If you have never really studied climate science but want to get to the essence of what the disagreements about the science amount to, then you may find this a helpful resource.

To be properly informed on one of the most important and politically charged issues of this century any engaged citizen needs to consider the facts, opinions and predictions emanating from climate science, as we would in any comparable area (such as public health).  This means asking the general questions of what is to be accepted as valid, what confidence can be placed in “expert” scientific opinion, and what additional conclusions we might draw from examining any controversies around the issue. The publication of popular summaries of the climate science arguments by both the Royal Society and the Global Warming Policy Foundation provided a rare opportunity to compare and contrast mainstream and sceptic arguments across a wide range of specific aspects of the science. This paper was informed and inspired by the author’s participation in the Institute of Actuaries 2104 Working Party on Climate Change.  Any views expressed here are however his responsibility alone.

Given that the science is central, a starting point is the information available from, and the assessments made by, the various bodies responsible for gaining scientific understanding on climate. These include the Meteorological Office in the UK, NASA and the NOAA in the US, and the IPCC, the inter-governmental panel charged with providing summaries and synthesis of the totality of the available research. The briefings provided by other bodies, such as The Royal Society and The American Academy of Sciences, represent perspectives endorsed by bodies representing science and engaged in maintaining standards with respect to science and public policy.  Finally there are the positions on climate science taken by other entities engaged with the sector, who will also have considered the science.  These include the International Energy Agency, the UK Committee on Climate Change, and others.

All these bodies endorse what one might loosely call the “informed consensus” on climate change. However it is worth testing opinions and judgements against some of the known sceptic arguments which dispute the “mainstream” or “consensus” view. Not least the quality of the arguments advanced in the public debate may be a guide to the confidence that should be attached to the mainstream view.

In December 2014 the Royal Society published A Short Guide to Climate Science, a layman’s introduction to the subject covering 20 separate questions on climate issues.  It may be generally assumed that, on this subject, the Society represents a wide body of opinion in the science community that broadly endorses the view that anthropogenic or man-made climate change is a demonstrable fact, a major threat and hence a challenge for public policy. This short guide was in turn condensed from a longer document.[1]

The Global Warming Policy Foundation (GWPF), in contrast, is associated with a much more sceptical view of the underlying science associated with this position, and commissioned a response to the “Short Guide” from a number of academics dealing with its content on a point by point basis, and offering what it described as a “fuller picture”.[2] The response does not represent a corporate position of the GWPF itself, but the authors are all members of the Academic Advisory Council of the GWPF.

This exchange therefore provides an opportunity to examine directly both sides of the case, without going into areas of ongoing scientific inquiry. We can conduct this examination as “informed” lay opinion, starting from the understanding that we already have of the fundamentals of climate science and the basic points at issue. To this we can add any professional familiarity with, for example, the interpretation of statistical trends, with the use of mathematical models, and with some of the general science that is a necessary part of understanding the issues in the first place.
This paper therefore takes each of the 20 summary points in the Royal Society guide in turn, together with the “fuller picture” presented by the GWPF, and adds our commentary. In some cases we have simply commented on the difference of view between the two, or drawn attention to matters of logic and consistency, or to assumptions implicit in the comment.  In others we have drawn on a wider range of opinion, including for example our own reading of the IPCC or other reports, or have explored the literature to a limited degree to determine whether a particular issue is the subject of a consensus or not. There are also a few points that, even if implicit, are not brought out fully or sufficiently clearly in the exchange between the RS and the GWPF authors. All these points are in our view essential to understanding the discussion and the wider policy issues.


The first observation is on measurement. In essence the proposition that the planet is now in long term warming mode is a statement about heat balances, with the Earth now being a net gainer of heat. To measure this directly through the temperatures of every physical component of the planet is impossible. The best measures we can currently hope for focus on surface temperatures, essentially air and sea surface temperatures, using sample observations to calculate averages across chaotic weather systems. These temperatures are also those that are of most immediate relevance to us in terms of weather and climate, but they are not the only elements relevant to understanding climate processes.
In heat balance terms it is estimated that changes in this measure are only capturing at most 10% of the change and are attempting to do so with sampling points that sit within chaotic weather systems. This means there are two major sources of measurement error and hence variability in temperature data. One is simply the sampling problem. The other is heat exchanges between air, land and ocean, and within those categories, which may be subject to quite substantial dynamic changes over quite long periods. 

The second is time lags that stem from the thermal inertia of a very large mass of land and ocean. A hypothetical permanent step change increase in greenhouse gases (GHG) concentration would tend, all else equal, to result in gradual warming until a new overall equilibrium was reached. Of course finding a way to measure time lags against a background of constantly changing GHG may be impossibly complex, but prima facie one might expect it to be measured in decades. One source suggests at least 25 years. Inter alia this makes it harder to gauge simple sensitivities of temperature to GHG increase. It also implies that, at any given time, more future change is already “baked in” than is already apparent.

The third is irreversibility. This does feature in discussion of how quickly climate change  might reverse if GHG emissions halted. But its main relevance ought to be in the area of policy. Theories of decision making under uncertainty (eg options theory), and indeed ordinary common sense, tend to counsel against irreversible decisions with potentially high cost. The cumulative nature of CO2, which seems to be generally accepted, makes current emissions an irreversible choice, to an extent that cannot be claimed for most economic and policy choices.

The fourth is modelling. The weaknesses of modelling are well known and most stem from the near impossibility of modelling systems of extreme complexity, as well as from significant areas of imperfect knowledge. This stems not just from the complexity of the numerous physical, chemical and biological processes involved, but from its combination with features of physical geography that have no simple mathematical representation (land and ocean relief maps and coastal indentation). Moreover, in the context of testing climate hypotheses, the outputs are being compared against observations that themselves represent only a small and imperfectly measured part of the general hypothesis of GHG induced change.

However at a more general level, models are doing no more than represent known or reasonably well estimated physical parameters (eg “Newton’s Laws of Motion” or the physics of heat transfer). Minimal models might be said consist of very simple arithmetic calculations, with obvious limitations, which may then expand to attempt to capture more of reality with more complex interactions. It is hard to see how progress on the key questions could be made without some form of arithmetic or mathematical calculation, ie modelling.

The fifth is the notion of a “mainstream” of “consensus science”.  Part of the issue for a non-specialist is to judge where this consensus rests. In most areas we tend to put most weight on findings endorsed by bodies with explicit responsibilities for work in the relevant field, including national bodies engaged with weather or climate matters. But none of these should be above challenge on such an important issue.

In this context we should take note the role of the Inter-Governmental Panel on Climate Change, the IPCC, to which we refer on a few occasions in our comments. The IPCC is not an international body conducting its own research. It is an inter-governmental body, closely supervised by representatives of national governments, with its most significant reports approved in plenary session.  Its role is to review the body of available research on climate within this framework, and it should be interpreted within this context.

One clear risk in the presentation of IPCC synthesis reports is that they present an unduly cautious picture of climate risks, based on a lowest common denominator of what can be agreed in plenary session of representatives from participating governments, rather than the best or most respected science.  From a risk perspective, and given the interests of producer and other nations with a strong interest in continuation of fossil based energy policies, this risk must be treated as seriously as any risk that climate threats are overstated.

The IPCC role differs therefore from that of individual bodies charged with climate and related research, such as NASA, or with national academies of science. But its reports constitute one of the most useful repositories summarising the current state of knowledge across a wide spectrum of climate science.

 Finally there is risk and uncertainty, a matter of particular concern to the actuarial profession, and one in which they have longstanding skill and experience.  We note that the subject will continue to include substantial uncertainties, our knowledge is incomplete, as it is in most areas of science, including those that form the basis of most of our daily lives. Moreover we cannot predict accurately the human or economic consequences of climate change, although we can identify very serious and possibly even catastrophic risks. But residual uncertainties do not detract from the validity of the science as the best guide to policy and decision making now, and to identification of what may be very dangerous risks for mankind as a whole.


This commentary details the 20 points included in the Royal Society briefing, together with the GWPF authors response. For each point it aims to comment on the argument as it might appear from an informed lay perspective, either comparing the arguments or checking on the extent to which particular statements.

1. Is the climate warming?

 Royal Society:

 Yes. Earth’s average surface air temperature has increased by about 0.8◦C (1.4◦F) since 1900, with much of this increase taking place since the mid-1970s. A wide range of other observations such as sea-level rise, reduced Arctic sea ice extent and increased ocean heat content provide incontrovertible evidence of a warming Earth.

 GWPF authors. A fuller picture:

 This is hardly an important question. The Earth’s surface is always warming or cooling, or on some occasions barely changing. What is important is that the change referred to is small and imperfectly measured. It should also be stressed that the Royal Society guide does not mention the role of the time window they are using for comparison. The climate has cooled since the mid-Holocene climatic optimum 8,000 years ago, and the warming of the past few decades is relatively small in comparison.

 Surface temperatures have increased on average by about 0.8C since 1900. There was a rise of around 0.5C at the start of the twentieth century, followed by a small fall from 1940 to 1970. From then until the late 1990s temperatures rose by around 0.5C. Differences of a tenth of a degree are insignificant. The temperature is virtually unchanged from that at the beginning of the century. The two periods of increase are indistinguishable, although the earlier increase cannot be attributed to increased carbon dioxide.

 The relation of other observations such as sea-level rise, Arctic sea ice extent and ocean heat content all depend on more factors than global mean temperature, and are hardly incontrovertible evidence of warming. That said, the possible acceleration of ocean heat content accumulation and sea level rise are close to the limits of our ability to detect and the values involved cannot be reconciled to each other. Depending on the time scale, other observational datasets are still more equivocal: global sea ice levels declined for several decades but are now above their long-term mean.


 “This is a hardly an important question.” 

 The GWPF authors response appears to disregard the issue at stake, which is not whether a small change matters per se but whether it tends to support, or not, the hypothesis that there is a long term threat from a much larger change. Interpretation and attribution of the measured warming is therefore important and the subsequent responses of the authors pay a lot of attention to it.

The profound threat described by the Royal Society, and that “mainstream” climate science claims to identify, is based on the analysis of real physical phenomena, notably the undisputed physics of the greenhouse effect itself and the empirical reality of rising CO2 concentrations. Gauging the extent of that threat, and the validity of the climate predictions, must surely be an important task, to which observations of actual temperature change will be an important contribution. Observations of an equivalent fall would surely have provoked a major re-examination of climate models, and of the state of climate science, and a search for negative feedbacks to better explain climate. Perhaps unfortunately from a human perspective, this is not where we find ourselves.

To be specific, the threat is that we are changing the climate and will in the future face much larger changes both from past and future emissions.  The value of firm evidence to support or dismiss the science will necessarily manifest itself in the first instance as small changes. Waiting for large changes, which will impose much larger consequences, will mean that we continue with decisions that promote increases in GHG that are irreversible. Given the lengthy time lags, these in turn may only be the pre-cursor of further changes already “baked in” to the system. Prima facie this seems irresponsible.

“The climate has cooled since the mid-Holocene climatic optimum 8,000 years ago, and the warming of the past few decades is relatively small in comparison.”
Systematic measurements capable of interpretation as direct measures of global temperature do not pre-date 1850, and earlier periods rely on reconstructions of various kinds, increasing the scope for differences in scientific opinion. This certainly seems to be true of the Holocene period.   At least one paper argues the exact opposite  “In terms of the global average, temperatures were probably colder than present day (depending on estimates of latitude dependence and seasonality in response patterns). While temperatures in the Northern Hemisphere were warmer than average during the summers, the tropics and areas of the Southern Hemisphere were colder than average.”[3]  Note that this comparison  is suitably qualified with the word “probably”.
Even if the statement about Holocene temperatures is correct, it seems largely irrelevant to the argument about whether future significant changes in temperature, from the period in which we now live, and as increases above what would otherwise have occurred, are a matter of importance or not.

 “The temperature is virtually unchanged from that at the beginning of the century (1998).”

This is a key statement in the GWPF briefing since it is treated as a given in a number of the subsequent comments, which rely on it heavily to support later arguments.  However its validity depends mainly on selection of 1998 as the year on which to base interpretation of the most recent period. 1998 was an el Nino year which is associated with higher estimated global average surface temperatures, at least on current methods of measurement. In statistical terms 1998 appears as an outlier in the time series. The point is explained in the fuller RS document. The chart below shows this.

A mischievous and similarly misleading interpretation of this chart, biasing by selecting “outlier” years to support a particular interpretation, could also restate the conclusion as follows: 
There was no discernible change in global average temperature between 1878 (selecting another obvious outlier) and the 1970s. After that it increased sharply.

This alternative interpretation chooses to ignore the significant movements in the first half of the 20th century.  In reality serious science and serious debate should try to avoid bias or “cherry picking” in interpreting time series. It often uses, as most informed readers with a statistical bent will know, smoothing techniques like moving averages to interpret time series.  This aims to reduce underlying statistical error or “noise” in the data. 

Interdecadal averages tell a different story (below).  It is certainly possible that there has been some slowing in the rate of growth, but it is equally possible that the small decline in interdecadal growth reflects statistical noise or minor data bias. Inter alia the inclusion of 2014 and 2015 in the data set (hottest year and hottest YTD respectively) may confirm the impression of a steadier trend.

[source: IPCC]

Finally the GWPF authors dismiss other sources of corroborative evidence. While it is correct to point out that sea levels and Arctic ice are subject to other complex influences, they rather obviously cannot be disregarded as pieces of evidence. Falling sea levels or increasing Antarctic ice mass, for example, would surely have been seen as evidence tending to contradict any warming hypothesis.  Some of these particular questions are discussed in more depth in later comments.

2. How do scientists know that recent climate change is largely caused by human activities?

Royal Society:

Human activity leads to emissions of greenhouse gases (causing warming), and of other pollutants that produce small particles in the atmosphere (which can have both cooling and warming effects). The dominant influence of human activities on recent climate change is clear from an understanding of the basic physics of the greenhouse effect and from comparing the detailed patterns of recent climate change with those expected from different human and natural influences. Only when human influences on the composition of the atmosphere are incorporated can models reproduce observed changes in climate.

GWPF authors. A fuller picture:

The warming effect of greenhouse gases is widely recognised. However, the direct effect is known to be relatively small: about 1◦C for a doubling of carbon dioxide levels. Most of the warming predicted in climate models arises from knock-on effects (‘feedbacks’) associated with changes to cloud cover, atmospheric humidity and so forth. Feedback processes are mostly hypothetical and are therefore much more uncertain, and some may even have cooling effects.

The Royal Society guide claims that models fail to explain recent warming unless they incorporate anthropogenic forcing. This assertion depends on the readily falsifiable claim that models correctly replicate natural variability. Models fail on natural variability, therefore the Royal Society’s claim fails in the real world. However, even if the conclusion were correct, it would still be consistent with the view that the climate is not very sensitive to greenhouse gases since the observed changes have been small (to the point of being indiscernible for the past 15 years).


“Feedback processes are mostly hypothetical ….”

 In the early days of efforts to estimate the impact of CO2 emissions on climate, this was certainly true. Numerous propositions were advanced, very reasonably, on ways that negative “dampening” feedbacks might act as automatic stabilisers, including re-absorption through the carbon cycle and increased plant growth.

Unfortunately some of these, notably reliance on parts of the carbon cycle, seem to have been largely discounted, but there are others that would clearly be positive and, in the absence of countervailing influences, would normally be assumed to accelerate or magnify change. Examples include increases in water vapour content of the atmosphere (for which there appears to be some evidence[4]), the prospect of reduced albedo from falling ice cover (evident in reducing Arctic sea ice), and the risk of methane release from a melting permafrost. Prima facie these are far from being purely hypothetical.

The effect of changes in clouds remains a complex and uncertain issue, with evidence that in different aspects clouds can have both warming and cooling effects but, on the basis of existing knowledge it would be hard to state with confidence that clouds offer the firm prospect of a substantial negative feedback. At the very least there must be a substantial probability that it does not provide a significant stabilising feedback.

This assertion depends on the readily falsifiable claim that models correctly replicate natural variability. Models fail on natural variability, therefore the Royal Society’s claim fails in the real world.”

Models typically generate a range of possible outcomes, based on different but credible initial or boundary conditions for particular physical parameters, and other uncertainties within chaotic systems, as well as limitations to the granularity or complexity of the modelling. If outcomes continue to lie within the predicted range, they can be said, as the Royal Society claims, to “reproduce” outcomes, and do not ”fail”.  This can obviously be a rather undemanding test under some circumstances, but it is combined with the claim that no alternative explanation has been proposed which is capable of providing a physical explanation of the significant observed changes in recent years.

In principle there should be numerous possible candidates for alternative “explanations”, including solar activity, changes in orbit, or changes in volcanic activity. So far none of these appears to be capable of providing a convincing physical explanation of the changing climate in recent years (post 1970). A prima facie case certainly exists more generally for solar activity, but there has been no change in solar radiation post 1970.  GHG on the other hand do meet the criterion for explaining this period. There is a known physical mechanism and a known increase in the quantity of GHG.

Given that natural variability (over all timescales) and measurement error exist and are recognised as part of the overall picture, models do not “fail” through inability to “explain” the totality of natural variability. Inability to explain everything does not imply inability to explain anything.

“It would still be consistent with the view that the climate is not very sensitive to greenhouse gases since the observed changes have been small (to the point of being indiscernible for the past 15 years).”

Consistency might well be possible if relevant observed changes were small, but a demonstration that climate is “not very sensitive”  depends on two propositions, first that there has been “indiscernible” temperature increase in recent years, and second that we can ignore any time lags due to thermal inertia of the planet as a whole. The first proposition has been discussed above and is clearly false.  The second is highly improbable.  Thermal inertia for the planet as a whole implies substantial time lags [5] before even the first round effects of a hypothetical step change in CO2 would have fully worked through to a new equilibrium of atmospheric concentration and ocean, land and atmospheric temperatures. It follows that we are not yet observing, and may still be decades away from experiencing, the full impact of the increase to even the current level of GHG concentration. 

3.  Carbon dioxide is already in the atmosphere naturally, so why are emissions from human activity significant?

 Royal Society:

Human activities have significantly disturbed the natural carbon cycle by extracting long-buried fossil fuels and burning them for energy, thus releasing CO2 to the atmosphere. The concentration of CO2 has increased by 40% since the Industrial Revolution.

GWPF authors. A fuller picture:

Carbon dioxide levels have been increasing steadily. A body of evidence points to this being due to human effects – emissions from burning of fossil fuels and land-use changes – although the Earth’s carbon dioxide budget is not sufficiently understood to accurately quantify the human and natural contributions. Natural fluxes in the carbon cycle are an order of magnitude higher than manmade emissions, so any natural imbalances, perhaps as a result of temperature changes, can swamp human contributions. Regardless, given the aforementioned evidence that the sensitivity to carbon dioxide is low, anthropogenic GHGs cannot by themselves explain 20th century warming.


 “… the Earth’s carbon dioxide budget is not sufficiently understood. Natural fluxes in the carbon cycle are an order of magnitude higher.”

 Prima facie this is very misleading. Annual emissions may indeed be small (c 3%) in relation to the large annual flux, but unless they are re-absorbed within the natural carbon cycle, they are net additions, and are cumulative. This means that they have and will become significant as they build up over several decades.

 The fact that we are now 40% above the observed 800,000 year range suggests substantial anthropogenic influence and that there has been limited re-absorption of any excess above natural carbon cycle levels.  Exact quantification would obviously require a counterfactual which could not be directly verified.

 “so any natural imbalances, perhaps as a result of temperature changes, can swamp human contributions”

Given that the GWPF authors are dismissive of amplifying feedback effects, necessarily so in order to claim low climate sensitivity, this suggestion is strange. Having earlier dismissed feedback mechanisms as purely hypothetical, the authors now introduce, admittedly hypothetically, a powerful feedback from temperature rise to net CO2 gain. If this feedback exists, and in principle at least it is quite credible that some changes could reduce the absorption capacity of carbon sinks (eg through impacts on vegetation), then it is akin to the methane release which is a major risk factor for “worse than predicted “ outcomes, which the authors later dismiss in 19 below. It also undermines the case for limited feedbacks described in 2 above.

 “given the aforementioned evidence that the sensitivity to carbon dioxide is low, anthropogenic GHGs cannot by themselves explain 20th century warming.”

The GWPF authors have not shown low sensitivity, for the reasons discussed in 1 and 2 above, and CO2 does not have to explain everything. The authors introduce an alternative factor, solar variations, in the next section, suggesting it has a significant role in the 20th century.  Prima facie there does appear to be at least a superficial statistical correlation in the early 20th century. This might have been argued to complement the explanation of a larger CO2 driven warming post 1970.  Given that the authors have introduced solar variations as an explanatory factor, one might ask whether this “two factor” explanation has been investigated and endorsed or discarded by climate specialists. This is discussed again below in 4.

4. What role has the Sun played in climate change in recent decades?
Royal Society:

The Sun has not played a major role in recent climate change. The Sun provides the primary source of energy driving Earth’s climate system and variations in the energy emitted by the Sun affect Earth’s climate. However, satellite measurements since the late 1970s show no overall increase in the energy emitted by the Sun, while the climate system has warmed.

GWPF authors. A fuller picture:

It is frequently claimed that the Sun has not played a major role in recent climate change because the overall energy emitted by the sun has changed little. This is simplistic. There is significant evidence that the Sun has played an important role in climate change, and over the 20th century in particular. Quantifications of these changes suggest forcing comparable to anthropogenic forcing. While variability of total solar irradiance may be small, variability of specific components of solar output can be large, and some of these are believed to affect the climate through mechanisms other than direct heating, for example by influencing cloud formation. These effects are a matter of current inquiry.


The Royal Society makes its claim only in relation to recent climate change, based on satellite measurements since the 1970s. The GWPF authors discuss a possible role in relation to 20th century warming (as a whole), and prima facie this does seem to be a hypothesis worth considering. The chart below [6] shows at least a superficial correlation between solar activity and temperature for the first part of the century, although it breaks down thereafter, and as the Royal Society says, does not appear to explain post 1970 changes.

 If we were to accept this connection as adequately supported by the evidence, then we would seem to have an explanation of the first part of the century with solar activity as possibly the major influence, with GHG predominating thereafter. In other words this would appear to deal with the alleged failure of climate science to explain the whole of 20th century warming in terms of anthropogenic GHG.  And the combined effect of both factors would appear to offer a powerful explanation of the whole of the century.

 A non-scientist might however be forgiven for remaining agnostic on the solar connection, at least in the absence of demonstrable direct radiation or other explicable physical effects. The authors suggest some specific components of solar radiation are believed to be important, but it is not suggested that these beliefs are anything other than unspecified hypotheses. Generally they have decried hypotheses in relation to feedbacks, some of which seem prima facie more substantial.
For these reasons, it is hard to see why acceptance of the potential influence of the sun should diminish either the credibility of GHG related theory or its significance for the future.

 5.  What do changes in the vertical structure of atmospheric temperature – from the surface up to the stratosphere – tell us about the causes of recent temperature change?
 Royal Society
 The observed warming in the lower atmosphere and cooling higher up in the stratosphere is the result expected from increases in CO2 and decreases in stratospheric ozone. Natural factors alone cannot explain the observed changes.

GWPF authors. A fuller picture:

 Not so: basic physics implies that increasing levels of carbon dioxide will lead to increased cooling in the stratosphere. This is quite separate from the greenhouse impact in the troposphere of increased carbon dioxide. However, measurements in the stratosphere indicate that although the overall trend is down, any cooling is only seen in the immediate aftermath of volcanic eruptions. Between such eruptions, stratospheric temperatures have been rising. This merely indicates that carbon dioxide levels here as elsewhere are not the only factor determining temperature. Similarly, temperatures in the troposphere over the tropics are predicted to rise faster than anywhere else, including at the surface. This too is a matter of basic physics, where the temperature profile follows what is known as the moist adiabat. Models are, indeed, consistent with this. However, observed warming in the tropical troposphere is very weak compared to warming at the surface, suggesting problems with observations at the surface or in the troposphere or both. Given the small changes that are being studied, neither possibility is implausible.


The fuller RS overview is more nuanced. “It is now known that the observed pattern of tropospheric warming and stratospheric cooling over the past 30 to 40 years is broadly consistent with computer model simulations that include increases in CO2 and decreases in stratospheric ozone, each caused by human activities. The observed pattern is not consistent with purely natural changes in the Sun’s energy output, volcanic activity, or natural climate variations such as El Niño and La Niña. Despite this agreement between the global-scale patterns of modelled and observed atmospheric temperature change, there are still some differences.“

 Similarly, and unlike some sceptic commentators, the authors treat this as essentially uncertain – “neither possibility is implausible”.  Elsewhere it has been claimed that elements in this controversy are critical tests of the validity of the mainstream science, but given the nature of the multiple uncertainties regarding both inadequate data and disputed understanding of the processes involved, there is little reason to suppose that much more can be said at this stage.  [More information may well at some stage provide further confirmation.]

 Without further research and exposition on the subject, it is not a subject on which it is currently easy or appropriate for a non-specialist to comment without significant further reading. We are not aware currently of a final compelling resolution of these differences in analysis and opinion.

Nothing in a limited search does other than tend to confirm that this is still a complex issue, governed by incomplete information and understanding, on which opinions differ, and that there has been a paucity of data.  At least one recent article[7],  has cited new data and in the process cast doubt on earlier general  understanding of these issues, while another[8] casts doubt on the view that volcanic activity can explain the changes. Overall, volcanic activity is believed to promote cooling, and is one of the factors contributing to natural variability.

Overall this appears to be an issue of secondary significance, from either perspective.

 6 Climate is always changing. Why is climate change of concern now?

 Royal Society:

 All major climate changes, including natural ones, are disruptive. Past climate changes led to extinction of many species, population migrations, and pronounced changes in the land surface and in ocean circulation. The speed of the current climate change makes it more difficult for human societies and the natural world to adapt.

GWPF authors A fuller picture:

The Earth has many and hugely varied climates. The climate also changes naturally on every timescale. Mankind is remarkably adaptable, living in almost all of these climates. It is impossible to know how rapidly climate changed in the distant past since the time resolution of the data we have is mostly inadequate for resolving the timescales that we are currently concerned about. However, there is ample evidence of rapid climate change associated with cold periods during the most recent glaciation (more than 12,000 years ago). Climate change is only one concern among many at the present time, and a disproportionate focus on it and its possible impacts detracts from our ability to address many other more pressing matters.


Historic evidence of actual climate change in the distant past does not relate to any period with a comparable need to feed 9 billion people.  Nor does history tell us in any detail what the many impacts of climate change may have been, even then, in terms of conflict or population decline. (Although in some instances modest changes in climate have been linked by historians to migrations, disease and population decline.)

The myth of Prometheus Unbound may suggest that mankind, unlike other species, is infinitely adaptable, but science and common sense suggest that ultimately there are or may be physical and biological limits to what is achievable, and to changes which will continue to accommodate 9 billion people.

“only one concern among many”

It is certainly true that there are plenty of other pressing and dangerous problems, such as the risk of global pandemics or the problem of antibiotic resistance, as well as the widespread nature of human conflict.  Many of these are intimately linked to climate change in the sense that they may be intensified by climate phenomena, particularly if climate affects land, water or other resource pressures – all of which are natural sources of conflict.

A distinguishing and fundamental feature of the climate issue is that decisions taken now – to continue high levels of GHG emissions, are essentially irreversible on less than geological timescales. This is much less true of the other threats.  Failures in planning now for pandemics or ebola type outbreaks, can, and almost certainly will, be remedied in response to a major public health incident.  Failure to take mitigating action in relation to climate simply means building in an increasing problem, and with less time to adapt to the consequences.

Finally the dichotomy is a false one. It is not obvious why attention to climate policy should prevent efforts to deal with other pressing dangers. It may indeed complement them, or be a necessary precondition to managing these other risks in the longer term.

 7.  Is the current level of atmospheric CO2 concentration unprecedented in Earth’s history?

 Royal Society:

The present level of atmospheric CO2 concentration is almost certainly unprecedented in the past million years, during which time modern humans evolved and societies developed. The atmospheric CO2 concentration was however higher many millions of years ago, at which time temperatures and sea levels were also higher than they are today.

GWPF authors. A fuller picture:

While carbon dioxide levels appear to be higher than they have been for hundreds of thousands of years, they are relatively low compared to most of the last 600 million years (when most lifeforms evolved), during which time levels were often from 2–20 times greater than today. Counter to the Royal Society, there were periods during which the carbon dioxide level was as much as 10 times higher than today but the climate was colder, for example the Silurian Period (about 443–420 million years ago). The fact that most plant life evolved during these periods is because plants thrive when carbon dioxide is increased. Moreover, our present estimates of carbon dioxide variations over the past 700,000 years are based on the analysis of ice cores, and these analyses may have inadequately dealt with diffusion, which could cause major adjustments to our estimates of early carbon dioxide levels.

The Royal Society statement focuses on the last million years.  The GWPF authors in contrast focus on astronomical measures of time when the Sun and Earth would have been very different, in respect for example of the strength of solar radiation – a different strength being consistent with higher “equilibrium” CO2 levels. So this is not a direct “counter” to the Royal Society.

The observation that the first life forms evolved in these earlier periods tells us little about how today’s humans, and today’s flora and fauna, would cope with what is (in geological terms) rapid change. Today’s life will have evolved to meet the environment of the last million years, or in many instances much less, rather than that of the last 400 million.

 The Royal Society confines itself to a period that includes human history, and points out the evidence, presumably based on reconstruction from ice core samples, that actual CO2 has remained in a comparatively narrow range. The authors dispute this with the speculative possibility that there may have been flaws in this analysis, but it is not evident that there is any empirical support for this possibility.

8.  Is there a point at which adding more carbon dioxide will not cause further warming?

 Royal Society:

 No. Adding more CO2 to the atmosphere will cause surface temperatures to continue to increase. The addition of extra CO2 becomes progressively less effective at trapping Earth’s energy, but surface temperature will still rise.

GWPF authors. A fuller picture:

 Each additional increase of carbon dioxide levels is expected to produce less and less greenhouse warming, so it takes far more emissions to produce the second degree of warming than the first. Thus unless carbon dioxide emissions rise exponentially in the long term, warming should slow down. In theory temperatures will always keep rising, but eventually at a rate indistinguishable from zero. As usual, the question is not about warming per se but about how much warming there will be compared to natural variability. The available evidence is entirely consistent with the answer ‘not much.’


The relationship is a log-linear function. If doubling were to produce one unit of warming, then doubling again would produce one more unit. This is potentially quite dramatic as a stabilising mechanism, so it is worth discussing and the GWPF authors’ comment is informative.

The main problem, intuitively and computationally, lies in establishing where we are in the process, given the very long time lags to which we referred earlier, and the fact that CO2 concentration does not proceed through step changes followed by decades when it is static. Today’s temperature cannot be interpreted as the outcome of today’s CO2 concentration.  It is the outcome of concentrations over previous decades.  Similarly we can only estimate (eg through climate modelling) what future “equilibrium” temperature, holding all else equal, would result from holding CO2 concentration at today’s level.

Recognition of a time lag effect will inevitably amplify a sensitivity calculated as if the time lag did not exist. And estimates of the time lag span several decades.[9]

It is worth noting that the loglinear effect does not necessarily translate through to feedback effects; these may move towards equilibrium at faster or slower rates.  If we look instead at the damage function, the impact from temperature rise in terms of adverse consequences of change are likely to increase faster as temperature rises.  Many if not most natural systems have threshold effects, where they can tolerate change or variation, but only up to a certain threshold.

9.  Does the rate of warming vary from one decade to another?

Royal Society:
Yes. The observed warming rate has varied from year to year, decade to decade, and place to place. These shorter-term variations are mostly due to natural causes, and do not contradict our fundamental understanding that the long-term warming trend since the mid-20th century is primarily due to human-induced changes in the atmospheric levels of CO2 and other greenhouse gases.

GWPF authors. A fuller picture:

Temperature and many other climatic measurements vary naturally on all timescales, from decades to centuries and longer. Long periods in which the climate warms or cools naturally are therefore to be expected. Because climate models do not incorporate all of the different factors that might affect the climate, many of which are as yet unquantified, unequivocal attribution of recent warming is not possible, although at least part of it may be due to human emissions of greenhouse gases.

A major error in modelling is the failure to account for natural variability. For the Royal Society to use this variation as an excuse for the obvious mismatches in models is strange indeed. Rationally, the fact that current models have greatly overestimated observed warming would suggest that models are too sensitive – a possibility that the Royal Society should have pointed out.


“unequivocal attribution of recent warming is not possible”

This is likely to be the case in any system with natural variation, and the Royal Society says cautiously that this is primarily due to human-induced changes, while the GWPF authors acknowledge that part of it may be so.

“have greatly overestimated observed warming”

 Prima facie, as argued earlier, this is an exaggeration based on selective interpretation of recent trends. It has also been suggested that adjustments to historic data, in the light of improved knowledge, may also make the models appear more accurate.

 Recent findings, described in more detail below, suggest the possibility of direct verification of the sensitivities used in the models.

 “A major error in modelling is the failure to account for natural variability.”

 This reflects a different conception of the nature and function of modelling. Climate models can project forced trends but they can only simulate stochastic (random) variability. This means they cannot be expected to predict global temperatures accurately over short periods, since these are dominated by stochastic and unforced variability. This is discussed again in 10 below.
Empirical verification

 Hitherto it has not been possible to obtain direct empirical verification of the calculations used in climate models. But according to a recent (2015) report,

 “Scientists have observed an increase in carbon dioxide’s greenhouse effect at the Earth’s surface for the first time. The researchers, led by scientists from the US Department of Energy’s Lawrence Berkeley National Laboratory (Berkeley Lab), measured atmospheric carbon dioxide’s increasing capacity to absorb thermal radiation emitted from the Earth’s surface over an eleven-year period at two locations in North America. They attributed this upward trend to rising CO2 levels from fossil fuel emissions.

 The influence of atmospheric CO2 on the balance between incoming energy from the Sun and outgoing heat from the Earth (also called the planet’s energy balance) is well established. But this effect has not been experimentally confirmed outside the laboratory until now. The research was reported Wednesday, Feb. 25, in the advance online publication of the journal Nature, and is claimed to detect the unique spectral signature of infrared energy from CO2.

The results agree with theoretical predictions of the greenhouse effect due to human activity. The research also provides further confirmation that the calculations used in today’s climate models are on track when it comes to representing the impact of CO2.”[10]

10.  Does the recent slowdown of warming mean that climate change is no longer happening?

 Royal Society:

No. Since the very warm surface temperatures of 1998 which followed the strong 1997–98 El Niño, the increase in average surface temperature has slowed relative to the previous decade of rapid temperature increases, with more of the excess heat being stored in the oceans. Despite the slower rate of warming, the surface temperatures in the 2000s were on average warmer than the 1990s.

GWPF authors. A fuller picture:

Surface temperatures have exhibited no warming since the start of the century. Weather satellite records suggest the pause has been going on even longer. The reasons for the pause are unknown. Numerous explanations have been proposed, the most high profile being a suggestion that the missing heat has found its way to the deep ocean. However, this is simply an obscure way of blaming natural internal variability, for which the ocean circulations (which are always exchanging heat between surface and deep water) are a major cause. There is no known way to distinguish these natural exchanges from the notion that ‘heat is hiding in the ocean’. What we do know is that these major ocean circulations are not correctly captured in the current climate models.


The GWPF authors repeat the assertion criticised earlier.  Nevertheless, although the GWPF authors have exaggerated the slowdown in warming, by focusing on 1998 as a starting point for comparison, it was true that there appears to have been some slowing in the rate of warming at least of surface temperatures. However since these were well within the range of outturns projected by the main models it would be foolish to conclude that the model sensitivities were wrong without further examination. Moreover the actual temperature observations for 2014 and 2015 make it much harder to justify the GWPF interpretations of the data that were already questionable. The numbers for these years suggest an objective interpretation much closer to the inter-decadal trend shown above.

 Perhaps the most important point to understand is that climate models project forced trends but only simulate stochastic (random) variability. This means that they should not be expected to accurately predict global temperatures over short periods, which are dominated by stochastic and unforced variability.

One possible test is to examine the performance of the models over much longer periods, to see whether there is any systematic bias or failure.  One recent paper[11] looks back over the past century and claims that global average temperature is extremely well captured by models, with no obvious pattern showing temperature trends are closer to the upper or lower end of model projection. In other words, while the models may not capture all the natural variability along the path of rising temperatures, they are slightly too cool just as often as they're slightly too warm.

 To avoid circularity, if some of the models were based on statistical estimation over the same period, one would require the test to be carried out over a different period, but to the extent that models use physical parameters estimated independently of the temperature data, this is less of an issue.

Ideally the global warming hypothesis should be expressed as a global heat gain balance, of which we know surface atmospheric temperatures are likely to be a very small part. Intuitively one would expect the much larger quantities involved to show much more stable  trends in net heat gain or loss. It would also eliminate much of the time lag issue which we have identified as very important.
Worryingly, in terms of concern about warming, there have been proposals that, using surface temperatures only, data revisions are necessary to reflect a changing appreciation of the details of climate effects, such as adequate sampling of polar regions. Some of these suggest that the recent hiatus in warming has been overstated.

 11.  If the world is warming, why are some winters and summers still very cold?

Royal Society: Global warming is a long-term trend, but that does not mean that every year will be warmer than the previous one. Day to day and year to year changes in weather patterns will continue to produce some unusually cold days and nights, and winters and summers, even as the climate warms.

GWPF authors.  A fuller picture:

Global warming refers to a long-term trend – spanning periods of decades to centuries – which has been very small compared to the weather, which varies a great deal from year to year. Cold weather is therefore not unexpected in a warm climate. By the same token, one should expect warmer episodes from time to time even if there were no global warming.


 No real difference of view is discernible in these observations, since this is essentially a statement drawing a necessary distinction between climate and weather. The GWPF authors correctly point out the converse, warmer episodes in times that are stable or cooling.

12 Why is Arctic sea ice reducing while Antarctic sea ice is not?

Royal Society:

 Sea ice extent is affected by winds and ocean currents as well as temperature. Sea ice in the partly-enclosed Arctic Ocean seems to be responding directly to warming, while changes in winds and in the ocean seem to be dominating the patterns of sea ice change in the Southern Ocean around Antarctica.

 GWPF authors A fuller picture:

 There is no basis for the assertion that winds are less important in the Arctic, and evidence exists that summer sea ice has often been low. Most climate models predict fast reductions in both Arctic and Antarctic sea ice, although the two are very different systems. The steady record rise in Antarctic sea ice is therefore not predicted by models, although there has been some speculation as to the reasons for the failure.  In any case, Arctic ice remains fully with us in winter despite summer lows. Even in the height of summer substantial ice remains (many millions of square kilometres).


The real distinctions are not brought out fully by either party in the above, abbreviated, discussion. As the GWPF authors say, these are very different systems.

 Sea ice is the more important in the Arctic. Melting sea ice does not increase sea levels, but is significant as a potential climate feedback because it reduces albedo in summer, ie causes the oceans to absorb more heat.

Land ice is a much more relevant factor in the Antarctic, because of the sheer quantity of ice on the Antarctic land mass. Melting land ice matters as a consequence effect of warming, because it does raise sea levels. 

Viewed purely as indicators or measures of actual warming, the target measure has to be to an estimate of changes in total heat content for the planet. In very simple terms Arctic sea ice is a useful indicator, but in the Antarctic, any changes in land ice (including temperature change within the land ice) need to be taken into the equation as well. (And of course temperatures in the immediately surrounding oceans matter too.)

Unfortunately available measures of Antarctic land ice appear to show significant reductions in volume, a further indicator of rising heat content for the planet as a whole.

Viewed as feedback effects and consequences, the albedo effect and the consequences for sea levels are probably the key issues.

“Arctic ice remains fully with us in winter despite summer lows.”

This is very little comfort since the main impact of the albedo effect is in summer. It might be thought that additional Antarctic sea level ice might be a countervailing effect, but if this were predominantly a winter effect then it would be less important than in the Arctic. It would be useful to confirm whether this is a reason and why it has not attracted much coverage.

The West Antarctic ice sheet has been identified as posing a major risk of much higher sea level rises, but this possibility is not included in the mainstream IPCC projections.

 13 How does climate change affect the strength and frequency of floods, droughts, hurricanes and tornadoes?

Royal Society:

Earth’s lower atmosphere is becoming warmer and moister as a result of human-emitted greenhouse gases. This means that more water is likely to be drawn into major rain storms, which could lead to more flooding events. There is considerable uncertainty over changes in hurricanes and tornadoes, but the extra energy available may make the strongest hurricanes stronger. Dry areas of the subtropics are expected to become drier in the future.

GWPF authors A fuller picture:

Climate models can have little to say about what happens below their level of resolution, which remains coarser than most storminess. There is no evidence of any increase in either intensity or frequency during the recent period of global temperature average rises. In fact, there has been a remarkable lack of landfalling hurricanes in the Atlantic. Tornadoes are unlikely to be affected by any global warming.

Extra energy does not cause storms. Nor does it necessarily increase their strength. Energy differences and gradients cause storminess. Changes in internal energy and moisture that do not affect gradients and differences can have little effect. Speculation that wet areas become wetter and dry areas become drier are claims about increases in gradients and differences, which the global warming hypothesis does not contain. In fact models call for a decrease in gradients between equator and poles, which would imply a reduction in storminess. Drought levels have, if anything, fallen worldwide in recent decades and there is little evidence of global changes in floods.


The IPCC, which attempts to summarise and synthesise the body of climate research, and whose reports are agreed by government representatives in plenary session, makes statements stronger than the rather cautious Royal Society guide statement above. It considers that some of the changes in extreme weather and climate events observed since about 1950 have been linked to human influence. Other individual climate scientists have expressed similar views.

“a decrease in gradients between equator and poles”.

The GWPF authors’ argument here depends on the assumption that it is this polar to equatorial gradient which is relevant to storminess. But this does not appear to be the only credible view of how hurricanes form. To quote NASA, “Tropical cyclones … use warm, moist air as fuel. That is why they form only over warm ocean waters near the equator.” This argument seems to represent quite a fundamental difference from the professional meteorological view.

Intuitively one might therefore expect some link between temperature, ie the levels of heat energy in the tropics, and the kinetic energy in storms.  Higher overall levels of energy, on this more local scale, would then tend to larger local “gradients”.  It is this intuition that the Royal Society seems to be expressing. It would be interesting to hear a contrary view from a meteorologist.

14.  How fast is sea level rising?

Royal Society:

Best estimates of the global-average rise over the last two decades suggest 3.2 mm per year (0.12 inches per year). The overall observed rise since 1901 is about 20 cm (8 inches). If CO2 and other greenhouse gases continue to increase on their current trajectories, it is projected that sea level may rise by a further 0.5 to 1 m (1.5 to 3 ft) by 2100.

GWPF authors.  A fuller picture:

In a warming planet sea levels would necessarily rise due to thermal expansion of the oceans and melting of glaciers and ice sheets. But sea level has been rising for thousands of years – since long before GHG emissions became significant. Claims of an acceleration in sea level rise from 2 to 3mm per year and its attribution to mankind must be treated with caution. In particular, it is not currently possible to reconcile estimates of sea level rise with estimates of the factors that are thought to contribute to it. The picture is even more unclear at the local scale where, depending on the location, many contributions have nothing to do with climate, such as tectonics, vegetation cover, hydrology, etc.


There is no direct contradiction here of the Royal Society position, as set out in their guide, and the GWPF authors’ identification of numerous other factors, and local variations, may be quite correct. However it is not necessary to assume precise attribution of sea level changes in order to regard the subject as potentially important.
First, the measurement of long-term changes in global mean sea levels can provide an important corroboration of predictions by climate models of global warming. This has certainly been a position taken by the NOAA.[12]  Falling sea levels would certainly have been a contrary indication.
Second, calculations for the future cover relatively new phenomena, including melting land ice (Antarctic), associated with radiative forcing and rising temperatures. These could result in a substantial increase in the rate of sea level rise and do form a major concern, with which the GWPF authors do not engage directly.

In terms of risk, some scientists have expressed fears over the stability of Antarctic ice sheets, which if realised could result in substantially higher and faster sea level rises.  Even the smaller rises projected in the RS guide would have very substantial consequences for low lying regions and coastal cities.

15. What is ocean acidification and why does it matter?

Royal Society:

About a quarter of the emissions of carbon dioxide from human activities are soaked up by oceans each year. The extra CO2 causes the chemical balance of seawater to shift to a more acidic state (lower pH) and some corals and shellfish have shells composed of calcium carbonate which dissolves more readily in acid. Acidification is likely to shift the competitive advantage among species, with as-yet-to-be determined impacts on marine ecosystems and the food web.

GWPF authors A fuller picture:

The oceans absorb some of the extra carbon dioxide released into the atmosphere. It would form a weak acid if it were not already mostly alkaline. Human emissions of carbon dioxide will tend to make sea water less alkaline and more chemically neutral. The projected change over the next century is between 0.1 and 0.5 pH units. However, seawater pH naturally varies from 7.5 to 8.5 between regions of the ocean, between habitats, between days, and even between times of day. It is therefore misleading to talk of ‘ocean acidification’. Shallow-water coral reefs are already subjected to hourly, daily and seasonal changes in pH that encompass the full range of ocean variability, hence the effects of changes in pH can be studied. Claims that corals and shellfish will find it harder to grow in acidic water are overly simplistic, not only because the water is not expected to be acidic but because the dissolved carbon dioxide forms bicarbonate and carbonate ions, the raw material for shellfish shells. Most studies find mixed effects, with some groups of organisms thriving as a result of increased dissolved carbon dioxide and some doing less well.


Some scientists appear to attach much more weight to this acidification than does the Royal Society guide, which is rather low key in this briefing.  The shift in competitive advantage between species reflects a very general phenomenon in terms of the natural world and would apply to even the smallest variation in climate. Almost any change will shift competitive balance[13][16], but is not necessarily disastrous except from a very static and unrealistic perspective on conservation of nature (which has always been changing and evolving).

The GWPF authors however appear to be responding to scientists who have expressed much greater and possibly legitimate concern than has the RS on this occasion. On a simple technical point the GWPF authors’ response lacks arithmetic logic. It appears to confuse the measures of average and range. The effect of any change on marine life, if it exists, is prima facie most likely to depend on the average or, possibly, on the extreme to which the organism is subjected.  An increase of 0.5 units (presumed to be an average) will either have a direct effect on the relevant chemical or biological process, or it will change the range from (7.5, 8.5) to (8.0, 9.0).

But neither party appears to treat this as a particularly important point. Seen purely in terms of this exchange, the ordinary reader might be forgiven for remaining agnostic or treating this as of lesser importance than other threats. But, if it is a serious issue, it negates to a significant degree any positive aspects of increased CO2 absorption by the oceans, whether through natural causes or through geo-engineering initiatives.

16.   How confident are scientists that Earth will warm further over the coming century?

Royal Society:

Very confident. If emissions continue on their present trajectory, then warming of 2.6 to 4.8◦C (4.7 to 8.6◦F), in addition to that which has already occurred, would be expected by the end of the 21st century. The range of values accounts for the fact that there are open questions as to how exactly some natural processes such as cloud formation amplify or reduce the direct warming effect of increasing levels of CO2.

GWPF authors A fuller picture:

Increasing carbon dioxide levels are likely to bring some warming. Climate models predict 0.6–1.8◦C by mid-century, but observational evidence indicates that they substantially overestimate how sensitive the climate system is to increasing carbon dioxide levels, and may well also overestimate how much of the emitted greenhouse gases will remain in the atmosphere. The failure of models to make correct predictions over the recent period diminishes our confidence in their ability to make correct predictions of the far distant future.


This repeats the GWPF authors’ earlier assertions on the overestimation of climate sensitivity, and on the weaknesses of modelling, which have been discussed earlier. Quite simply these assertions do not seem to be correct or to be supported by a fuller analysis.

17 Are climate changes of a few degrees a cause for concern?

Royal Society:

Yes. Even though an increase of a few degrees in global average temperature does not sound like much, global average temperature during the last ice age was only about 4 to 5◦C (7 to 9◦F) colder than now. Global warming of just a few degrees will be associated with widespread changes in regional and local temperature and rainfall as well as with increases in some types of extreme weather events. These and other changes (such as sea level rise and storm surges) will have serious impacts on human societies and the natural world.

GWPF authors. A fuller picture:

There is little indication of serious problems in the short-term. Links to extreme weather are not supported by observational, theoretical or even model evidence, and suggestions that rainfall patterns would vary are no more than hypotheses (and counterfactual hypotheses at that). A warmer climate would also bring substantial benefits, for example longer growing seasons and fewer cold-related deaths. Higher carbon dioxide levels will fertilise plants, including many important crops. Estimates of the economic impact of temperature changes suggest little net impact until the temperature is several degrees above pre-industrial levels.

As concerns the ice age comparison, the Royal Society guide is patently absurd. Changes in temperature averages represent effects, not causes. They cannot discriminate between very different processes of change. It is even possible for there to be climate change where the global mean temperature doesn’t change at all. It is well known that the ice ages were driven by huge changes in Arctic insolation in summer (changes that are of the order of 50 times larger than changes in the mean radiative budget), and that changes in mean temperature are simply the small residue of the larger high-latitude changes. Any familiarity with fluid dynamics would show that the Royal Society has things backwards in asserting that the small changes in the mean drive the much larger regional changes.


The fuller RS publication claims that both theory and direct observations confirm that global warming is associated with greater warming over land than oceans, moistening of the atmosphere, shifts in regional precipitation patterns and increases in extreme weather events. There is no obvious reason to assume that changes expressed in temperature terms would be uniform, particularly in complex dynamic systems within which there is little uniformity (eg by latitude) in any case.

“the ice age comparison”

Surely this is intended as a simple illustration that a (relatively) small average temperature difference can be associated with a world that looks completely different.  In this context the causes of ice ages are irrelevant.

“Any familiarity with fluid dynamics would show that the Royal Society has things backwards in asserting that the small changes in the mean drive the much larger regional changes.”

The relevance of fluid dynamics is unclear, since the physics of climate is not confined to fluid dynamics.

Moreover a well-known counter example to the authors’ view is the el Nino climate phenomenon. It is generally held that a “small” change in ocean current temperatures in one part of the world (the NOAA definition is a 3-month average warming of at least 0.5 °C), is associated with or has substantial consequences for climate and rainfall around the globe, occasionally manifesting as or at least accentuating events such as droughts. These may be both beneficial and otherwise, and appear to affect crops and commodity prices.

Others are far less sanguine than the GWPF authors about the future. It is however much harder to be specific about particular regional impacts and their economic consequences. Some of these may well be beneficial, at least on some timescales, as the authors correctly claim. Others, particularly with dramatic effects such as sea level rise, but also lesser changes disrupting agricultural patterns, could write off huge amounts of physical and human capital and real estate.

The real issue here is the nature and scale of the risks that will be imposed on the world as a whole, and the disruptive changes implied.

18.  What are scientists doing to address key uncertainties in our understanding of the climate system?

Royal Society:

Science is a continual process of observation, understanding, modelling, and testing. The prediction of a long-term trend in global warming from increasing greenhouse gases is robust and has been confirmed by a growing body of evidence. Nevertheless, understanding (for example, of cloud dynamics) remains incomplete. All of these are areas of active research.

 GWPF authors.  A fuller picture:

 Scientists continue to address some of the unknowns regarding the climate system. Some are concerned that funding continues to be focused on characterizing human influences on the climate rather than investigating natural variability, and on developing complex computer models rather than improving observational data-gathering systems or fundamental theory. One would be hard-pressed to identify the ‘growing body of evidence’ that the Royal Society guide refers to. Certainly, the evidence of the past 40 years points clearly to exaggeration by existing models. Moreover, the ‘incomplete understanding’ that the Royal Society guide so glibly acknowledges happens to be fundamental to the crucial question of climate sensitivity. ‘Nullius in Verba’, the society’s motto, means we do not even take the word of the Royal Society guide on these things. We don’t.


Again the authors’ comments are assertions that have been discussed earlier. Most readers of IPCC summaries will probably view their reviews as suggesting a “growing body of evidence”, leading to stronger statements, and recent evidence as tending to reduce uncertainties around climate sensitivity. 

The idea that “incomplete understanding” invalidates useful analysis is not self-evident. It is certainly and obviously not true in fields as diverse as quantum theory and evolutionary biology.

The GWPF authors’ views on research priorities might be taken to imply that important areas such as focusing on precise measurement and data accuracy were being neglected, or that the bulk of research consisted of computer model building.  The reality is that research and continuous advance is evident across the board, some of which is instanced in these comments, in areas as diverse as satellite measurement of Antarctic ice and direct measurement of warming.

19.  Are disaster scenarios about tipping points like ‘turning off the Gulf Stream’ and release of methane from the Arctic a cause for concern?

Royal Society:

 Results from the best available climate models do not indicate any abrupt changes or ‘tipping points’ in the climate in the near future. However as warming increases, the possibilities of major abrupt change cannot be ruled out.

GWPF authors.  A fuller picture:

 While these disaster scenarios are raised from time to time and have been discussed in previous IPCC reports, they are largely discounted in the current  IPCC report, as is appropriate.


 These two comments differ mainly in emphasis, with the Royal Society refusing to rule out “tipping points”.

 One of the most notable tipping points is the potential feedback associated with methane release from Arctic permafrost.  Implicitly the Royal Society says this is not indicated from the best current models.  The most recent IPCC reports support a similar view.

 However if we accept that there are significant possibilities of current science and “best estimates” overstating risks, then we should also recognise the possibility  that they are understated. Certainly some scientists continue to maintain that this particular risk has been understated.

20.  If emissions of greenhouse gases were stopped, would the climate return to the conditions of 200 years ago?

 Royal Society:

No. Even if human emissions of greenhouse gases were to suddenly stop, Earth’s surface temperature would not cool and return to the level it was at before the Industrial Revolution for thousands of years because CO2 is only removed from the atmosphere over these very long time scales.

GWPF authors.  A fuller picture:

This question insinuates that there was some sort of ‘steady state’ of the climate before industrialisation. Indeed, before industrialisation, the Earth was in the Little Ice Age, and few would want to return to such a period. However, because the Earth’s climate varies naturally on all known timescales, it is not possible to make definitive statements about what the climate would be like today had there been no manmade greenhouse gas emissions. Similarly, it is not possible to say what will happen in the future, regardless of the levels of future GHG emissions. Moreover, the statement that carbon dioxide is only removed from the atmosphere over timescales of thousands of years is highly misleading. The majority of the excess carbon dioxide over preindustrial levels should be removed from the atmosphere within a century after a sudden halt to emissions, assuming that there are no long-term changes in the natural carbon cycle, which would swamp human contributions even for small changes. It is only a modest proportion that would take thousands of years to be removed.

The Royal Society does not “insinuate” a steady state, and earlier points in the guide discuss other natural variations in climate. 

 “the excess carbon dioxide … should be removed … within a century after a sudden halt to emissions, assuming that there are no long-term changes in the natural carbon cycle”  
CO2 does not “decay” of its own accord and can therefore only be removed as part of processes in the natural cycle. This statement therefore places a remarkable degree of confidence in an implicit ability to predict the future behaviour of the natural carbon cycle, which is in contrast with the authors’ comments on ability to predict future climate more generally. 
Moreover the presumption of continuing net absorption within the natural cycle implicitly accepts that this is a further addition to carbon within the oceans (as a major carbon sink), exacerbating potential adverse impacts there. 
If the carbon cycle itself returned quickly to a new equilibrium position, ie with a zero balance of entry and exit, then CO2 would persist in the atmosphere for a very long time, or indefinitely.  

“it is not possible to make definitive statements …”

This seems to confuse precise statements about “exactly what might have happened otherwise”, which can never be known, with statements about impact or likely impact, albeit imprecise, which can reasonably be deduced from known elements in the science.


This analysis, and the 20 points, have focused primarily on matters of science. Any assessment of the impacts of climate change, its human and economic consequences, and possible policy implications are only covered tangentially.  These conclusions are therefore confined to the discussion of the science.

It should be clear from the above statements and comments that, at least in the context of this exchange and its implications for acceptance or otherwise of the main propositions of mainstream climate science, that there are quite a small number of key issues. Both parties clearly accept the science underpinning the principle that GHG can drive warming, although the GWPF authors seek to qualify this in a number of ways.

The most important single issue, at least in this exchange, is the interpretation of recent (post 1970) data on actual global surface temperatures, both as evidence of an actual impact and as a measure of climate sensitivity to increasing GHG. A neutral party looking at the statistics for the first time, and without any preconceptions as to the theoretical underpinnings, would see a clear increase since 1970, but with a less clear cut trend since about 2000. However attempts to base comparison on an obvious statistical outlier such as 1998, should automatically invite suspicion. Moving or interdecadal averages tell a steadier story. Purely in terms of risk, it would be impossible to conclude from this data that we could rely on the assumption that warming had now paused and would only continue, if at all, at very low levels. (Question 1.)

We might expect that additional data will increasingly tend to resolve any residual doubt on this issue, both through the passage of time, for example if the trends through 2014 and 2015 YTD (hottest on record) are continued, and, perhaps more importantly, with improving data on new approaches to estimating other elements in the global heat balance.

In terms of climate sensitivity to GHG, the discussion fails to bring in time lags and thermal inertia. It should be clear that a simple division of temperature change by CO2 change over the same period, which is implicit in the GWPF arguments, is fallacious, and that such a calculation has the potential to seriously and systematically understate sensitivity, even ignoring selective bias in choice of base years. (Comment on Question 2.)  We might argue that this intuitively rather simple phenomenon, the time lag between changes in GHG concentration and the full effect of that change, is not stressed sufficiently in the RS briefing, and is simply ignored by the GWPF authors. The ratio of committed to observed warming has certainly been stressed by other IPCC authors.

A second issue is whether the GHG increase is predominantly attributable to human activity, but the attempt to cast doubt on this proposition is not supported by any convincing argument or evidence. (See comment on Question 3.)

Other arguments are essentially around questions of corroboration for the basic propositions presented by a mainstream view of CO2 induced warming, and particular consequences such as sea level rise or extreme weather which are important per se.  While some of these indicators and corroborations may not be clear cut, are subject to additional explanatory factors, and fall well short of an absolute proof, they remain as bits of evidence that appear to point consistently in one direction.

Finally there is scope for discussion of rather philosophical questions, on the nature of truth in scientific inquiry. This perhaps deserves a much longer discussion, but the current issues are about what practical judgements are to be made about balances of risk, and what policy measures and what strategies are appropriate to deal with them.

Residual uncertainties and incompleteness in science will remain, but increasingly it appears that these are much less relevant than what we do now know.  The Royal Society statements, which are intended for a wide public consumption and not for more specialist audiences, may well be less than complete, but they do offer at least a good description of a very large, complex and fundamentally important subject.

[1]  Royal Society. Climate Change Evidence & Causes. An overview from the Royal Society and the US National Academy of Sciences. February 2014.
 [2] The Small Print. What the Royal Society Left Out.  The GWPF report was prepared by, and endorsed by the following authors: Professor Robert Carter, Professor Ross McKitrick, Professor Vincent Courtillot, Professor Ian Plimer, Professor Freeman Dyson, Dr Matt Ridley, Professor Christopher Essex, Sir Alan Rudge, Dr Indur Goklany, Professor Nir Shaviv, ProfessorWill Happer, Professor Fritz Vahrenholt, Professor Richard Lindzen.
[3] Kitoh, Akio; Murakami, Shigenori (2002). "Tropical Pacific climate at the mid-Holocene and the Last Glacial Maximum". Paleoceanography 17 (3): 1047.

[11] a collaboration between scientists at the Max Planck Institute in Germany and the University of Leeds, is published in the journal Nature.

Also. Berkeley Lab News Centre.

No comments: