It just so happens that most of the posts on this site have tried to counteract arguments from those who would sow fake “uncertainty” in the climate debate. But lest our readers feel that we are unjustifiably certain about our knowledge, let us look at a recent example of the opposite tendency: too much certainty.
A recent BBC Horizon documentary (transcript) raised the issue of ‘global dimming’ and argued that this ‘killer’ phenomena’s newly-recognised existence would lead to huge re-assessments of future global warming. As part of the hyperbole, the process of global dimming was linked very clearly to the famines in Ethiopia in the 1980s and the implication was left that worse was to come. Media reports with headlines like “Fossil Fuel Curbs May Speed Global Warming” swiftly followed. So what’s the real story?
Global dimming is indicated by measurements over land areas in many regions in the world and may therefore be a real phenomena. Though there are serious issues with the quality of some of the data (birds drinking out of uncovered evaporation pans, drift and inhomogeneities in the solar radiation measuring instruments), in the most global assessment, Beate Liepert estimated that there was globally a reduction of about 4% in solar radiation reaching the ground between 1961 and 1990. While more recent indications are that the trend may have reversed in the last decade, it could still be significant. Assuming for the sake of argument that these data are valid, what could have caused this? A change of that magnitude in the incoming solar radiation itself is not possible since satellite observations would have seen it. Thus, it must be something that is happening in the atmosphere to intercept solar radiation. There are only a few possibilities: clouds, water vapour or aerosols.
First of all it is important to note that even pure greenhouse gas forcing will lead to a slight decrease in surface solar radiation (due to the concurrent increased humidity) and potential cloud feedbacks. Cloud cover and thickness are both like to vary as a function of climate change.
Contrails (those wispy trails left behind high flying jets) have increased over the period and may be important. But estimates of their global effect, even making very generous assumptions about their spread are small (Minnis et al, 2004). Aerosols are also known to have increased over this time, and so they are a natural candidate. However, simulations using the relatively straightforward ‘direct effect’ of aerosols (the increase in albedo of the planet due to the particle brightness) do not match the inferred changes. The final candidates are numerous interactions of aerosols with clouds, the so-called ‘indirect effects’.
There are an ever increasing number of these ‘indirect effects’, but the two most discussed are the aerosol/cloud opacity interaction (more aerosols provide more sites for water to condense in clouds, thus cloud droplets are smaller and clouds become more opaque), and the cloud lifetime effect (smaller droplets make it more difficult to make drops big enough to rain, and so clouds live longer). Estimates of the importance of such effects vary widely, and while they are thought to be significant, the uncertainty associated with them is very large. These effects are nevertheless a necessary part of the suite of human-related forcings that are being assessed in order to understand the climate of the 20th Century.
It should however be stressed that there are as yet no completely convincing explanations that quantitatively match the (admittedly uncertain) observations of this phenomena (Liepert and Lohmann, 2004). However, the Horizon documentary nevertheless confidently asserts that:
Global dimming is a killer. It may have been behind the worst climatic disaster of recent times, responsible for famine and death on a biblical scale. And Global Dimming is poised to strike again.
The reference is to the 1980s famine in Ethiopia, partly caused by the failure of the Sahel monsoon (but clearly exacerbated by the poor governance of the Menghistu regime then in power). The link is based on a single modelling sensitivity study (Rotstayn and Lohmann, 2002) which looked at only the changes in the indirect effect from the pre-industrial (ca. 1850) to the present day. In this study, there was a shift southward of the rainfall belts in a similar way to that observed over the whole century (i.e. not necessarily just the 1980s). While this is indeed very interesting and does suggest that aerosol indirect effects can have important climatic consequences, it is merely the first step to attributing any particular climatic effect (failure of Sahel monsoon) to a particular cause (aerosol indirect effects). The obvious open questions relate to the importance of other forcings, in particular, greenhouse gases (which were not changed in this experiment), and the robustness of any transient response (i.e. does a simulated drought occur in the Sahel in the 1980s more often than at any other time). Absent this further study (which we expect is ongoing as part of the assessments related to the IPCC 4th Assessment report), it is horribly premature to declare ‘global dimming’ the cause of this event. Note that while this study looked at the aerosol effects, it makes no claim to actually match the ‘global dimming’ results. Additionally, other model experiments (Giannini et al, 2003) point to warmer Indian Ocean temperatures to explain the 1980s droughts.
Aerosols are however much more clearly responsible for serious respiratory problems in big cities (London prior to 1950s, Beijing today), and their health impacts are well known. This was one of the big pushes behind initiatives like the Clean Air acts in many countries which reduced aerosol emissions from power stations. While in the developed world (US, Europe, the ex-USSR) emissions have been falling, the global burden is increasing because of development in India and China. Since, on average, aerosols have a cooling effect (although some absorbing aerosols like black carbon (soot) are actually adding to global warming), reducing current aerosol levels (particularly sulphates) is equivalent to an extra warming effect.
An important point to note is that while cooling from aerosols and warming from greenhouse gases may have a slight cancelling effect in the global mean, this is not true regionally. Ideas that we should increase aerosol emissions to counteract global warming have been described as a “Faustian bargain” because that would imply an ever increasing amount of emissions in order to match the accumulated GHG in the atmosphere, with ever increasing monetary and health costs.
Does this all have either an implication for the global climate sensitivity (how much warming would result from a doubling of CO2) or the scenarios used by IPCC to project climate changes out to 2100? This is where I have to disagree most strongly with the commentary in the program. First, if we were trying to estimate climate sensitivity purely from the response over the 20th century, we would need to know a number of things quite exactly: chiefly the magnitude of all the relevant forcings. However, the uncertainties in the different aerosol effects in particular, preclude an accurate determination from the instrumental period alone. While it is true that, holding everything else equal, an increase in how much cooling was associated with aerosols would lead to an increase in the estimate of climate sensitivity, the error bars are too large for this to be much of a constraint. The estimate of 3+/-1 deg C (for doubled CO2) based on paleo-data and model studies is therefore still valid, even after this program.
Secondly, would a re-evaluation of the aerosol effect imply that projections to 2100 must be worse than previously suggested? If the climate sensitivity lies within the bounds considered in IPCC TAR (which I would argue is still the case), the answer is no. The most extreme scenario postulated in TAR (A1F1) already has a big reduction in sulphate aerosol forcing, and so the temperature changes by 2100 are almost purely a function of the GHG forcing. They are therefore unaffected by a re-evaluation of the aerosol indirect effect.
The suggested ‘doubling’ of the rate of warming in the future compared to even the most extreme scenario developed by IPCC is thus highly exaggerated. Supposed consequences such as the drying up of the Amazon Basin, melting of Greenland, and a North African climate regime coming to the UK, are simply extrapolations built upon these exaggerations. Whether these conclusions are actually a fair summary of what the scientists quoted in the program wanted to say is unknown. However, while these extreme notions might make good television, they do a dis-service to the science.