Fall 1994, Issue No. 20

Atmospheric Heating and Cooling from Fossil-Fuel Combustion

Stephen E. Schwartz
Brookhaven National Laboratory

The burning of fossil fuels produces not only carbon dioxide but also sulfur dioxide. Atmospheric oxidation converts this SO2 into sulfate aerosols that scatter solar radiation, increasing clear-sky planetary reflectivity. Increased concentrations of aerosol particles also lead to increased reflectivity of clouds. Both of these effects are thought to cool the atmosphere and to offset to some extent the warming from increased CO2.

A major difference between CO2 and sulfate aerosols is the residence time of the materials in the atmosphere: decades to centuries for CO2, days to weeks for sulfates. Another key parameter is the sulfur content in the fuel. The greater the sulfur content, the greater the climatic effect.

The global-warming potential (GWP) concept, which is used to compare the climatic influence of different greenhouse gases, was adapted to compare the globally averaged radiative forcing of sulfate aerosols to that of CO2. Because the two substances have very different atmospheric residence times, their greenhouse-warming integrals (GWIs) were evaluated independently. These integrals reflect the residence-time profiles of a substance by considering both

• the instantaneous forcing resulting from the introduction of a given amount of the material into the atmosphere and
• the time for which one is considering the forcing.

Because CO2 stays in the atmosphere for a substantial time, its GWI must be considered as a function of time. Figure 1 [in the newsletter] shows this time dependence for three exponential decay profiles (reflecting CO2 residence times of 70, 100, and 130 years) and for a multi-exponential profile that reflects the coupling of several reservoirs for CO2. The infinite-time value for the exponential profiles is 27 plus or minus 8 microjoules per square meter per kilogram of carbon emitted into the atmosphere as CO2. For the multi-exponential decay profile, the value is greater and may increase indefinitely, depending on the ultimate fate of the CO2.

In contrast, because of the short atmospheric residence time of sulfate aerosols, the GWI of those aerosols is the same as the total or "infinite-time" GWI. That value is -360 microjoules per square meter per kilogram of sulfur emitted into the atmosphere as SO2, uncertain to a factor of 2. The negative sign indicates that sulfate aerosol exerts a cooling influence.

The ratio of the infinite-time GWI for sulfur to that for CO2 is about 0.075. That value is roughly four times the mass ratio of sulfur to carbon in emissions from fossil-fuel combustion (0.019 for the period 1860 to 1987). Consequently, when CO2 forcing is considered for its entire residence time, the positive (warming) influence caused by CO2 substantially exceeds the negative (cooling) influence of the sulfate aerosol, and the net influence of emissions from fossil-fuel combustion is one of warming.

This conclusion does not hold, however, when one considers the CO2 forcing for shorter time horizons. In particular, the analysis indicates that, during the period of exponential growth in fossil-fuel combustion that has occurred since the beginning of the industrial era, the positive CO2 forcing has been essentially equal to the negative sulfate forcing produced by direct light scattering. Including the effects of cloud forcing further increases the sulfate forcing by a factor of about 2. If these estimates are accurate, the net radiative forcing resulting from fossil-fuel combustion over the industrial era has been one of cooling, not warming. However, this conclusion must remain tentative in view of the uncertainty in estimates of the radiative forcing of sulfate aerosol, and the issue can be resolved only by decreasing this uncertainty.

Several points should be noted:

• The calculations were carried out for a global-mean forcing, and because of sulfate aerosols' short residence time, substantial short-term regional forcings of differing sign may be expected.
• Because of differing feedbacks in different regions (e.g., a greater water vapor feedback over tropical oceans where the sulfate forcing is minimal), the mean global response to these disparate forcings might not be the same as that to a uniform forcing at the mean value.
• In contrast to the sulfate forcing, the full effect of the CO2 emissions has yet to be exerted, leading to a much greater positive global-mean forcing than has been experienced thus far.

This article is based on Issue No. 28 of the DOE Research Summary Series published by CDIAC.