Climate change

The cloud conundrum

One of the great uncertainties in projecting global warming is accounting for the effects of small particles in Earth's atmosphere. Progress is nonetheless being made with this fiendishly complex problem.


Generally accepted climate projections for the year 2100, compared with today, predict a global average temperature increase ranging from around 5.8 °C to a more benign, but still worrisome, 1.4 °C. Which of these futures awaits us depends, in part, on aerosols — tiny particles in the atmosphere — and their effect on clouds. Ackerman et al. (page 1014 of this issue)1 conclude that the role of these particles in increasing the water content of clouds, and so cloud reflectivity, is smaller than previously thought.

As a consequence of air pollution, the mass of small particles in the atmosphere is about 40% larger than it would be in a pristine atmosphere2. These particles mainly reflect solar radiation, a direct effect on climate that has received considerable attention over the past 15 years. But when clouds form, these tiny particles also act as seeds for water condensation. When a cloud forms in a region containing a large number of particles, it will have higher concentrations of smaller droplets if the liquid water in the cloud is constant. The smaller droplets have a higher surface area; so the cloud is brighter and able to reflect more solar radiation, thereby cooling the climate. This is the so-called first indirect effect of aerosols on clouds.

But climate scientists have also explored the consequences of smaller droplets on precipitation. Smaller droplets are less likely to collide with each other and form precipitation. This change in ‘precipitation efficiency’ has been termed the second indirect effect, and has been calculated to increase both the total cloud cover and the total amount of liquid that is held within clouds. These two effects increase the total reflected solar radiation even more and lead to a conundrum. Because some climate models are highly sensitive to this change, they predict a total cooling effect that is larger than the warming by greenhouse gases: but the net effect of greenhouse-gas and aerosol changes over the past 100 years is observed to be a warming.

The conundrum may be solved in one of three ways. Some particles, especially the ‘black carbon’ produced by biomass and fossil-fuel burning, can absorb radiation rather than reflect it, and so counteract cooling. Perhaps this effect has been underestimated: if these particles are much more abundant than thought3, or if they cause more absorption of solar radiation after deposition on snow4 (which is normally highly reflective), they may be counteracting the effects of particles on precipitation efficiency to a greater extent than expected.

Or perhaps the effects of particles on ‘ice clouds’ — clouds high in the atmosphere — have been underestimated. Because high clouds tend to absorb more energy in the form of thermal radiation than they reflect in the form of solar radiation, an increase in ice-cloud amount and particle number would warm the climate. Almost no work has addressed the effect of air pollution on ice clouds; but if increased pollution also causes increases in ice particles within high clouds, it might balance the effect of increased air pollution on low, warm clouds.

A third possibility is that those climate models in which clouds are very sensitive to changes in particle concentrations have got it wrong. Ackerman et al.1 show that this might indeed be the case, which is perhaps not unexpected. Because climate models are typically run at a resolution of only 250 km, they cannot resolve individual clouds, and must parametrize the effects of particles on clouds — that is, they invoke comparatively crude equations of cloud physics, rather than simulate real clouds.

Ackerman et al.1 used a large eddy-simulating, cloud-resolving model to examine the effects of increased particle concentration on low-lying stratocumulus clouds. They picked a number of different cases that have been studied in the past and show that an increase in the number of particles does not lead to an increase of total liquid — unless, that is, the overlying air in the region is humid. The reason is that when precipitation decreases as a result of increased particle concentrations, the mixing of air from above the cloud with that beneath it increases. If the air above the cloudy layer is dry, the increased mixing leads to less liquid water in the cloud. But if the air above the cloud is humid, then the mixing resupplies the cloud with moisture. Ackerman and colleagues' new results restrict the number of regions that may be affected by the second indirect effect to a smaller percentage of the globe, and so mean that the total impact of that effect may be much smaller than previously thought.

If the first indirect effect — cloud brightness — is as large as some models predict, aerosols are still acting strongly to increase reflection of solar radiation, and so are masking the current effect of greenhouse gases to a large extent. If so, then the net radiative forcing that is the cause of the 0.6 °C average temperature increase of the past 100 years must be small, and climate models must be much more sensitive to this small difference if they are to agree with past observations. The future, then, might be more at the upper range of climate projections. But if aerosols do not increase low-level cloud brightness as much as some models have it, then greenhouse gases may be only slightly masked by aerosol-induced cloud changes, and projections of future climate might follow the more benign path.

How fast the world needs to address the issue of greenhouse-gas warming still depends on which of these two paths is correct. We still have the conundrum. But it is now less daunting than it was, given that the results of Ackerman et al. mean we can rule out some of the large effects of aerosols on cloud water content.


  1. 1

    Ackerman, A. S., Kirkpatrick, M. P., Stevens, D. E. & Toon, O. B. Nature 432, 1014–1017 (2004).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Penner, J. E. et al. in Climate Change 2001: The Scientific Basis (eds Houghton, J. T. et al.) 289–348 (Cambridge Univ. Press, 2001).

    Google Scholar 

  3. 3

    Sato, M. et al. Proc. Natl Acad. Sci. USA 100, 6319–6324 (2003).

    ADS  CAS  Article  Google Scholar 

  4. 4

    Hansen, J. & Nazarenko, L. Proc. Natl Acad. Sci. USA 101, 423–428 (2003).

    ADS  Article  Google Scholar 

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Penner, J. The cloud conundrum. Nature 432, 962–963 (2004).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing