Climate Science Glossary

Term Lookup

Enter a term in the search box to find its definition.


Use the controls in the far right panel to increase or decrease the number of terms automatically displayed (or to completely turn that feature off).

Term Lookup


All IPCC definitions taken from Climate Change 2007: The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Annex I, Glossary, pp. 941-954. Cambridge University Press.

Home Arguments Software Resources Comments The Consensus Project Translations About Donate

Twitter Facebook YouTube Pinterest

RSS Posts RSS Comments Email Subscribe

Climate's changed before
It's the sun
It's not bad
There is no consensus
It's cooling
Models are unreliable
Temp record is unreliable
Animals and plants can adapt
It hasn't warmed since 1998
Antarctica is gaining ice
View All Arguments...

Keep me logged in
New? Register here
Forgot your password?

Latest Posts


Lessons from Past Climate Predictions: IPCC FAR

Posted on 26 August 2011 by dana1981

The Intergovernmental Panel on Climate Change (IPCC) First Assessment Report (FAR) was published in 1990.  Its purpose was to assess the available scientific information related to the various components of climate change, and to formulate realistic response strategies for the management of the climate change issue.  In the process, the FAR made some projections of future global warming, whose accuracy we will evaluate in this post.

The FAR used energy balance/upwelling diffusion ocean models to estimate changes in the global-mean surface air temperature under various CO2 emissions scenarios.  Details about the climate models used by the IPCC are provided in Chapter 6.6 of the report.

The IPCC FAR ran simulations using various emissions scenarios and climate models. The emissions scenarios included business as usual (BAU) and three other scenarios (B, C, D) in which global human greenhouse gas emissions began slowing in the year 2000.  In 2010, the atmospheric CO2 concentration in BAU projected by the FAR was approximately 400 parts per million (ppm), and in Scenarios B, C, and D was approximately 380 ppm.  In reality it was 390 ppm, so we ended up right between the various scenarios.  The FAR greenhouse gas (GHG) radiative forcing and CO2-equivalent for the scenarios is shown in Figure 1.

FAR projected forcing

Figure 1: IPCC FAR GHG forcing and CO2-equivalent projections for the four  emissions scenarios

As you can see, the FAR's projected BAU GHG radiative forcing in 2010 was approximately 3.5 Watts per square meter (W/m2).  In the B, C, D scenarios, the  projected 2010 forcing was nearly 3 W/m2.  The actual GHG radiative forcing was approximately 2.8 W/m2, so to this point, we're actually closer to the IPCC FAR's lower emissions scenarios. 

However, aerosols were a major source of uncertainty in 1990.  In an improvement over Kellogg's 1979 projection study, the IPCC FAR was aware that an increase in atmospheric aerosols would cause a cooling effect.  However, they had difficulty quantifying this cooling effect, and also did not know how human aerosol emissions would change in the future.

The IPCC FAR ran simulations using models with climate sensitivities of 1.5°C (low), 2.5°C (best), and 4.5°C (high) for doubled CO2 (Figure 2).

FAR temp projections

Figure 2: IPCC FAR projected global warming in the BAU emissions scenario using climate models with equilibrium climate sensitivities of 1.5°C (low), 2.5°C (best), and 4.5°C (high) for double atmospheric CO2

We digitized these projections and compared them to the observed average global surface temperature change from GISTEMP (Figure 3).

FAR 1880 vs. observations

Figure 3: IPCC FAR BAU global warming projections (blue) vs. observed average global surface temperature change from GISTEMP five-year running average (red)

As you can see, the observed warming since 1880 has been between the IPCC BAU "best" (2.5°C sensitivity) and "low" (1.5°C sensitivity) projections.  However, as noted above, the actual GHG increase and radiative forcing has been lower than the IPCC BAU, perhaps because of steps taken to reduce emissions like the Kyoto Protocol, or perhaps because their BAU was too pessimistic. 

Regardless of the reason, we're not really interested in how well the IPCC scenarios projected the GHG changes; we want to know the accuracy of the model temperature projections.  We can take the observed atmospheric GHG changes into account, and see what the model would look like with the up-to-date estimates of the GHG forcings from the 2007 IPCC Fourth Assessment Report (Figure 4).

IPCC adjusted projections since 1880

Figure 4: IPCC FAR BAU global warming projections reflecting the observed GHG changes (blue) vs. observed average global surface temperature change from GISTEMP five-year running average (red)

Obviously the IPCC model is a bit oversimplified, failing to take into account the natural factors which contributed to the pre-1940 Warming, or the factors (primarily human aerosol emissions) which contributed to the mid-century cooling.  However, the IPCC "best" projection matches the long-term warming trend, particularly since about 1965, very closely.  As with Broecker's 1975 prediction, this is strong evidence that human greenhouse gas emissions have been the main driver behind the observed global warming over this period, and suggests that CO2 became the dominant climate driver in the mid-20th Century.

Since the IPCC projections were made in 1990, we can also evaluate how accurately they projected the global warming over the past two decades (Figure 5).

IPCC adjusted projections since 1990

Figure 5: IPCC FAR BAU "best" global warming projection reflecting the observed GHG changes (blue) vs. observed average global surface temperature change from GISTEMP (red) since 1990.  This figure has also been added to our hi-rez graphics page.

Now we see that had the IPCC FAR correctly projected the changes in atmospheric GHG from 1990 to 2011, their "best estimate" model with a 2.5°C equilibrium climate sensitivity would have projected the ensuing global warming very accurately.

It's also important to note once again that the IPCC models did not account for changes in human aerosol emissions, which have had a significant cooling effect at least over the past decade, or natural factors like solar activity, which has declined since 1990 as well.  This suggests that the IPCC "best" model equilibrium sensitivity of 2.5°C may be somewhat too low.

Accurate Climate Models

Figure 5 in particular shows once again that even two decades ago, global climate models were making very accurate projections of future global warming.  As with Broecker (1975) and Hansen (1988), the accuracy of the IPCC FAR global warming projections put a dagger in the myth that models are unreliable.  These results also add to the mountains of evidence that climate sensitivity is in the ballpark of 3°C for doubled atmospheric CO2.

0 0

Bookmark and Share Printable Version  |  Link to this page


Comments 1 to 12:

  1. What is the source of data for Figures 4 and 5 ?

    Did you adjust both aerosols and GHG forcings, or just GHG ?
    0 0
  2. Charlie - which data? The caption says that the observational data is from GISTEMP. If you mean the model adjustment, the source is the IPCC FAR projected GHG forcing (Figure 1) vs. the actual forcing (from the IPCC AR4, stated below Figure 1).

    It's a GHG forcing model, so I took observed GHG changes into account. As I said, the model doesn't account for aerosols. That's one reason why the "best" 2.5°C sensitivity is probably a bit on the low side, even though the model matches observations quite well (also noted in the post).
    0 0
  3. Base period for figures 2 to 5 is the year 1765?

    Thanks for the informative post, Dana. I've been curious about this comparision, which is in my view even more telling than Rahmstorf 2007.
    0 0
  4. Alexandre - correct, the baseline is 1765 as used in the IPCC FAR, and thanks!
    0 0
  5. #2 Dana I think your intro to fig 4 would be clearer if you changed it to something like

    ...and see what the model would look like with the more modern estimates of the GHG forcings from AR4 (Figure 4).

    As I read the text I had the impression you were only updating the projected forcings since 1990 and was rather surprised that the graphs differed before 1990. I gather from your reply to Charlie that you have replaced all the GHG forcings since 1880.
    0 0
  6. fair point mdenison; text revised.
    0 0
  7. I note that figures 3 and 4 have different "projections" for years such as 1940 and 1960.

    Did the "actual" forcings prior to 1990 change between FAR and AR4 ?

    I assume that the 3 blue lines in Fig 4 are for climate sensitivities of 1.5C, 2.5C and 4.5C per CO2 doubling. Correct?

    dana1981 says "It's a GHG forcing model, so I took observed GHG changes into account." --- please explain how you took the changes into account in a way to change the plot before 1990.
    0 0
  8. However, as noted above, the actual GHG increase and radiative forcing has been lower than the IPCC BAU, perhaps because of steps taken to reduce emissions like the Kyoto Protocol, or perhaps because their BAU was too pessimistic.

    Global carbon emissions have actually accelerated since the Kyoto Protocol was ratified. Over the 90s emissions growth was relatively slow, partly due to the collapse of the Soviet Union and the Mt. Pinatubo eruption. That would explain some of the difference.

    I think large factors in the lower GHG forcing have been the Montreal Protocol, which wasn't fully in place when the FAR was released, and the abrupt slowdown of methane growth (with various proposed causes). Current concentrations of CFCs are below even the Scenario D projections.

    Could you do a version of Figure 5 with an overlaid comparison to the original post-1990 BAU projection? That would show the importance of forcing differences.
    0 0
  9. Where is the algorithm and source code for the IPCC model published?
    0 0
  10. In Figure 5, why is the IPCC projection linear with slope changes in 2000 and around 2008? Also, why does the IPCC projection after 1880 have two sharp linear slope changes when actual data in this time period was nowhere as complete or accurate as today? How did the IPCC have data to use energy balance/upwelling diffusion ocean models to make projections in this time period? For scientific accuracy, I believe all these charts should show the key years when data collection methods and quality of data changed. After all, science is supposed to be about the data and vetting that data, not who is stating conclusions about the data. In Figure 4, how can a projection be below the actual at the starting point of the projection? There has to be a known starting point where both values were equal.
    0 0
  11. Ken,
    The link to the IPCC model description is highlighted in blue in the second paragraph of the lead post. In 1990 it was not the custom to post the algorithms and source code, but a detailed description of the models exists so that you can reproduce their data if you wish. Unfortunately, in the past 5 years once the algorithms were posted online skeptics have chosen not to examine them.

    The answers to your second post would be in the IPCC FAR report if you read it. Complaining that you do not understand how the data was graphed when you have not read the original report does not make a very convincing argument. Since the graph is of projections made in 1990 they are probably centered on the 1960-1990 time period, but they might be centered 1950-1980. The projection is below the measured line at the start of the graph because they do not perfectly model the data. Using my eyecrometer it seems that they are pretty close to the measured data. We are now seeing how well they projected into the future.
    0 0
  12. Charlie: If you rescale figure 4 to be 712 pixels high and superpose it on figure 3, you'll see that the projections are identical. I presume therefore that the calculation here is just a simple rescaling to fit the current GHG levels, without redoing the whole calculation.

    As you've obviously realised, that's a very crude approximation, since it changes the hindcast values, and in reality the forecast values will also be slightly out because of the lagged effects of the rescaled earlier emissions. And I'm guessing you're already thinking about the following: If the forcings and projections are both available, then we can deduce the response function of the IPCC model and redo the calculation properly. That would be an interesting exercise.

    I see a problem however - the curves are not very featureful, and so the response function may be ill determined.
    0 0

You need to be logged in to post a comment. Login via the left margin or if you're new, register here.

The Consensus Project Website



(free to republish)



The Scientific Guide to
Global Warming Skepticism

Smartphone Apps


© Copyright 2015 John Cook
Home | Links | Translations | About Us | Contact Us