Roy Spencer presentation to the Heartland Institute conference

The Heartland Institute, a US free-market think tank, held a conference on climate change in New York in early March. It was a forum for some of the climate change sceptics to discuss their research. The conference got very little coverage in the media and was ignored by the science pages of the newspapers. This seems a mistake. A large section of the population of the US and the UK, and smaller numbers elsewhere, believe that the apparent scientific consensus on global warming is a result of selective coverage by TV and press. The failure to cover presentations by some of the leading sceptics is support for the accusation that global media, and mainstream climate scientists, are refusing to engage with the dissenting views of reputable scientists who do not share the standard view.


Roy Spencer is one such scientist. He has good scientific credentials and his sceptical book on climate change is selling extremely well in the US. The main theme of his presentation was that climate sensitivity to increases in CO2 is much less than conventionally thought. He doesn’t deny the human sources of climate change; he suggests that the standard models exaggerate the impact of greater amounts of CO2 in the atmosphere because they fail to comprehend the true impact of changes in cloud cover. In summary, he says that changes in cloud cover may damp temperature changes resulting from CO2 increases, rather than exaggerate them.

The scientific consensus is that doubling the amount of CO2 and other GHGs in the atmosphere will, if all other things remain equal, increase temperatures by about 1.2 degrees Celsius. In the standard models, this increase is multiplied by two or three because of the effects of changes in cloud cover. These reinforcing effects are usually called ‘positive feedbacks’. Other positive feedbacks include higher temperatures decreasing the amount of ice cover, causing less light radiation to be re-emitted to space.

Why is it generally thought by climate modellers that changes in cloud cover will amplify the effect of the CO2 increase? There are two forces at work:

  • High-level cirrus clouds act as a blanket around the earth, trapping heat. A generally hotter atmosphere is generally assumed to increase the amount of cirrus cover.
  • Low-level clouds reflect the sun’s light back into space tending to decrease temperatures. Global warming is usually thought to result in decreased low cloud cover, amplifying the effect of warming.

Sir John Houghton’s standard undergraduate textbook on global warming says that ‘climate is very sensitive to possible changes in cloud cover or structure’. In other words, if the standard models are even slightly wrong about the nature of the relationship between clouds and warming, changes in cloud cover may amplify or repress the temperature variations induced by greenhouse gas increases.

Scientists often talk about the global warming impact of greenhouse gases in terms of watts (which can be thought of as a unit of heat) per square metre. This measure is usually called ‘radiative forcing’. The IPCC analysis suggests that the man-made greenhouse gases will increase the radiative forcing by about 4-5 watts per square metre by 2050. Houghton’s book says that forcing by clouds is typically a many times multiple of this. It varies by latitude and by time of year. But the crucial point is that clouds matter a great deal.

As I’ve said, the standard view is that clouds amplify the impact of global warming induced by CO2. Houghton says that there is ‘encouraging agreement’ between this hypothesis and actual observations of cloud behaviour. Roy Spencer’s presentation asked us to consider two pieces of work from his team that tend to contradict this view:

  • A 2007 paper that suggested that tropical rainstorms result in only short-term increases in high-level cirrus clouds that dissipate quickly. (Cirrus acts as a blanket.) Spencer used temperature and other readings collected by satellite.
  • A paper waiting for publication that says that the theory that higher temperatures reduce low cloud cover is inadequate. (Low clouds tend to reflect sunlight back into space.) He says that the causality may be different. Perhaps lower levels of cloud cover result in higher surface temperatures, a phenomenon that we might all instinctively recognise? He claims that previous measurements have simply assumed a causality that sees higher temperatures reducing the coverage of low clouds. He says we haven’t done the measurements properly to ascertain which comes first, higher temperatures or lower cloud cover.

What if Spencer is right? His work suggests that a doubling of CO2 levels from pre-industrial levels – which will occur some time around 2075 if today’s rate of increase persists – will not result in temperatures three or four degrees above pre-industrial levels, as pessimists fear, but perhaps about one degree.

We have already seen about 0.7 degrees, with greater increases in high latitudes. So his theory implies strong negative feedback from now on. As Roy Spencer said in his presentation, the world is engaged in a huge experiment to determine the true sensitivity of climate to changing greenhouse gas levels. His view on the outcome of that experiment is strikingly different to the world’s academic consensus and it would be good to have a rebuttal from those who disagree with him.

Note Dr Spencer was one of the scientists who reported some years ago that the troposphere was not warming. His research findings were undermined by further analysis of the data that his team had collected. Almost all scientists now believe that the degree of tropospheric temperature change is broadly compatible with the standard model of global warming.