As we have stated at several junctures, water vapor is the most effective greenhouse gas. However, human activity does not have a direct impact on the concentration of water vapor; relative to other greenhouse gases, there is a lot of it and its highly variable concentration is controlled largely by weather and climate patterns. All of the other greenhouse gases affected by human activity are found in minute quantities in the atmosphere ranging in concentrations from a few parts per million (ppm) to a few parts per trillion (ppt), which is a millionth of a part per million. The table at the bottom of the page summarizes some of the characteristics of the other greenhouse gases. The reason that these gases are important, even though they are present in very small amounts, is that the molecules which compose them are very efficient at absorbing infrared radiation.
CO2 CH4 CFC-12 N2O Pre-industrial 280 ppm .8ppm 0 .288ppm concentration Current 353 ppm 1.7ppm 484ppt .310ppm concentration Accumulation .5% .9% 4% .25% rate residence 3 years 10 years 130 years 150 years timeWe focus here on three additional greenhouse gases: methane (CH4), a representative halocarbon (CFC-12), and nitrous oxide (N20). Notice that at the present time, all of these gases are found in concentrations smaller than those of carbon dioxide. For example, roughly 2 methane molecules are found in the atmosphere for every 350 carbon dioxide molecules. Concentrations of methane since the industrial revolution began have roughly doubled, while the halocarbons such as CFC-12 are not found naturally in the atmosphere as evidenced by their absence prior to the industrial era. Note also in the table that all of the concentrations of these gases are increasing substantially each year. Finally, methane molecules in the atmosphere are exchanged quite rapidly with other parts of the biosphere while the other two greenhouse gases have much longer residence times.
We will now review briefly some of the key aspects of each of these gases.
Figure 7.1. Sources of methane emissions into the atmosphere in units of 10[12] kg per year of carbon.
Methane is produced as a by-product of a number of anaerobic (oxygen deficient) chemical reactions that take place in the biosphere. As summarized in Fig. 7.1, the main sources of methane result from emissions from natural wetlands, rice paddies, and animal enteric fermentation (more familiar as belching and passing gas). Additional sources include methane emissions released as a result of drilling for natural gas and oil, burning of vegetation, the prolific eating habits of termites, and the anaerobic decay of organic material in landfills. Other minor sources (from the ocean and freshwaters, coal mining, etc.) have been lumped above into a single category. Combined, all of these sources represent a flux of carbon into the atmosphere of .4 units per year.
The major sink for methane are complex, oxidizing chemical reactions in the atmosphere. These reactions remove .345 units of carbon in the form of methane each year with another .022 units removed by processes at the air-soil interface. It is estimated that there is a net increase per year of about .033 units of carbon in the atmosphere in the form of methane. This amount represents a tiny fraction of the carbon gained each year in the atmosphere in the form of carbon dioxide. As we will discuss in the next chapter, a methane molecule is much more effective at absorbing infrared radiation than a carbon dioxide molecule is. Thus, methane is an important greenhouse gas, even though its concentration is small in the atmosphere.
Figure 7.2. The increase in methane concentrations (in ppm) based on air in ice cores.
As we can see from Fig. 7.1, many of the sources of methane are independent of human actions. However, increased rice production and the accelerated growth world-wide of herds of ruminant animals (cattle and sheep) undoubtedly affect the amount of methane in the atmosphere. Figure 7.2 shows estimates of the concentration of methane over the past 400 years measured from air in ice cores. The concentration of methane is increasing very rapidly during the 20th Century.
There is one heartening aspect of methane as a greenhouse gas. As a result of the efficient removal of methane through chemical reactions in the atmosphere, its residence time is relatively short, on the order of a decade. Thus, procedures invoked to control methane emissions may limit its concentration in the atmosphere faster than similar procedures applied to carbon dioxide emissions.
Halocarbons contain chlorine and flourine and are effective greenhouse gases, even though their concentrations are quite small in the atmosphere. In the form of chloro-flourocarbons (CFCs), they have a variety of harmful effects, including the destruction of ozone in the stratosphere. Many governments signed the "Montreal Protocol on Substances that Deplete the Ozone Layer" in 1987 as a means to reduce the emissions of a CFCs and the principle producers of this chemical, including this country, have agreed to a complete halt to production by the year 1996.
CFCs are a nemesis for a number of reasons. They were developed for a variety of industrial purposes: refrigerants, propellants, solvents, etc. Their chief advantage for these industrial applications was that they do not react readily with other chemicals. You probably remember that CFCs were banned in aerosol cans (hair spray, whipping cream, and so forth) during the late 1970's over concerns about their effect on the environment. The inert (non-reactive) characteristic of CFCs is their chief liability once they manage to enter the atmosphere. They have a long lifetime in the atmosphere and are principally destroyed by chemical reactions in the stratosphere. Since this sink is very inefficient, the concentration of CFCs has been increasing at a rate in excess of any of the other greenhouse gases. Recent monitoring of CFCs in the atmosphere does provide some evidence that the rate of increase in their concentration is slowing down. As we shall see in Chapter 9, CFCs are also the most effective greenhouse gases when compared on a molecule by molecule basis.
Nitrous oxide is produced naturally from a variety of chemical reactions in the biota and oceans. Human activity in terms of combustion and the increased reliance on fertilizers have led to increasing concentrations of N2O during the industrial era. Like CFCs, nitrous oxide is destroyed in the stratosphere as part of reactions that play a role in other controversies, including the depletion of ozone in the stratosphere.
The quantity that is used to measure the effects of a greenhouse gas on the radiation balance is called radiative forcing. It is the amount by which a given quantity of greenhouse gas reduces the infrared radiation emitted to space. No direct measurements are available to determine what is the radiative forcing that has arisen specifically from the observed increase in carbon dioxide over the past hundred years (or the past ten years for that matter). We must then rely on estimates of this forcing based on models that simulate as accurately as possible the radiative behavior of the atmosphere. As opposed to numerical weather prediction models that simulate the behavior of the atmosphere in all three dimensions, these radiative transfer models simulate the atmosphere in one dimension only, the vertical. Thus, they are designed to evaluate the effect of changes in the radiation for the global atmosphere as a whole. In this respect, our models developed in Section 2.2 are crude facsimiles of radiative transfer models.
The result of simulations with radiatiave transfer models is this: the radiative forcing due to a doubling of atmospheric carbon dioxide is 4 Wm[-2]. Note that this is the radiative forcing due to the increased carbon dioxide alone. Other feedback mechanisms to be discussed in the next chapter can further increase the radiative forcing beyond 4 Wm[-2]. The mechanism by which the additional carbon dioxide achieves a positive radiative forcing (that is, a reduction in emitted infrared radiation) was examined in Section 2.2 with the model that we developed there. We recall that the response of that model to an increase in a greenhouse gas was that the ERL went to a higher level, where the temperature was lower. The resulting decrease in emitted infrared constitutes the radiative forcing. The atmosphere and underlying surface then warm in response to the radiation imbalance that this radiative forcing represents.
Recall that the equilibrium temperature for the earth-atmosphere system is 254 K. What would be the increase in this temperature that would follow from a radiative forcing of 4 Wm[-2]? We can easily determine from the Stefan-Boltzmann Law that it would be 1[o]C. We could stop now and conclude that a doubling of atmospheric carbon dioxide will result in a global warming of 1[o]C except for the fact that this estimate ignores contributions from the rest of the climate system. The feedbacks arising from the change of water vapor content, the exchange of heat with the ocean, and the potential changes in cloud amount and height as a result of the changing concentration of carbon dioxide must be considered. This requires simulations with 3-dimensional climate models. The results of such simulations will be covered in the next chapter.
We turn now to the objective of quantifying the effect of other greenhouse gases. As in the case of the carbon dioxide increase we have just been discussing, these results have been obtained from radiative transfer models with no water vapor or clouds included. We will use as our baseline the radiative forcing that we think prevailed in 1765. Figure 7.3 shows the additional radiative forcing since then broken into several eras. Focus first on the sum of the contributions from all of the gases (the total height of each of the columns.) From 1765-1900, human activity caused an increase in the radiative forcing of about .55 W m[-2]. The radiative forcing increased by another 1.2 W m[-2] during the period from 1900-1960. During each of the next three decades, the radiative forcing has been rising at an accelerating pace. The radiative forcing contributed from the 1980-1990 period is nearly as large as that which came from the thirteen and a half decades from 1765-1900.
Now look at the relative magnitudes of the contributions to the radiative forcing that came from each of the greenhouse gases. Prior to 1900, increased CO2 emissions were responsible for most of the increased radiative forcing. During the 1900-1960 era, methane and CFC emissions play a larger role. CFCs in the last decade (1980-1990) are estimated to have contributed about 35% as much radiative forcing as that provided by CO2.
Figure 7.3. Contribution to radiative forcing (in W m[-2]) due to increases in greenhouse gas concentrations since the onset of the industrial era. The sum of all of these contributions represents a change in radiative forcing since the onset of the industrial era equal to 2.7W m[-2].
We know that CO2 molecules are much more abundant in the atmosphere than methane or CFC molecules. How could the latter two greenhouse gases have contributed significantly to the radiative forcing during the past few decades? It turns out that a CFC molecule is especially effective at absorbing infrared radiation, in part because it absorbs infrared selectively at wavelengths in the atmospheric window (8-11 microns). Remember from Chapter 2 that the peak emission of infrared occurs in this span of electromagnetic spectrum. For comparison, one molecule of CFC is 12,000 times more effective at absorbing infrared than one molecule of CO2. Methane also absorbs infrared selectively (see Fig. 2.20). On a molecule to molecule basis, it is 21 times more effective at absorbing infrared radiation than carbon dioxide.
Let's examine in Fig. 7.4 the possible increases in radiative forcing that may result in the future on the basis of the four policy scenarios introduced in Section 6.4. For reference, Scenario A assumed that world development takes place without any limitation due to concern about the global warming problem; this is the Business-As-Usual Scenario. The other three scenarios (B, C, and D) represent successively more stringent controls on greenhouse emissions. Remember that none of these scenarios expect that the concentration of carbon dioxide in the atmosphere will begin to diminish within the next hundred years.
Focus first on the sum of the contributions from all of the gases, shown in Fig 7.4 as the uppermost curve in each panel. In Scenario A, the radiative forcing is expected to increase from below 3 W m[-2] now to 10 W [-2] by the year 2100, an increase of over 300% from the present value. Only in Scenarios C and D is the radiative forcing kept within a factor of two of the present value.
Figure 7.4. Projected changes in radiative forcing (in W m[-2]) due to estimated increases in greenhouse gas concentrations using the four policy scenarios.
Now look at which greenhouse gases contribute to this projected increase in radiative forcing. In the Business-as-Usual Scenario (Scenario A), much of the projected increase arises from uncontrolled emissions of CO2. with methane emission contributing the second largest amount[*]. The other scenarios reduce the contributions from CO2 and methane with increasing effectiveness, according to these calculations.
The preceding section was devoted to a description in terms of radiative forcing of the effect of increasing concentrations of greenhouse gases on the radiation balance of the earth-atmosphere system. The effect of clouds on the radiation balance can likewise be quantified in terms of radiative forcing. Consider some local region of clear sky anywhere on earth. There the infrared radiation emitted to space is coming from the clear-sky ERL. If we now add to the region some mix of clouds whose tops are above the clear-sky ERL, then the radiation emitted from the region will be less than it was. Such a decrease constitutes a positive radiative forcing. Recall from Section 2.3 that clouds have an additional effect: The cloud albedo effect. Thus, when we begin with a clear sky and then add clouds, there will be an increase in the amount of visible solar radiation reflected back to space from the region. This corresponds to a negative radiative forcing which acts to cancel and possibly even dominate the positive radiative forcing representing the decrease in emitted infrared. Such is the competing effect of these two types of radiative forcing by clouds that we must carefully define our terms.
We make these definitions in the context of a presentation of the present observed global field of cloudiness on the radiation balance of the earth-atmosphere system. The data comes from three polar orbiting satellites that carry the instruments for the so-called "Earth Radiation Budget Experiment" (ERBE) that has been in progress since 1984. Each of the ERBE satellites carries instruments for measuring emitted infrared radiation and reflected visible radiation at points along a "scanning swath" under its orbit. Fig. 7.5 illustrates the geometry of the situation. The narrow column with a "single-pixel" area at its base represents the field of view of a camera on board the satellite. The camera instantly records infrared radiation (X Wm[-2]) emitted from some mix of clouds and clear sky in the region represented in the figure by a pattern of dots. This reading it compares with a reading obtained from the nearest region of clear sky along the same or, if need be, adjacent scanning swath. Let this reading be denoted by Xcs, where the subscript stands for "clear sky." We define the infrared cloud radiative forcing to be the quantity Xcs - X. Since X is less than or equal to Xcs, the quantity is positive or zero.
At the same time that the infrared observation takes place, another camera
records the visible radiation, (Y Wm[-2]),
that consists of light reflected back from clouds together with light reflected
from the earth and blue light scattered back toward space by the molecules of
the atmosphere. Let Ycs denote the reading obtained by the same camera
from the nearest region of the clear sky. We define the visible cloud
radiative forcing to be the quantity Ycs - Y. Since Ycs
is less than or equal to Y, the quantity is negative or zero.
Figure 7.5. The cloud radiative forcing is determined by comparing the visible reflected and infrared radiation emitted to space from cloud free regions (shown schematically here on the left edge of the scan) to the radiation from cloudy regions (on the right edge of the scan).
Let us look at the product of such a global monitoring of the effect of clouds on the radiation balance of the earth-atmosphere system. We turn to the color plates that are found after page XX of the text. In the discussion that follows we shall abbreviate the term cloud radiative forcing to the initials CRF. Plate 1a is a global map of infrared CRF for the month of July 1985. The color code follows the scale appearing below the map. According to the definition given above, the infrared CRF is a positive number or zero.
We first focus attention on the band of strong infrared CRF (red and yellow) found in tropical and in northern subtropical latitudes. This is the hallmark of a sky persistently and heavily populated with clouds whose tops are well above the clear-sky ERL. Here "persistent" means present on most days of the month and "heavy" refers to a spatial coverage that is overcast or nearby so. This region stands in marked contrast to that which is immediately north and south of it over the ocean basins, Africa, and Australia. There the infrared CRF (purple) is a small positive number or zero. This is indicative of regions where, if there are clouds at all, their tops are close to or below the clear-sky ERL. Finally, the areas of moderate infrared CRF (green) are evidently populated with clouds whose tops are above the clear-sky ERL, but are not as high or as widespread and persistent as in the tropical and northern subtropical latitudes.
The visible CRF for the same month is shown in Plate 1b. A whole new color code is employed here. From its definition, the visible CRF is a negative number or zero. The first thing to take note of is the band of strong visible CRF (blue) found in tropical and northern subtropical latitudes. Strong visible CRF is the hallmark of persistent and heavy cloud cover and does not depend on whether the cloud tops are high or low. Comparison of the spatial pattern of this band with the spatial pattern of its counterpart in Plate 1a leads to the following conclusion: the field of clouds with high tops that characterizes much of the tropics and subropics has the property that its visible CRF cancels most of its infrared CRF. These clouds therefore have a relatively small effect on the radiation balance of the earth atmosphere system.
Notice now in Plate 1b the deep blue and purple coloration exhibited by the North Atlantic and North Pacific regions of the globe. This region of very strong visible CRF is produced by a solid overcast of clouds that are present every day of the month. When we consult Plate 1a we see that these clouds produce only a moderate infrared CRF, which indicates that the top of this widespread overcast is not much above the clear-sky ERL. Putting the two things together, we have the following conclusion: the rather low overcast cloudiness field that persists over the high altitude ocean basins of the Northern Hemisphere in summer produces a very strong visible CRF that is not cancelled by the infrared CRF. In other words, these clouds have considerable effect on the radiation balance of the earth-atmosphere system. The visible CRF is negative, which is to say that these clouds are acting to cool the earth-atmosphere system.
Now that we have identified and described two rather different regimes of cloud radiative forcing, we can go on to examine the pattern of net CRF. This we obtain by adding the infrared CRF (a poistive number or zero) to the visible CRF (a negative number of zero). Plate 2a shows the net CRF for the month we have been discussing (July 1985). What clearly emerges is the region of very strong negative net CRF over the ocean basins of the Northern Hemisphere. The region extends southward along the western coast of North America and emerges also along the western coast of South America. These are the regions of strong upwelling whose relation to the deep ocean circulation was described in Chapters 4 and 6. The cold upwelled surface water is conducive to the maintenance of an overcast layer of low clouds. Areas of less intense upwelling near the west coast of Africa foster low cloudiness which results in the broad region of moderate negative net CRF seen there. With the exception of some small isolated areas, the net CRF elsewhere over the globe is small.
So much for the pattern of net CRF that prevails over the globe in the Northern Hemisphere summer season. Let us turn now to the opposite season. Plate 2b shows the net CRF as observed for January 1986. Little needs to be said to interpret this. The condition of low overcast which characterized the high latitude ocean regions in the Northern Hemisphere six months earlier has simply shifted to the Southern Ocean, as the sea which extends around Antarctica is called. Elsewhere over the globe, the net CRF is small.
Figures such as the ones we have just been examining contain a lot of information. Now that we have seen where on the globe the regions of significant net CRF lie and understand what is responsible for this, we can eliminate some of the information. This is done by averaging in space. First we form what is known as the zonal average of the net CRF. A good place to start this averaging process is at the equator. We return to Plate 2b and from there take the values of net CRF at equally spaced points around the equator add them up, and divide by the number of points. We repeat the process at any number of latitudes between the equator and the poles. The result is the solid-line curve (January) in Fig. 7.6. A repeat of the same operation on the data represented in Plate 2a yields the curve labelled July. What then stands out clearly is the zone of strong negative net CRF situated in high latitudes of the summer hemisphere. This zone can still be discerned in the intermediate-season months of April and October.
It is now evident from the information that is displayed in Fig. 7.6 that the net effect of clouds is to cool the earth-atmosphere system. Data such as what we see here in Fig. 7.6 can be further averaged over latitude for each month of the year and these 12 numbers can be averaged over the year. What we have then is the global and annual average for the year in question. Those numbers can then be averaged over the number of years for which the record exists.
We show in the following table numbers generally agreed to be representative of the global and annual average state of the present climate system.
Cloud Albedo effect -44Wm-2 Cloud Greenhouse 31Wm-2 effect Net effect -13Wm-2
The negative sign of the net CRF indicates that present effect of clouds is to make the system cooler than it would be if there were no clouds. What is particularly significant is the size of this number: it is a little more than three times larger than the (oppositely signed) radiative forcing that characterizes a doubling of atmospheric carbon dioxide, which was 4 W m[-2]. What's more, we can see from the table that the net CRF is the difference between two numbers that are fairly large compared to the difference itself. This signifies that the net is the result of a fairly delicate balance between the two different types of radiative forcing that clouds provide.
The existence of the balance just identified has implications for climate models that are used to predict the response of the global earth-atmosphere system to a doubling of atmospheric carbon dioxide. There is every reason to believe that the response of the system to the increased carbon dioxide will, among other things, consist of a change in the characteristics of the global field of cloudiness. The models have the task of calculating quite accurately both the infrared CRF and the visible CRF in order to get the difference between these two comparably sized numbers right. That the models are not yet up to this task will become evident when we compare them against one another in the next chapter. The lack of consensus as to which of a variety of ways currently used to simulate clouds in climate models is best, together with the large net radiative forcing that clouds are observed to provide, jointly render simulation of clouds the largest single uncertainty afflicting global warming predictions at this time.
1. What are the three greenhouse gases other than carbon dioxide that are rapidly accumulating in the atmosphere as a result of human activities?
2. Why do CFCs last so long in the atmosphere while methane molecules do not?
3. What is the definition of the term "radiative forcing"? What (in Wm[-2]) is the radiative forcing due to a doubling of atmospheric carbon dioxide alone?
4. Why are 1-dimensional radiative transfer models useful for studying feedbacks in the 3-dimensional climate system?
5. Why do we say that doubling CO2 would cause a reduction by 4 W m[-2] in the radiation emitted to space and, at the same time, we say that the atmosphere remains in radiative equilibrium?
6. If cloud top height changes in a given location, how does that affect: (a) cloud infrared radiative forcing and (b) cloud visible radiative forcing?
7. If cloud amount changes in a given location, how does that affect: (a) cloud infrared radiative forcing and (b) cloud visible raiative focring?
8. What (in W m[-2]) is presently the net cloud radiative forcing in the global and annual average?