increasing CO2 in the atmosphere would amplify the initial warming.
The situation today is different -- CO2 is not just a feedback, we're now adding it directly to the atmosphere in large quantities.
Does that help clarify things?
If temperatures stopped rising, and dropped steeply many times in the past, why should I believe that they can go higher now? You had rapidly rising CO2 in the past, but the temp. rise was unable to continue, and quickly dropped back. Why?
CO2 levels are higher now (by more than 35%) than at any time in the last 600,000 years. The last time they were this high (with similar solar output), temperatures were almost 5C warmer, IIRC.
So, in fact, temps *can* go higher if CO2 levels are pushed higher. That's the essence of AGW theory.
Why did an obvious violent feedback mechanism suddenly stop, and go into equally violent reverse? Until you can answer that, it's surely unjustified to claim that the warming that's happening today won't meet the same "barrier" that warming did in the past.
"...it's surely unjustified to claim that the warming that's happening today won't meet the same "barrier" that warming did in the past. "
The present warming's not going to meet the same barrier, that barrier's efficacy won't be the same today because we've created a novel situation.
"Past is prologue" arguments always face the fundamental problem that we're not living in the past, we're living in the present, us and our effluent.
The information you are looking for is available on this site. I can't post direct links, since this is being tapped out from my phone. Read up on the Milankovitch cycles and climate sensitivity in regards to CO2 as a feedback agent.
Currently, CO2 is the forcing agent rather than a feedback. This is why archiesteel's comment was spot on.
I think about the four peaks in temp. in the graph above. Just before that peak, temp was rising at an enormous rate, and so was CO2. Something stopped it suddenly, and sent it the other way, EVEN THOUGH CO2 WOULD CONTINUE RISING STEEPLY FOR ANOTHER 800 YEARS !!
According to the models, 800 years of steeply rising co2 should cause at least 750 years of more temperature rises, not a steep plunge.
On point 1) You were told to read up on Milankovitch cycles. Milankovitch cycles are periodic shifts in the earths orbit. When the orbit moves closer to the sun we get the warming trend, when the orbit moves further from the sun the warming trend is reversed and we get a cooling trend. The orbital cycles drive the direction of temperature change while CO2 only acts as a feedback or amplifier of said changes. Get it?
On point 2) You asked how current theory explains the temperature trends. You can't very well expect someone to explain current theory without using the theory itself. Also, Milankovitch cycles on their own have nothing to with CO2, CO2 just acts as a feedback.
If your question is what evidence do we have that CO2 levels are causing GW, then that is a separate topic and is covered here.
And nothing in any Milankovitch cycle that could suddenly negate 800 years of steeply rising co2 levels, if the models are correct.
On a side note, it's not an "assumption" that increasing concentrations of CO2 cause the atmosphere to be a more efficient insulator. If you want to argue about that, however, use the search function to find a suitable venue for chasing that topic.
Amplifying e's comment about your point 2: The main evidence for CO2's warming effect is not passive observation of its historic relationship to temperature. Rather, it is fundamental physics.
CO2's warming effect was postulated in the early 1800s, long before there was evidence of its historic relationship to global temperatures. The exact mechanism by which CO2 does that was discovered in the early 1900s, again before there was much historic evidence. Later the historic relationships were discovered and found to match those predictions. For a good summary of the physics see Ramanathan's paper. For a good history see Weart's The Discovery of Global Warming.
Yes, which is why feedbacks are required to account for the degree of temperature swings. The models accurately recreate these temperature swings, so your claim that the models are inconsistent with the models is wrong and nonsensical.
Multiple, interrelated, not entirely synchronous factors were responsible for both the onset and offset of ice ages. They are not understood completely, to say the least. But just one example of how those factors can have sudden onsets and offsets is the recently discovered CO2 "burp" from the deep ocean.
The problem I'm pointing out with the graphs above, is that at the peak of previous warming, you get an 800 year period, when co2 levels are high, and rising steeply, and yet temperatures are plummeting.
This is in stark contrast to todays models.
And our point in the cycle now is at or near that historic maximum.
I don't see any explanation why the models are so wrong, for so long, for these events.
The Milankovitch cycles just don't cut it. The insolation curves are extremely gentle.