Guest Post by Willis Eschenbach
Today I came across an IPCC figure (AR4 Working Group 1 Chapter 2, PDF, p. 208) that I hadn’t noticed before. I’m interested in the forcings and responses of the climate models. This one showed the forcings, both at the surface and at the top-of-atmosphere (TOA), from the Japanese MIROC climate model hindcast of the 20th century climate.
Now, do you notice some oddities in these two figures? Here’s what caught my eye.
The first oddity I noticed was that the surface forcing from the long-lived greenhouse gases (LLGHG) was so small compared to the top-of-atmosphere LLGHG radiative forcing. At the end of the record, the TOA forcing from LLGHG was just over two watts per square metre (W m-2). The surface forcing from LLGHG, on the other hand, was only about 0.45 W m-2. I don’t understand that.
This inspired me to actually digitize and measure the surface vs TOA radiation for a few of the components. For each W m-2 of TOA radiative forcing from a given source, the corresponding surface forcing was as follows:
Aerosol Direct: up to 15 W m-2 (variable)
Land Use: 1.5 W m-2
Volcanic Eruptions: 0.76 W m-2
Solar: 0.72 W m-2
Cloud Albedo: 0.67 W m-2
LLGHG: 0.21 W m-2
With the exception of the Aerosol Direct these relationships were stable throughout the record.
I have no idea why in their model e.g. one W m-2 of TOA solar forcing has more than three times the effect on the surface as one watt of TOA greenhouse gas forcing. All suggestions welcome.
The next oddity was that the sum of the radiative forcings for “LLGHG+Ozone+Aerosols+LandUse” is positive, about 1.4 W m-2. The surface forcing for the same combination, on the other hand, was strongly negative, at about -1.4 W/m2. The difference seems to be in the Aerosol Direct figures. It seems they are saying the aerosols make little difference to the TOA forcings but a large difference to the surface forcings … which seems possible, but if so, why would “Land Use” not show the same discrepancy between surface and TOA forcing? Wouldn’t a change in land use change the surface forcing more than the TOA forcing? But we don’t see that in the record.
In addition, the TOA Aerosol Direct radiative forcing changes very little during the period 1950-2000, while the corresponding surface forcing changes greatly. How can one change and not the other?
The next (although perhaps not the last) oddity was that the total surface forcing (excepting the sporadic volcanic contribution) generally decreased 1850-2000, with the total forcing (including volcanic) at the end of the period being -1.3 W m2, and the total forcing in 1950 being -0.6 W m-2 … why would the total surface forcing decrease over the period during which the temperature was generally rising? I thought perhaps the sign of the forcing for the surface was the reverse of that for the TOA forcings, but a quick examination of the corresponding volcanic forcings shows that the signs are the same. So the mystery persists.
In any case, those are the strangenesses that I found. Anyone with ideas about why any of those oddities are there is welcome to present them. What am I missing here? There’s some part of this I’m not getting.
In puzzlement,
w.
PS – I’m in total confusion regarding the albedo forcings that go all the way back to 1850 … if I were a suspicious man, I might think they just picked numbers to make their output match the historical record. Do we have the slightest scrap of evidence that the albedo changed in that manner during that time? Because I know of none.
PPS – Does anyone know of an online source for the surface and TOA forcing data in those figures?
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
With all due respect Sir, if you are confused what chances have we got ?
Perhaps it’s a Japanese thing ?
Willis: I think you have that backwards. Or rather, the authors do. The top panel is TOA. The bottom panel, with the higher forcing, is the surface. According to the caption under the graph.
Of course, that means the legends or the captions are wrong. The graph legends say one thing, the caption under the graph another.
Either way, the authors cocked it up.
I might think they just picked numbers to make their output match the historical record.
Well, it wouldn’t be the first time.
If solar forcings are not following model forecasts then the modelsa are wrong. What is the surprise?
LLGHG forcings lower than solar, I suggest you look again at this poor theory of GHG’s.
I prefer the Cosmic ray theory which now seems to be on the way to proof. All forcings can be explained using this. No need of GHG’g at all!
In situations like this the answer is simple , because the models said so
Willis,
The caption below the graphs doesn’t agree with the labels used on the graphs themselves. In the caption, it says that the top panel is the radiative component. Perhaps the caption is correct and the labels aren’t.
/dr.bill
Albedo
I have what might be a dumb question on albedo at the Arctic. A standard MMGW feedback argument is that with warming, the ice melts causing a change in albedo and so more warming as less incoming HF radiation is reflected back space.
Yet I remember from my O level physics that we regard the suns ray as travelling in parallel lines. If that is true, the surface at the Arctic is virtually parallel to the rays and thus they pass through the atmosphere with very limited opportunity to melt the ice.
Regards
Paul
Les Johnson says:
June 5, 2011 at 1:58 am
Good catch on the mixed-up labels, Les, I’d missed that. But the bottom one has to be the TOA radiation, as the title (but not the caption) says, because otherwise the TOA CO2 forcing would only be about 0.4 W/m2, and that makes no sense at all.
w.
Paul Maynard says:
June 5, 2011 at 2:29 am
The only dumb question is the one you don’t ask.
Well … kinda. You are right in the general concept. Near the poles the low sun angle causes greatly reduced insolation due to a) the angle itself, b) the dependence of albedo on angle, c) the increased distance through the atmosphere, and d) the increased cloud shadows due to low solar angle.
In addition, the changes in the ice don’t make as much difference as you might expect for another reason—in general, the ice is where the sun isn’t. That is to say, for example, that increases in the ice area during the month-long polar night don’t increase the albedo at all, because the sun’s not around to see it.
Finally, much of the polar areas are often covered by low-lying stratus clouds, and changes in the underlying surface mean little to the albedo in that condition.
Despite that, the changes in the ice area (and to an even greater extent in the snow area) can make a difference in the planetary albedo. Think about Nebraska on a clear winter day, and the difference in albedo if it’s snow-covered.
It’s just that, for the reason you cited as well as other reasons, the change in the actual albedo is not as large as you might initially think.
w.
Willis: In my mind, the TOA should be less than surface. If the TOA is less, then that means heat is accumulating (less LWR loss to space). If TOA is higher, that should mean that there is a net cooling.
But I agree that something is wonky besides the captions. The thick black line on both charts should cancel out, as they are about the same value, with one positive and the other negative. Are they saying there was no warming?
At least they show cloud albedo as an increasing negative forcing. That is a first for these models, no?
Willis:
You say;
“I have no idea why in their model e.g. one W m-2 of TOA solar forcing has more than three times the effect on the surface as one watt of TOA greenhouse gas forcing. All suggestions welcome.”
Choosing arbitrary values of ‘TOA solar forcing’ and ‘TOA greenhouse gas forcing’ to obtain a curve fit would be consistent with the known practice of choosing arbitrary values of ‘climate sensitivity’ and aerosol negative forcing to obtain a curve fit.
The models use wide ranges of ‘climate sensitivity’ and aerosol negative forcing to obtain reasonable hindcasting of global temperature. The difference between AOGCMS for used values of these parameters is a factor of ~2.5.
So, it seems reasonable to suggest (n.b. I merely “suggest” because I have no evidence) that
“solar forcing has more than three times the effect on the surface as one watt of TOA greenhouse gas forcing”
in the models is an effect of curve fitting to obtain an acceptable hindcast.
Richard
“The first oddity I noticed was that the surface forcing from the long-lived greenhouse gases (LLGHG) was so small compared to the top-of-atmosphere LLGHG radiative forcing.”
In the lower atmosphere there is too much CO2 and, above all, H2O.
If CO2 concentration increase, its effect is felt higher up in the troposphere.
IR radiation from the surface is not seen from space (IR windows excluded, of course).
It’s difficult to make any sense out of this incoherent mess with labels inverted but there are a couple of things to note.
The LLGHG surface forcing is probably being amplified by the hypothetical positive feedback “somewhere” in the the atmosphere. Remember the basis of all these models requires multiplying “what the science says” the CO2 forcing is by some voodoo science to produce the required TOA imbalance.
It appears that what you are noting is the “climate sensitivity” of this model.
Measuring this response of the model and pretending this is now “data” then allows you to multiply how much global temps will rise by the same number and scare the shit out of the populus.
In answer to your other question , yes they do actually make most of the forcings up as well since for most no data exist. I saw recently a post that looked into this (sorry can’t remember a link) . Most of them seem to based on handwavy estimates that thing must go up with increasing industrialisation, so they make them up to all look about the same and it fits the plan so, fine.
The one where there is some data , black carbon, actually bore no resemblance to real data and had the same hand-made, man-made century long trend.
wrt albedo change
Goldewijk, K. (2001), Estimating Global Land Use Change Over the Past 300 Years: the HYDE Database, Global Biogeochem. Cycles, 15(02), 417-433.
http://www.mnp.nl/hyde
Important error in your first line : the fourth assessment report is called AR4 *not* FAR which refers to First Assessment Report.
I was wondering why you bothered with something that old until I checked what you linked it to.
Willis Eschenbach says:
June 5, 2011 at 2:48 am
There is also the inconvenient fact that as the ice melts, the sea itself, which is pretty much always warmer than the ice, is no longer insulated and can radiate heat to be lost to space. Indeed, that could even make a low sea ice coverage (esp the Arctic), a negative feedback.
Have you considered contacting the authors and asking them to explain?
“PS – I’m in total confusion regarding the albedo forcings that go all the way back to 1850 … if I were a suspicious man, I might think they just picked numbers to make their output match the historical record. Do we have the slightest scrap of evidence that the albedo changed in that manner during that time? Because I know of none.”
Please Willis such crude language!
The models are parameterized. Each forcing is a guessed value and its effect is a guessed value. The model is then run varying the guessed values until the output appears to match the real world. Any real world data that still does not match the model guessed value is then ‘adjusted’ and the original raw data discarded.
Considering Cloud cover has decreased by 4% during the satellite period resulting in a large 0.9w/m2 positive forcing it seems strange to me that the model shows the cloud albedo having a very constant decline over the century resulting in a negative 0.6w/m2
forcing.
Talk about making things up!
The models can’t predict forward so how can they hindcast? If you force fit the hindcast to the historical actuals based on the preconception of how the model should work I’d suspect some of the input parameters/forcings would need to be pretty wierd. But hey it’s the “right answer” that counts not reality, science or chaotic behaviour that is important, just the message. Lesson is don’t look at data, it’s not designed for that.
These modelers are trying to simulate something that cannot be simulated because of lack of data. There simply is not enough quality data to do it. Now that being said we learn a lot about the system by trying. The modelers have been caught making up data and starting with the answer. Of course the tired old saying applies here. Garbage in Garbage out. Forcings = Fudge Factors Who decides what Fudge Forcing to use; that’s the question?
Julian in Wales:
At June 5, 2011 at 3:56 am you ask the reasonable question;
“Have you considered contacting the authors and asking them to explain?”
But there is a problem with that.
IPCC AR4 Chapter 2 which Willis is questioning says:
“As an example, the temporal evolution of the
global and annual mean, instantaneous, all-sky RF and surface
forcing due to the principal agents simulated by the Model for
Interdisciplinary Research on Climate (MIROC) + Spectral
Radiation-Transport Model for Aerosol Species (SPRINTARS)
GCM (Nozawa et al., 2005; Takemura et al., 2005) is illustrated
in Figure 2.23.”
At the end of that Chapter, the references indicate that Figure 2.23 is illustrating information obtained from:
Nozawa, T., T. Nagashima, H. Shiogama, and S.A. Crooks, 2005: Detecting
natural influence on surface air temperature change in the early twentieth
century. Geophys. Res. Lett., 32, L20719, doi:10.1029/2005GL023540.
Then, looking for that paper one soon discovers it is behind a $25 pay wall.
It is not known if the unknown authors of IPCC AR4 Chapter 2 generated the graphs of their Figure 2.23 as their statement (above) suggests, or if they copied it from the reference?
So, without spending $25, whom does one ask? And how long will it take to get answers?
Richard
I would say there’s some substantial “tuning” involved here. The caption to the figures says “as simulated.” The text that refers to this set of figures is at http://www.ipcc.ch/publications_and_data/ar4/wg1/en/ch2s2-9-5.html and says:
…Radiative forcing time series for the natural (solar, volcanic aerosol) forcings are reasonably well known for the past 25 years; estimates further back are prone to uncertainties (Section 2.7). Determining the time series for aerosol and ozone RF is far more difficult because of uncertainties in the knowledge of past emissions and chemical-microphysical modelling. Several time series for these and other RFs have been constructed (e.g., Myhre et al., 2001; Ramaswamy et al., 2001; Hansen et al., 2002). General Circulation Models develop their own time evolution of many forcings based on the temporal history of the relevant concentrations. As an example, the temporal evolution of the global and annual mean, instantaneous, all-sky RF and surface forcing due to the principal agents simulated by the Model for Interdisciplinary Research on Climate (MIROC) + Spectral Radiation-Transport Model for Aerosol Species (SPRINTARS) GCM (Nozawa et al., 2005; Takemura et al., 2005) is illustrated in Figure 2.23. Although there are differences between models with regards to the temporal reconstructions and thus present-day forcing estimates, they typically have a qualitatively similar temporal evolution since they often base the temporal histories on similar emissions data…
Look at some of the original Ramanathan articles on this.
A question on the albedo of water. I thought that when the angle of incident was very low the albedo of water increased tremendously. In the high arctic and antarctic the sun is always low to the horizon so wouldn’t the change from ice to open water have little effect?
Just thinking.
ArtR