The Greenhouse Delusion - Chapter 7: by Dr Vincent Gray

Forecasting the Future

Forecasting the future can hardly be considered as a science because there are too many unknown variables. Weather forecasters are expected to tell us what the weather will be, but despite a large array of scientific equipment they are still often wrong. Economists are employed to predict the future behaviour of the economy, but they are rarely right. It has been said that they are the only profession who continue to be employed even when they are always mistaken.

Because of future uncertainties, economists are usually reluctant to provide forecasts further ahead than a decade or so. Yet climate scientists are now routinely telling us what will undoubtedly happen a hundred years, or sometimes several hundred years ahead.

The previous Chapter has shown that computer models of the climate based on the assumption that the greenhouse effect is the only influence on the climate have never successfully predicted any future climate change, and have such large uncertainties that they are quite unsuitable for future forecasting.

The models usually calculate the consequences of increasing atmospheric carbon dioxide concentrations, typically the effect of it doubling. Such a model does not, however constitute a future prediction unless it also includes a forecast of how long it will take for the concentration of carbon dioxide to double.

In order to do this, it must have a scenario of how carbon dioxide and other greenhouse gases will change their atmospheric concentrations in the future. Then there has to be a model which converts greenhouse gas emissions to atmospheric concentrations, and then the climate model which can provide radiative forcing the extra radiation at the top of the atmosphere due to the greenhouse gas changes and then yet another model to convert this forcing to a projected temperature or sea level change at the earths surface.

It should be made plain that all the estimates provided in the scenarios are purely a matter of the opinions of "experts", comparable to the opinions of the "confidence" to be placed on models.

A range of scenarios has been used by the IPCC in order to forecast, predict, or as they would claim to prefer it, project the future.

The typical, and still a highly popular scenario, is the assumption that the carbon dioxide concentration in the atmosphere will increase by 1% a year. This assumption is made by 30 of the 99 models listed in Chapter 9 of Climate Change 01 (1).

The assumption merely multiplies the measured rate of rise of CO2 of 0.4% per year for the past 27 years, by a factor of 2, and is typical of the wild, irresponsible exaggeration which pervades almost all of the IPCC scenarios.

Climate Change 01 (2) gives the following excuse.

"A common standardised forcing scenario specifies atmospheric CO2 to increase at a rate of 1% a year compound until the concentration doubles (or quadruples) and is then held constant. The CO2 content of the atmosphere has not, and likely will not, increase at this rate (let alone suddenly remain constant at twice or four times an initial value). If regarded as a proxy for all greenhouse gases, however, an "equivalent CO2" increase of 1% a year does give a forcing within the range of the SRES scenarios"

In addition to admitting the gross exaggeration involved in multiplying a well-established rate of rise in CO2 by 2, this passage then claims that there could be a 1% rise in "equivalent CO2", which includes other greenhouse gases. Yet the most important of the other greenhouse gases, methane, has been increasing at a declining rate for the past 14 years, and its concentration is currently falling. The only way the scenarios can provide a 1% increase in "equivalent CO2" is to combine an exaggerated rise in CO2 itself with a reversal of the recent methane trend, and a forecast concentration increase. The above passage shows that the irresponsible exaggeration of 1% CO2 increase a year still pervades the most recent, SRES scenarios.

There have been three sets of more detailed scenarios than the 1% solution.

The first, launched with Climate Change 90 (3) consisted of four scenarios, A, B, C, and D. Scenario A was termed "Business As Usual" (BAU) or as SA90, and was considered by many as a plausible forecast for the future.

The next series of six scenarios was developed for Climate Change 92 (4), and the details are available from a supplementary Report (5). The scenarios were designated IS92a, IS92b, IS92c, IS92d, IS92e and IS92f.

A new set of scenarios was used in Climate Change 01 (6).  The scenarios were prepared by a special committee of IPCC Working Group III. Their Report Special Report on Emission Scenarios (7) was produced without any input from the scientists involved with Working Group I, and its conclusions were foisted on Climate Change 01 without opportunity for discussion. Six teams of specialists from 18 countries drew up a total of 40 scenarios, all based on four "storylines" claimed to represent different views of what might happen in the future. The 40 scenarios were originally summarised in the form of four "Marker Scenarios", A1, A2, B1. and B2, but as time went on, they split A1 into three; A1F1, A1B and A1T.

The existence of all these different scenarios has led to considerable confusion, as the modellists did not know which to use. The Special Report on Emission Scenarios 00 (7) was published a year before Climate Change 01 which has provided additional information on the scenarios which was not present in the Special Report . Additional confusion was caused by partial attempts to correct the IS92 scenarios to remedy the fact that they were unable even to predict 10 years ahead (8). Only one of these has been currently listed (9), an amended version of the carbon dioxide and methane concentrations for IS92a, which , confusingly has the same name, while the original IS92a is referred to as IS92a/SAR. Presumably there are also amended versions of the other IS92 scenarios which are not revealed.

The IPCC has gone to some trouble to explain that the scenarios should not be used to forecast the future.

Climate Change 92 (10) , says

"Scenarios are not predictions of the future and should not be used as such"

The Special Report on Emissions Scenarios 00 (11) puts it this way

"Scenarios are images of the future or alternative futures. They are neither predictions nor forecasts"

Despite these assertions the scenarios are widely used as predictions and forecasts, not only by politicians and "activists" of various colours, but often by the scientists who are responsible for them, in their public appearances. They are also assumed to be genuine forecasts by Climate Change 01 :Impacts (12) and by Climate Change 01 : Mitigation (13)

The Special Report on Emissions Scenarios 00 (14) also insists

"The possibility that any single emissions path will occur as described in the scenario is highly uncertain", and

"No judgement is offered in this Report as to the preference for any of the scenarios and they are not assigned probabilities of occurrence, neither must they be interpreted as policy recommendations"

This statement amounts to an endorsement of the wildest unlikely scenarios. The scenarios, are, as a matter of course, used extensively as policy recommendations despite this caveat.

Also (13)

"there is no objective way to assign likelihood to any of the scenarios. Hence there is no "best guess" or "business-as-usual" scenario"

There is an obvious objective way of assigning likelihood which is ignored by the IPCC.  Projections that prove to be correct over a period of, say, ten years, are more likely to be reliable than projections that are seriously incorrect over the same period.

The most priceless statement is (15);

"These tools are less suitable for analysis of near-term developments and this report does not intend to provide reliable projections for the near term"

Near-term projections that are unreliable must surely cast serious doubt on the credibility of long-term projections, particularly, as will emerge, when most of them are seriously exaggerated.

Climate Change 94 (16) says

"Scenarios deal with the future so they cannot be compared with observations"

Unfortunately, the future tends ultimately to become the present, and then the past. It then becomes possible to compare the scenarios with what has actually happened, and to assess whether they have been successful.

Tables 1 to 5 show the assumptions made by the various scenarios between the years 1990 and 2100 for a range of parameters. They are compared with the actual measured quantities for the years 1990 and 2000. Scenario figures are from Climate Change 90 for the A, B. C and D scenarios, Pepper et al (5) for the IS92 scenarios, and Special Report on Emissions Scenarios 00 (7) and Climate Change 01 (9) for the SRES scenarios. Figures often had to be interpolated by estimates from enlarged graphs.

Table 1. Scenario estimates of world population (in billions)


Table 1 shows how scenario estimates compare with measured population figures. The data are from UNFP (17), and the US Census Bureau (18).

All the "projected " values for the year 2000 are exaggerated by amounts between 0.6% and 6%.  Of the IS92 scenarios the most successful were IS92c and IS92d. The IPCC continues to place its faith on IS92a, however. The SRES scenario assumptions are too high for the year 2000 despite the publication date of 2001, with A2 nearly 2% too high.


Coal Production

An example of the future projections for energy usage in the scenarios can be shown by the figures for coal productionFigures are from the United Nations Energy Handbook (19), updated by BP (20)

Table 2 Scenario Estimates of Coal Production ( in Exajoules)

IS92 scenarios all exaggerate coal production in the year 2000, by between 3 and 17%. As before, the most successful scenarios are IS92c and IS92d. IS92e overestimates the 2000 figure by 17%, assumes an increase of 65% by 2010,  and envisages an increase in world coal production of eleven times over the 2000 figure by the year 2100.

The 1990 figures for the SRES "illustrative" scenarios are something of a mystery.  They are obtained by multiplying "Share of coal in primary energy" by "primary energy" in Table 2a of the "Summary for Policymakers of Special Report on Emissions Scenarios 00 (21). Individual scenarios (22) give much higher figures for 1990 that vary between 82 and 105 EJ. This, surely, shows the sloppiness that pervades these scenarios. There was, after all, a measured value of 96EJ in 1990 (19).

Despite the initial low figure, the A1 scenarios exceed the coal production figure for 2000 by as much as 16%.

A1F1 envisages a coal industry in 2100 7.2 times the size of 1990, and A2 10.8 times. The scenario A1C AIM (22) puts it up to 12.2 times. On the other hand, A1T projects a coal industry one quarter the size of 1990 by 2100, and the scenario B1T MESSAGE (22) gets it down to 2%. These extreme values are rather absurd.


Carbon Dioxide Emissions

Table 3 shows the assumed emissions of carbon dioxide from fossil fuels and cement, compared with the measured values (23, 20).  Figures assuming that the current rate of increase since 1974 (0.8 GtC/yr) might continue have been included . The Special Report on Emissions Scenarios (21, 22) confuses us by adding (or subtracting) emissions from land use changes to give "total anthropogenic emissions",  but the figures for fossil fuel combustion have been chosen here.

Most of the figures appear to include emissions from cement production. The Special Report on Emissions Scenarios 00 (21, 22) includes "industrial processes"

Table 3 Emission of CO2 by fossil fuel combustion, in GtC/yr as projected

The IS92 scenarios had one scenario ( IS92c) which seriously assumed that there would be hardly any increase in emissions from 1990 to 2000, and another (IS92d) which assumed an increase of only 0.2GtC less than what actually happened. Apart from these two, however, all the rest of the IS92 scenarios grossly overestimated the actual emissions figure for the year 2000 by between 3 and 5%.The SA90, ("Business as Usual") scenario, overestimated 2000 by 19%.

All of the SRES scenarios expect a sudden increase. as of now, in the rate of emission shown by the last 26 years trend (0.8% a year), to give figures for the year 2010 between 11 and 34% above that recent long-term trend. A1F1 and A2 more than double that trend by 2100. However, A1B and B2 have got back to the current trend by 2100, and A1T and B1 are below the 1990 emissions by 2100.


Carbon Dioxide Concentrations

Table 4 shows the atmospheric carbon dioxide concentrations used in the various scenarios, compared with the measured values (24), which have shown a constant rate of rise of 0.4% a year since 1974. The effects of this rate continuing are shown.

An innovation, in Climate Change 01 (9) is that the carbon dioxide concentrations resulting from the use of different carbon cycle models are given, so that there is a range of figures, depending on the model chosen. This example should be followed more often, as it reminds us how uncertain these figures really are. In Table 4 are included the results from the highest (TOP) and the lowest (BOTTOM) individual scenarios.

Table 4 Atmospheric Carbon Dioxide Concentrations in ppmv assumed by the scenarios

The differences between the different models illustrates the fact that the model outputs have very large uncertainties, which are much higher than the differences between individual scenarios.

Of the IS92 scenarios, IS92c and IS92d are compatible with a continuing 0.4% a year.  IS92e is the only one following the 1% solution, but IS92a gets there if you use the right model.

A1F1, and A2 have carbon dioxide concentrations for 2100 which are in the region of the 1% increase a year, again, with the right model. The TOP assumption, is well above even this level


Methane Concentrations

Table 5 shows the atmospheric methane concentrations assumed by the scenarios, compared with the measured values (24). As noted previously, the rate of increase of methane in the atmosphere has been falling for the past 17 years, and the actual concentration is currently falling. This trend must surely continue for a while, at least, and it may turn out to be a long-term effect. None of the scenarios are prepared to recognise this, even as a possibility.

This whole Table therefore has an air of unreality: a total unwillingness to believe actual measured values, or their trends. Only scenario B1 has a methane concentration in 2100 corresponding to the trend of the past 16 years, but even this scenario goes well above the trend as of now, so it assumes an immediate sudden reversal of the current trend. The whole set of scenarios can be dismissed as completely unrealistic from this Table alone

Table 5 Methane concentrations assumed by the scenarios in ppbv



The whole purpose of the IPCC is to provide estimates of the global temperature rise in the future, supposedly as a result of increases in greenhouse gases. In order to do this two sets of unreliable, unvalidated models (both climate models and carbon cycle models) are applied to mostly highly unlikely emissions scenarios, some of whose assumptions are described above. This means that almost any result can be obtained, depending on the choice of model, model parameters and scenario. This choice, therefore can readily be manipulated to secure a desired result, which is a set of projections which satisfies the political authorities that sponsor the research.

The process is well illustrated by the history of the currently published set of IPCC temperature projections

Climate Change 90 (25) combines a range of "climate sensitivity" (High estimate 4.5C, Best Estimate 2.5C and Low Estimate 1.5C) with the four scenarios A , B, C and D to give Figure 8, page xxii , which shows simulated temperature rise between 1765 and 2100 for the three Climate Sensitivities and Scenario A (Business as Usual, BaU). It shows a rise over this period of between 2.9C and 6.2C (1990 to 2100, 2.1C to 4.8C).

Figure 9, page xxiii, shows the rise from 1765 and 2100 for the Best Estimate climate sensitivity, and the four scenarios; a rise between 2.0C and 4.2 C (1990-2100 of 1.0C to 3.2C)

The Executive Summary (26) states:

"This will result in a likely increase in global mean temperature of about 1C above the present value by 2025 and 3C before the end of the next century". They then point out that the lowest scenario, D would give about 0.1C per decade

Climate Change 95, on pages 322 and 323 (27) has five figures giving possible temperature changes between 1990 and 2100.

Figure 6.20 (page 322) uses scenario IS92a, four different climate sensitivities and changing and constant aerosols. The range of temperature increase from 1990 to 2100 is from 1.4C to 3.5C.

Figure 6.21 (page 322) uses IS92a and climate sensitivity 2.5C, CO2 forcing alone, and with aerosols. The temperature rise 1990 to 2100 ranges from 1.8C and 2.4C

Figure 6.22 (page 323) shows the effect of all the IS92 emissions scenarios using full aerosol forcing and climate sensitivity of 2.5C. The temperature increase 1990 to 2100 ranges from 1.3C to 2.5C

Figure 6.23 shows the effect of IS 92 scenarios and constant aerosol, with climate sensitivity 2.5C. The range of increase 1990 to 2100 is 1.3C to 3.2C.

Figure 6.24 shows the extreme range of possible changes, between IS92c with constant aerosol and climate sensitivity of 1.5C, giving a rise of 0.8C,to IS92e with constant aerosol and climate sensitivity of 4.5C, giving a rise of 4.5C.

All these Figures used a single carbon cycle model.

It is difficult from this variety to decide which figures to take, and the "Summary for Policymakers" expresses this difficulty. as follows

"For the mid-range IPCC emission scenario IS92a, assuming the "best estimate" value of climate sensitivity, and including the effects of future increases in aerosol, models project an increase in global mean surface air temperature relative to 1990 of about 2.0C by 2100. This estimate is approximately one third lower than the "best estimate" in 1990.....Combining the lowest IPCC emission scenario with a "low" value of climate sensitivity and including the effects of future changes in aerosol concentrations leads to a projected increase of about 1C by 2100. The corresponding projection for the highest IPCC scenario (IS92e) combined with a "high" value of climate sensitivity gives a warming of about 3.5C"

There was little exaggerated public attention drawn to the obviously extreme estimate of 3.5C warming by 2100 after the report was published.

Figure 1 shows the current IPCC projected range for the temperature rise from 1990 to 2100 from Figure 9.14 of Chapter 9 of Climate Change 01 which is the same as Figure 22a of the Technical Summary (29). A slightly different Figure appear as Figure 5(d) of the Summary for Policymakers (30). For the first two the alleged range between IS92c low and IS92e high is from 1.2C to 3.4C. For the Summary for Policymakers diagram (28) the range shown for "All IS92" is between 1.0C and 3.5C.

The big change from Climate Change 95 is the upper end of the range of the increase in projected temperature rise from 1990 to 2100, which is now 5.8C instead of 3.5C. There is also an increase in the lower end of the range from 1.0C to 1.4C.

Climate Change 01 (31) explains this as follows :

"It is evident that the range in forcing for the new SRES scenarios is wider and higher than in the IS92 scenarios. The range is wider due to more variation in emissions of non-CO2 greenhouse gases".

The range for the annual emissions most important non-CO2 greenhouse gas, methane, for the year 2100 changed only from 546-1072Mt for the IS92 scenarios to 236-1069Mt, which should have led to a reduction in the lower range figure.

Then (31)

"The shift to higher forcing is mainly due to the reduced future sulphur dioxide emissions of the SRES scenarios compared to the IS92 scenarios"

It is true that the range of sulphur emissions of 87-254 MtS/yr for the IS92 scenarios for the year 2100 was reduced to 11-83 MtS/yr for the SRES scenarios, but it was not explained why increased amounts of use of coal with increasing sulphur content could justify such a reduction.

But this is not the whole story, as can be seen by studying the First and Second Drafts of Climate Change 01.

In the First Draft Figure 9.13a on page 66 showed a range of projected temperature increase between 1990 and 2100 of between 1.7C and 4.0C.

In the Second Draft, Figure 9.18, page 88 showed a range of projected temperature increase from 1990 to 2100 of between 1.3C and 5.0C

Since the effects of lower sulphur dioxide had already been incorporated into the First Draft, the escalation from 4.0C to 5.8C in the maximum of the range must have been secured by other means.

This was done by two devices. Firstly the original four "Marker" Scenarios from the 40 SRES scenarios were expanded into six, notably by scenario A1F1 which has an extremely high use of fossil fuels. If that was not enough, the range was increased still further by applying the "envelope of all scenarios" to Figure 1, so that the maximum assumptions of all the scenarios could be utilised.

Then, instead of using only one climate model, as was the previous custom, no less than seven models were used, particularly a very high one, GFDL-R15-a.

Section 9.3.3 of Climate Change 01 (32) is a record of how Figure 1 was designed to suit the requirements of the politicians, by including some uncertainties, omitting others, choosing parameters to suit, and leaving others out. It is perfectly clear that by a suitable choice of parameters, models and uncertainties any required temperature projection can be supplied. They make this point clear in the following passage

"The range for these two parameter settings for the full set of SRES scenarios is 1.4 to 5.8C. Note that this is not the extreme range of possibilities, for two reasons. First, forcing uncertainties have not been considered. Second, some AOGCMs have effective climate sensitivities outside the range considered."

They say this in order to suggest that the upper end of the range could be even higher, but they fail to mention that the lower end of the range could also extend, to zero or negative temperature change. If they were really honest they would confess the truth, which is that the models and the projections are compatible with almost any conceivable temperature change in the future and are therefore worthless as a guide to future action.


  1. Climate Change 01 Chapter 9, Projections of Future Climate Change. Table 9.1, pages 538-540

  2. Climate Change 01 Chapter 9 page 533

  3. Climate Change 90 Appendix 1 page 341

  4. Climate Change 92 Chapter A3 Emissions Scenarios for IPCC: An Update, pages 69-95

  5. Pepper, W, et al 1992 Emission Scenarios for the IPCC: An Update . IPCC supplementary paper

  6. Climate Change 01 page 799.

  7. Special Report on Emissions Scenarios 00

  8. Gray, V R 1998 The IPCC future projections: are they plausible? Climate Research 10 155-162

  9. Climate Change 01 Appendix II, SRES Tables pages 807-810

  10. Climate Change 92 page 9

  11. Special Report on Emissions Scenarios 00 Technical Summary, page 1

  12. Climate Change 01 : Impacts

  13. Climate Change 01 : Mitigation

  14. Special Report on Emissions Scenarios 00.. Summary for Policymakers page 3

  15. Special Report on Emissions Scenarios 00 Technical Summary page 2

  16. Climate Change 94 page 252

  17. United Nations Population Fund . Annual Reports

  18. United States Census Bureau

  19. United Nations Energy Handbook

  20. British Petroleum Energy Statistics

  21. Special Report on Emissions Scenarios 00 Summary for Policymakers, Table 2a page 91

  22. Special Report on Emissions Scenarios Appendix VII, Tables

  23. Trends; Carbon Dioxide Information and Advisory Center. http:/

  24. Carbon Cycle Group, NOAA,

  25. Climate Change 90 pages xxii and xxiii

  26. Climate Change 90 page xi

  27. Climate Change 95 pages 322 and 323

  28. Climate Change 95 page 5

  29. Climate Change 01 Chapter 9, page 555, Technical Summary page 70

  30. Climate Change 01 Summary for Policymakers , page 14

  31. Climate Change 01 page 555

  32. Climate Change 01 pages 554-557

Return to Chapter list and Summary 

Return to `Climate Change Guest Papers' page

Return to `Still Waiting for Greenhouse' main page

FastCounter by bCentral