That’s it: It was 4% cloudier in 1985, then roughly the same after 2000 — that’s the Pause and the Cause
A new paper in Russian, by OM Pokrovsky, shows that global cloud cover decreased markedly from 1986 to 2000. This is a very large decline in terms of the planetary atmosphere. Pokrovsky uses ISCCP satellite data (the “International Satellite Cloud Climatology Project” — a US program). It’s the best cloud data there is. The effects of clouds are so strong that most of the differences between IPCC-favoured-models comes from the assumptions the models make about clouds. Cloud feedbacks are the “largest source of uncertainty”. [IPCC, 2007]
Clouds cover two-thirds of the Earths surface, reflecting around 30% of the total energy from the Sun back to space. A small change in cloud cover can easily warm or cool the planet, like a giant pop-up shade-sail.
This, on its own, explains all the warming that occurred from 1986 – 2000. It explains the pause. We don’t know why clouds decreased, but we know it wasn’t due to CO2, which kept rising relentlessly year after year, and even faster after the turn of the century.
Something else is driving cloud formation, or density or longevity, and the global climate modelers don’t know what that is.
“Thus, cloud cover changes over three decades during the period of global warming can explain not only the linear trend of global temperature, but also a certain interannual variability.”
What drives the clouds?
Cloud cover changes could be caused by changes in the solar magnetic field, which may drive cloud seeding via its effect on the cosmic rays that bombard Earth (see Henrik Svensmark). But clouds could also be affected by the solar wind or by solar spectral changes, neither of which are included in GCMs. Clouds could also be driven by changes in aerosols due to volcanoes, bacteria, and plankton. Clouds could also form differently with changes in jetstreams or ocean currents. Meandering jet streams put huge “fingers” of cold air into warm air zones — surely a recipe for more cloud formation. (see Stephen Wilde’s work).
Global Climate Models have no chance of predicting cloud cover. They assume cloud changes are a feedback, not a forcing. So, right from the start, the models don’t even recognise that some outside force might be independently changing cloud cover. In 2012, Miller et al. reported that models got cloud feedbacks wrong by 70W/m2 — an error that’s nearly 20 times larger than the total effect of CO2. What a farce.
Calculating the warming effect
The effect of clouds is complicated. High clouds cause warming. Low clouds cause cooling. Clouds over the dark oceans change the albedo of Earth more than clouds over a bright desert. Clouds in the tropics will reflect more incoming light than clouds over the poles. But at its most brutally simple, the more clouds there are, the more the world cools.
Figure 9 below, describes the relationship between global temperatures and cloud cover. It appears Pokrovsky used it to calculate the effect of the reduction in clouds. A 0.07C warming effect for each 1% decrease in cloud cover, means a fall of 4% in cloud cover would lead to 0.3C of warming. This is just from 1986 – 2000AD and is roughly the same amount of warming as was seen in Hadley. In this situation, no matter how much the trend of Hadley temperatures is “adjusted up,” as long as an analyst uses Hadley temperatures to estimate the linear trend, the increase due to clouds will fit. (Expect Hadley 5.0 to start adjusting key turning points next to mess with this clear signal.)
The conclusions in the paper:
Figure 9 presents the corresponding regression analysis results. As global temperatures, we used the data of CRUTEM 3 (University of East Anglia, Great Britain, http://www.uea.ac.uk). The number of points for statistical analysis was 318. The regression equation has the form Y = – 0.0659 X + 19.637. The determination coefficient characterizing the accuracy of the regression is 0.277. The latter means this model accounts for about 28% of the observed dispersion of surface air temperature. High global cloud cover is associated with low global temperatures, demonstrating the cooling effect of clouds. The regression linear approximation model suggests that a 1% increase in global cloud cover corresponds to a global decrease in temperature of about 0.07oC and vice versa.
In the case of global cloudiness of the lower tier, the regression equation changes slightly: Y = – 0.062 X + 16.962. The determination coefficient characterizing the accuracy of the regression increases and in this case is 0.316. From a statistical point of view, this model accounts for about 31% of the observed dispersion of surface air temperature. High low clouds are associated with low global temperatures, demonstrating the cooling effect of low clouds. A simple linear regression model suggests that a 1% increase in global low cloud cover corresponds to a global temperature drop of around 0.06oC and vice versa.
Thus, cloud cover changes over three decades during the period of global warming can explain not only the linear trend of global temperature, but also a certain interannual variability. But the inclusion of a block describing the temporal evolution of cloud cover in climate models remains a problem due to the stochastic nature of cloud variability. However, climate models are deterministic and cannot be directly combined with stochastic cloud blocks. Nevertheless, the factor of cloud cover on climate change cannot be ignored due to the significant contribution of this climate-forming parameter and should be studied more carefully to improve climate forecasts.
IPCC, Assessment Report 4, 2007, Working Group 1, The Physical Science Basis, Chapter 8. Page 636 126.96.36.199 “Clouds” The original link is now broken: http://www.ipcc.ch/pdf/assessment-report/ar4/wg1/ar4-wg1-chapter8.pdf. The link at the front here is my copy of their PDF in 2009. The current IPCC AR4 Chapter 8 version.
Pokrovsky OM (2019) Cloud Changes in the Period of Global Warming: the Results of the International Satellite Project Russian Academy of Sciences, DOI: https://doi.org/10.31857/S0205-9614201913-13