Now it's 2°C climate change target 'not safe'

Rob Potter says:
And the whole 6-9 m of sea level rise will take just 100 years……. Wow, the resolution of their historical reconstruction is fabulous!
That would require 6-9cm per year. Or 60-90 mm/year, 600-900mm/decade…
I’m thinking 6-9 mm per decade is what they meant, based on a 6-9 m rise over a millennium (which they might just be able to resolve).
6-9mm per decade would mean that the 6-9m change would require 10,000 years.
Which is plenty of time for plenty of other processes to alter the level of the land.

Ammonite,
Hegerl and Knutti are just presenting a review of the literature, and the published literature, mostly model based, does generally agree on a climate sensitivity of 3 degrees C. But there is more to be learned from reading paper and its references than that. Consider for instance that they don’t discuss the model diagnostic literature. Since all the AR4 models have correlated error in several different diagnostics, including representing less than half the observed increase in precipitation (Wentz), failing to reproduce the amplitude of the observed response to the solar cycle (Camp and Tung) (Lean), more than 3 W/m^2 globally and annually averaged bias in surface albedo feedback (Roesch), errors in tropical radiative suggestive that the net feedback is negative while the models have it positive (Spencer) (Lindzen), etc., Hegerl and Knutti failed to discount the model based results. I don’t see how they can be discounted in a way which has them significantly and credibly contributing to the final conclusion. Please note that in their discussion of “various observations” they admit, not only model dependency, but dependency on just one climate model. It was just slab ocean model, by the way:
“The main caveat is that all three studies are based on a version of the same climate model and may be similarly influenced by biases in the underlying model.”
Let’s now consider the model independent estimates of climate sensitivity. Upon closer examination, we find that much of it is actually model dependent, one of the papers of Annan for instance, that claims to arrive at a model independent estimate, uses models to evaluate the maunder minimum and (if I recall correctly) the Holocene optimum and a previous glacial/interglacial transition.
Note another limitation disclosed by Knutti and Hegerl:
“The concept of radiative forcing is of rather limited use for forcings with strongly varying vertical or spatial distributions.”
This does not come as a revelation to those who have thought much about nonlinear dynamic systems. So which forcings have strongly varying vertical or spacial distributions? Well, solar for one, couples more strongly to the stratosphere and surface including 10s of meters of the ocean than it does to the troposphere. The well mixed GHGs couple mainly to the troposphere, with H2O having an uneven distribution, and aerosols couple to the troposphere is a vertically and spatially varying way. For the purposes of the AGW hypothesis, it is climate sensitivity to CO2 and the other non-H2O well mixed GHGs that is over interest. There is no reason to assume that the climate sensitivity to CO2 is the same as that to solar or aerosols, so we can eliminate those estimates based upon solar and aerosol forcings.
Next we can consider the paleo estimates of climate sensitivity to CO2. The errors in estimates of temperature from proxies and CO2 levels are both high, and the estimates of climate sensitivity based on long term equilibrium responses are of questionable relevance to what will happen in the next century or two, because they either involve a crossover to a different climate mode, or because higher estimates of climate sensitivity require time frames in which significant melting of the Greenland ice sheet can occur. All the glacial outlets of the Greenland ice sheet would have to be galloping continuously at rates only seen infrequently for hours at a time and the elevation of the top of the ice sheet would have to be considerably lowered.
Rather than just summarily rejecting all these studies, lets also try to understand why the estimates of climate sensitivity might be high. The model diagnostics help here, because they show that the models under represent the observed response to the solar cycle variation. Models that under represent solar might need to be excessively sensitive to CO2 in order to reproduce the recent warming despite the fact that solar activity was at a grand maximum during the latter half of the 20th century. Furthermore, the models might also need excessively high sensitivity to CO2 because they were unable to reproduce the multi-decadal climate modes that were in the positive phases during the recent warming. The under representation of the negative feedback of the water cycle and the questionable positive tropical cloud feedback might be how they achieve this higher sensitivity.
Lets look at why the estimate of climate sensitivity to solar radiative forcing might be high. Solar variation is admitted by the IPCC FAR to be poorly understood and possibly off by a factor of two. If variation in solar radiative forcing is actually higher, the sensitivity required to explain past climate variation is lower, most studies just use Lean’s estimates without allowance for that uncertainty. But I think the real reason that estimates of climate sensitivity to solar radiative forcing is likely to be too high, is because the climate response to solar is not just to radiative coupling to the climate but to other nonlinear coupling mechanisms. Solar UV variation couples chemically to the stratosphere and a lesser extent to the troposphere through chemistry, the creation of the greenhouse gas ozone. It was already being understood in 2008 that solar UV varied by several percent over the course of a solar cycle. Recent work in the current extended minimum show that these minima are not like the shorter solar cycle minimum. UV varies even more greatly, perhaps by an order of magnitude. Therefore models might have required much higher sensitivities to solar radiative forcing to understand the maunder and dalton minima and the medival warm period because the climate response was not just due to radiative forcing but to this other nonlinear coupling which they didn’t represent.
The models implement radiative coupling similarly for solar and CO2, for instance CO2 is usually coupled to the whole mixing layer of the oceans even though it barely penetrates microns into the surface while solar penetrates 10s of meters. Similar treatement of the CO2 and solar wavelengths occur at the land surface as well. Therefore model implementations which amp up the response to solar radiation in order to explain past climate response which was due to coupling that wasn’t just radiative, will also have amped up the response to CO2 radiative forcing.
Past aerosol forcing is poorly understood and is so poorly constrained that it supplies most of the explanation of how models with more than a factor of two difference in climate sensitivity can “match” the same climate. The Pinatubo eruption aerosols are probably better understood, but estimates of climate sensitivity from this event are usually also tainted by models.
Basically, there is practically no model independent evidence of climate sensitivity to CO2 in the current climate regime. The direct effects of CO2 can only explain about a third of the recent warming and a climate sensitivity of 1 to 1.1 degrees C. Anything more requires net positive feedback. Recent model independent evidence suggests that the net feedback might actually be negative (Lindzen) (Spencer) and at the very least that the models have another important conflict with the observations.
The science is not settled.

Ammonite says:
October 2, 2010 at 6:25 pm
————
The last interglacial period, which this thread was started on, actually invalidates all those papers you cited (which I have read thoroughly before). Temperatures rose 2.25C above the most recent interglacial yet all of the forcing factors that the papers are based are exactly the same between the two periods. CO2 increased enough to cause about 1.6C of warming (at 3.0C per doubling) in the last interglacial yet temperatures increased 7.0C.
The Lorius paper starts out by defining the planetary energy balance N as:
N = (1 – Albedo)/4*Solar – 0.96 * 5.67e-8 * (Temp Surface)^4
– from there, one can determine all kinds of sensitivities including the CO2 sensitivity because we know Solar, we know most of the Temp Surface numbers, we know CO2/GHG forcing numbers: yet not a single climate paper ever provides what Albedo estimate they used. How can you determine the change in total forcing and energy balance and the sensitivities without it. Well, one can just use whatever estimate one wants to [And they do – It is the major fudge factor in all the paleo climate model reconstructions – basically they just choose to use an Albedo number which gives them 3.0C per doubling of CO2/GHGs].
Covey’s 1996 paper contains the following quote (in reference to the simulation of the Eocene, a period in which there was no sea ice and no Greenland or Antarctic glaciers) –
“Covey et al. (1991) estimated that 2 to 3 W m -2 radiative forcing would
result from the complete disappearance of sea ice from the present-day Earth, but
we have excluded sea ice changes from our definition of [change in] Q because we want to include such changes in the feedback processes measured by the paleocalibration technique. Accordingly we consider only the remaining contributor to changes in surface albedo, namely higher sea levels and the resulting decrease in the fraction of relatively high-albedo land areas.”
– ie we just changed our previous estimate so that we could use “the paleocalibration technique” – they just ignored something that is -3 watts/m2 (almost the same number Hansen uses for the forcing change from all the ice and vegetation changes in the ice ages) – which is itself incompatible with the Covey 1991 estimate of losing just sea ice since the ice sea ice and glaciers should be something like -15 watts/m2).
– etc. etc.

It’s a pity really – think of the manure potential for the veg-garden!
As a lad, I was very fortunate to be taken under the wing of Prof. F.W. Shotton, who was something of a big-hitter on the Quaternary of the UK. He used to take me on his research-trips and they were a major education into how the UK had changed from glacial to interglacial and how the resultant deposits & fossil faunas developed as a consequence.
I know a lot of regulars on here are hostile to my take on climate change, but I’m happy to put that to one side if you are, and to discuss any questions of a more geological nature. My background is in mineralogy & precious metals exploration primarily, but also in the geological evolution of the British Isles, which has some of the most diverse geological history for any comparably-sized landmass around the world. The events of late have again reminded me that, in the climate debate, both sides screw up at times: however, that should not prevent us from talking about the scientific method and the need to apply it rigorously. For all the posts that I have read on here attempting to rubbish climate science, I still stick with what I see as the overwhelming balance of evidence on that front, which is counter-wise to what most posters on here seem to think, but in geology I am happy to drop in and contribute to less controversial topics, from time to time. Apart from anything else, I think it is important to engage with people, rather than the “traditional” wars between “warmists” and “deniers”. We’ll know either way who was right on that within a few years, after all.
Cheers – John

Bill Illis says: October 3, 2010 at 7:00 am
“yet not a single climate paper ever provides what Albedo estimate they used”
Thankyou for your response Bill. I will take this as a reading assignment (no sarcasm).

Passed the point of arguing how many angels can dance on the head of a pin
– reached the stage of mandating the maximum number of them that will do it

Ammonite,
Yes, you are correct that the annually averaged surface albedo bias documented by Roesch is extremly small, however, at about 3W/m^2 it is comparable in size to the Anthropogenic greenhouse forcing and more than 3 times larger than the energy imbalance of less than 0.8W/m^2 that Hansen calculated was being stored into the oceans during the 1990s. The 20th century warming was a small difficult to detect phenomenon, so it should be no surprise that climate models are not up to the task of attributing and projecting the phenomenon yet.
You sound as if you aren’t familiar with the Wentz publication in the journal science showing that none of the models produced as much as half of the observed increase in precipitation. Not only does correlated under representation of a key negative feedback by more than a factor of two create a credibility and validation problem for attributing and projecting warming, it calls into question all the model based conclusions that there will be increased risk of droughts, that are being funded by countries, states and regions and reported without discussing the model diagnostic literature. The money is just too hard to turn down.