San José State University
Thayer Watkins
Silicon Valley
& Tornado Alley

Clouds, Cloudiness, Surface Temperature,
the Greenhouse Effect and Global Climate Change

The Net Effect of Cloudiness
on Surface Temperatures

The greenhouse effect is not only produced by the greenhouse gases, clouds absorb long wavelength (infrared) radiation from the surface of the Earth and radiate some of it back down. In addition to this absorption and re-radiation of infrared radiation from the Earth's surface they may simply reflect it back to the surface.

Clouds also have a major role in reflecting some of the Sun's short wavelength (visible light) radiation back into space. The proportion of incident radiation reflected by a substance is called its albedo. The albedo of low thick clouds such as stratocumulus is about 90 percent. The albedo of high thin clouds such as cirrus may be as low as 10 percent. The albedo could vary with the wavelength of the radiation, but for clouds it does not as evidenced by the fact that they are white under white light. At sunrise and sunset the incident light is red, orange or yellow and the clouds reflect this light without modification. The albedo of clouds for infrared radiation is likely the same for visible light. There are two sides, top and bottom, to clouds that may be involved in the reflection of radiation.

Thus clouds share a role with the greenhouse gases and also share a role with the ice and snow fields of the high latitudes. (The role of clouds in reflecting the thermal (infrared) radiation back to Earth's surface has generally been neglected.) Altogether Water; in its three forms as vapor, liquid droplets, and particles of ice; is the overwhelmingly dominant substance in Earth's climate.

The effects of cloud cover on temperature is a familiar experience. Without a cloud cover in an area the temperature drops sharply at night whereas with clouds the temperature drop is noticeably more moderate. On the other hand in the daytime in the summer with no clouds the temperature goes much higher than it does when there is a cloud cover.

The effect of clouds on surface temperature is the net effect of three things:

The effect of clouds depends upon their type and the time of day. The more interesting and important type is the low thick clouds. At night the reflection effect is zero so the greenhouse effect and reflection of thermal radiation dominate and the low thick clouds have a warming effect. One can easily see that the reflection of thermal radiation is far more important than the greenhouse effect. The greenhouse effect could at most return 50 percent of the outgoing radiation back to the Earth. Reflection from the underside of clouds probably returns 90 percent of the radiation. The two effects are not in competition. Clouds could return 90 percent from reflection and half of the unreflected 10 percent. Thus it is easy to see why there is such a difference in temperature between a clear night and a cloudy night in the winter. Since the greenhouse effect from the atmospheric gases would be the same on a clear and a cloudy night one could say that the effect from greenhouse gases is negligible compared to the effect of low thick clouds.

The effect of high thin cirrus clouds at night would be very small compared with that of the low thick stratocumulus clouds.

In the daytime the reflection effect can dominate the greenhouse effect and thus clouds have a net cooling effect. The cooling effect of a cloud shadow is familar to everyone.

An even more homey illustration is the effect of a hat or cap on head temperature. A head covering keeps in the body heat. It has essentially a greenhouse effect. But despite that greenhouse effect, in the bright sunlight one is cooler with a hat or cap than without one, as one can see by the amount of sweat produced. Of course, in the shade the reverse is true.

The effect of clouds in the daytime also depends upon cloud type and their height. Thin clouds reflect less sunlight so their net effect may be a slight net warming. The thick, puffy, beautifully white cumulus are highly reflective and so they have a net cooling effect in the daytime but a net warming at night. The dark rain-laden clouds are not reflective but they nevertheless intercept the nearly all of the Sun's radiation and prevent it from reaching the surface. The dark clouds themselves would be warmed by the absorbed radiation and some of absorbed energy would be re-radiated toward the surface. The dark clouds could be net warmers or net coolers depending upon conditions but the general perception is that dark clouds are net coolers in the daytime.

The effects of cloudiness on surface temperature as a function of cloud type are summarized below:

Net Effect on Surface Temperature
of Various Cloud Types by Time of Day
 Thick whiteThick darkThin

The effect of the thick clouds would shift from net cooling to net warming as the Sun's angle changes during the day.

Cloudiness and the Climate Models

The climatic models used by the Intergovernmental Panel on Climate Change (IPCC) are not very good at replicating the current cloudiness as seen from the following diagram. The black line is the observed values and the colored lines are for the various climate models used by IPCC.

Source: IPCC, Third Asssessment Report: Climate Change, 2001

This is the observed latitudinal profile of the proportion of cloudiness during the Northern Hemispheric winter (DJF).

J.R. Houghton expressed the situation as follows:

Clouds are, in fact, probably the dominant influence in the radiative budget of the lower atmosphere but adequately taking them into account raises many problems […]
The Physics of Atmospheres, p. 41.

It is perplexing that the models do so poorly at replicating the current cloudiness characteristics yet they are supposed to be more accurate at replicating the latitude profile of temperature as shown in the following graph.

Source: IPCC, Third Asssessment Report: Climate Change, 2001

The answer is that the climate models deal in terms of energy flows. The role of temperature is to determine the thermal radiation from the Earth. This radiation is proportional to the fourth power of the absolute temperature. This fourth power relationship means that temperature is the fourth root of the thermal radiation flow and thus if the radiation flow is in error by 20 percent the temperature is in error by only one fourth this much or five percent. Thus a model could be in error by a whopping 40 percent and yet the temperature seems to have a modest error of only ten percent.

The role of cloud cover in the determining albedo is illustrated as follows. Albedo is the proportion of incoming radiation that is reflected from surface back to where it came from. The average albedo for the Earth is about 37 percent. The albedo of an area depends upon the nature of its surface and the angle of inclination of the Sun. Snow has a high albedo and black earth has a low albedo. White clouds have an albedo comparable to a snow field as airplane passengers often observe.

If a surface area has an albedo of 10 percent and clouds an albedo of 90 percent then with 70 percent cloudiness the albedo for the area is 0.7(0.9)+0.3(0.1)=0.66=66 percent. This means that 34 percent of the Sun's radiation for the area is not reflected back into space. If a climate model computes the cloudiness to be 40 percent then the albedo is 0.4(0.9)+(0.6)(0.1)=0.42=42 percent. This means the model would be using a solar input of energy to the area that was 42/66=0.636 of the actual, or roughly 64 percent. It is at first hard to see how the model could have an accurate computation of the temperature if the energy input was so far off. The answer as previously noted is that the temperature depends upon the fourth root of the energy input. The fourth root of 0.636 is about 0.893 or roughly 0.9, so the error in temperature would be only about 10 percent. Thus an error of only about 10 percent in temperature corresponds to an error of 34.4 percent in energy flow.

There is vastly more of the Sun's radiation reflected from clouds than from the polar ice caps. This is because there is vastly more area covered by clouds outside of the polar regions than the ice caps. It is also because angle of incidence in the polar regions is so low compared with the other regions of the world.

The situation is shown below for the Northern Hemispheric winter (DJF).

The area of the Earth's total surface is about 718 million square kilometers. The area covered by clouds during December, January and February is 453 million square kilometer. The total amount of sea ice in the polar regions during that time of the year is about 17 million square kilometers. However much of this sea ice is redundant as far reflectance of the Sun's radiation is concerned because there are clouds above it. The cloud coverage in the Arctic is about 60 percent in the winter so only 40 percent of the 15 million square kilometers of sea ice is reflecting the Sun's radiation. The cloud coverage in the Antarctic in December, January and February is about 10 percent so 1.8 million square kilometers of the 2 million is effective. Thus the sea ice area that is effective in reflecting the Sun's radiation is 7.8 million square kilometers. This 7.8 million square kilometers of effective reflective sea ice is about 1.7 percent of the cloud coverage. Thus a 1.8 percent increase in cloud coverage would more than replace the total loss of all the polar sea ice. It would take only 0.66 of 1 percent increase in global cloud coverage to replace a 3 million square kilometer loss of Arctic sea ice. Actually the polar sea ice is even less important to Earth's energy budget than the above computation indicate. The amount of energy reflected in the polar regions is much less because the angle of the Sun. The figure below illustrates this point.

This graph uses some rough approximations to obtain an order of magnitude figure for the importance of the polar ice and snow field to Earth's energy budget. For this graph the albedo of clouds is taken to be 0.8. This same value is used for ice and snow fields. The ice and snow fields are presumed to cover the Earth from 67.5° of latitude to the poles. The albedo of rest of the Earth is taken to be 0.2. In the graph the reflectance is shown in arbitrary units.

Using the above stated values for the albedos and snow and ice coverage, total disappearance of the snow and ice fields would reduce the Earth's reflectance by only 5 percent. This is without any change in clouds.

Seiji Kato and his associates at the Langley Center of NASA published in 2006 the results of an investigation of the effect of decreasing sea ice in the Arctic on the amount of radiant energy reflected from the Arctic. This study concluded that although there was a decrease in sea ice in recent years there was an increase in cloudiness that more than made up for the loss of albedo from the sea ice. Thus there was not only no ice-albedo positive feedback presumed by climate modelers, there was in fact a negative feedback. Kato's result illustrates the admonition that in climatology every theory has to be checked empirically.

A small change in cloudiness over the rest of the Earth's surface can be far more important than major changes in the area of the ice caps. It is important to keep such things in perspective. Climate modelers have a distinct tendency to focus on a sensational minor topic while neglecting the major topics of climate. Clouds and cloudiness are the major factors in the Earth's climate. Clouds rule the Earth's climate. Everything else, including the atmospheric greenhouse gases, is marginal.

A very interesting and important discovery along these lines was made by Richard Lindzen and his group at M.I.T. They found that in the central Pacific region when the sea surface temperature rises there is less cirrus cloud cover and thus more energy radiates out into space. This thermal vent is a negative feedback in the Earth's climate system and one that is not incorporated in the computer climate models used to project global warming. It is estimated that the mid-Pacific thermal vent would reduce by two thirds the projected global temperature increases. Once again it is a matter of the cloud system ruling the Earth's climate system. For more on this see Mid-Pacific Thermal Vent.


Climate models focus on the effect of greenhouse gases, primarily carbon dioxide and water vapor, to the neglect of cloud cover. As shown above this effect is much smaller than that of clouds. Does this mean that the projected temperature change over the next century is larger than the climate models suggest? In principle that would be the case, but the climate models have been tweaked to give plausible projections. For example, the climate models use a rate of increase of the concentration of carbon dioxide which is two and a half times the current rate. There is no justification for this other than to produce scarier projections. The climate models are probably just worthless and should be scrapped. The ones that ventured to provide validation by carrying out backcasts failed miserably. Some people think that because the climate models contain only equations based upon fluid dynamics and thermodynamics that that makes them valid. The climate models are in error from what they have left out rather from what they contain.

HOME PAGE OF applet-magic
HOME PAGE OF Thayer Watkins