1
Climate Forcing by Aerosols

It is easy to take for granted the constancy of Earth's climate and the ability of humans to cope effectively with its natural variability. After all, the human race has survived ice ages and, more recently the Little Ice Age; the effects of large volcanic eruptions; and myriad events of violent weather, floods, and droughts. However, this oft-held assumption is misleading in its naivete and may even be dangerous. Recent advances in atmospheric science have shown that

  • the chemical composition of the entire atmosphere of the planet has been changed due to human activity;

  • these changes in gases and airborne particles result in changes in the heat balance of the planet; and

  • the meteorological processes that, when taken together, constitute climate are dependent on and driven by local variations in this energy balance.

Thus, there are fundamental reasons to believe that changes in climate may result from these human-induced changes in atmospheric composition. The magnitude of these changes in climate, however, are poorly known.

In order to approach this complex subject systematically, the scientific community has divided the problem into two major parts, referred to as climate forcings and climate responses. Climate forcings are changes in the energy balance of the Earth that are imposed upon it; forcings are calculated or measured in units of heat flux—watts per square meter (W m-2). Responses



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change 1 Climate Forcing by Aerosols It is easy to take for granted the constancy of Earth's climate and the ability of humans to cope effectively with its natural variability. After all, the human race has survived ice ages and, more recently the Little Ice Age; the effects of large volcanic eruptions; and myriad events of violent weather, floods, and droughts. However, this oft-held assumption is misleading in its naivete and may even be dangerous. Recent advances in atmospheric science have shown that the chemical composition of the entire atmosphere of the planet has been changed due to human activity; these changes in gases and airborne particles result in changes in the heat balance of the planet; and the meteorological processes that, when taken together, constitute climate are dependent on and driven by local variations in this energy balance. Thus, there are fundamental reasons to believe that changes in climate may result from these human-induced changes in atmospheric composition. The magnitude of these changes in climate, however, are poorly known. In order to approach this complex subject systematically, the scientific community has divided the problem into two major parts, referred to as climate forcings and climate responses. Climate forcings are changes in the energy balance of the Earth that are imposed upon it; forcings are calculated or measured in units of heat flux—watts per square meter (W m-2). Responses

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change are the meteorological results of these forcings, including a large number of variable factors such as temperature, wind, rainfall, the probabilistic distribution of these, and extremes of weather. The changes in energy balance as a result of changes in composition are calculated to be a few watts per square meter; some forcings are positive and others negative. Changes of this magnitude are a small, but finite, fraction of the average total energy flux into and out of the Earth's surface, which is approximately 200 W m-2. Nonetheless, changes of a percent or two in heat flux are calculated in some climate models to produce significant changes in key meteorological parameters such as temperature and rainfall. In order to be confident in our ability to predict future climate, we must accurately quantify these forcings, use them to predict the response of the climate system, and verify that the response is accurate through comparison to data (e.g., the historical temperature record). Whereas observing systems are now in place and adequate for quantifying greenhouse forcings, those for quantifying the forcing by anthropogenic aerosols do not exist. Further, the inclusion of aerosols and their effects in climate models is highly simplified and may contain errors. An incorrect or uncertain calculation of forcing by anthropogenic aerosols could significantly alter the understanding of climate response because of the requirement that the climate response to any given forcing should not be much different from that allowed by comparison to historical data. Thus, a great deal is at stake in correctly and accurately quantifying all climate forcings. Without this, it will be impossible to develop any reliable predictive capacity for climate responses. In this report, we consider just one family of forcings—those resulting from aerosols. We conclude that there is substantial evidence that these forcings are significant compared to those by greenhouse gases (GHGs), but that the current quantification of them is much more uncertain than for forcings by GHG. We further conclude that the ability to adequately predict climate change will depend on reducing these uncertainties through an integrated research program. ATMOSPHERIC AEROSOLS An aerosol is defined as a suspension of solid or liquid particles in a gas. Atmospheric aerosols are ubiquitous and often observable by eye as dust, smoke, and haze. Particles comprising the atmospheric aerosol range in sizes from nanometers (nm) to tens of micrometers (µm), that is, from large clusters of molecules to visible flecks of dust. Most of the smallest particles (less than about 0.1 µm) are produced by condensation, either from reactive gases in the atmosphere (e.g., sulfur dioxide) or in high-temperature

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change processes (e.g., fire). Particles larger than about 1 µm are usually produced mechanically (windblown soil, sea spray, etc.). As a result of myriad production processes, atmospheric aerosol chemical composition is highly variable, with respect both to size and to spatial and temporal distribution. Because of different sizes and chemical compositions, aerosol particles have a wide range of lifetimes in air: from minutes to hours for the largest dust particles, from days to weeks for submicrometer smoke and haze particles in the troposphere, and up to two years for volcanic aerosol in the stratosphere. There are two layers of the lower atmosphere: the troposphere and the stratosphere (i.e., below and above the tropopause at about 10 kilometer (km) altitude, respectively). Tropospheric aerosols vary significantly in amount and composition by region, with a characteristic horizontal spatial scale of variation ranging from 1 km to a few thousand kilometers. Notable examples of regionally defined tropospheric aerosol types are marine aerosol, industrial haze, desert dust, and smoke from biomass combustion. Because of its much longer residence time, stratospheric aerosol is substantially more homogeneous chemically and spatially than tropospheric aerosol. The presence of aerosols in the stratosphere is most evident following large volcanic eruptions. For example, the June 1991 eruption of Mt. Pinatubo in the Philippine Islands caused brilliant sunsets and sunrises worldwide through most of 1992. A substantial fraction of today's tropospheric aerosol is anthropogenic . The highly visible haze that persists in all of the industrialized regions of the world consists mainly of sulfate and organic compounds from emissions of sulfur dioxide, organic gases, and smoke from biomass combustion. These emissions have changed dramatically over the past century; SO2 emissions now amount globally to approximately 65-80 teragrams (Tg) (as elemental sulfur) per year, mainly from burning of solid fuels and smelting of metal ores. Figure 1.1 charts the growth of Northern Hemisphere sulfur emissions during the past century and, for comparison, provides estimates of Northern Hemisphere natural fluxes. [Southern Hemisphere anthropogenic emissions are perhaps 10 percent of those in the Northern Hemisphere.] Because the time scale for tropospheric air transport across the equator is much longer than that for the removal of sulfur compounds from the air and because most anthropogenic sulfur emissions are located in the Northern Hemisphere, most of the anthropogenic sulfate aerosol is found in the Northern Hemisphere. As a result, anthropogenic sulfur emissions are estimated to contribute about 80 to 90 percent of the total burden of sulfate aerosol in the Northern Hemisphere. In their letter to the chair of the Board on Atmospheric Science and Climate (BASC) of the National Research Council, the director of the Office of Global Programs of the National Oceanic and Atmospheric Administration

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change FIGURE 1.1 Estimated Northern Hemisphere and regional anthropogenic sulfur emissions over the past century. SOURCE: Dignon and Hameed (1989); Dignon and Gene (1995). (NOAA), the director of the Environmental Sciences Division of the U.S. Department of Energy (DOE), the director of the Division of Atmospheric Sciences of the National Science Foundation (NSF), and the chief of the Climate and Hydrologic System Branch of the National Aeronautics and Space Administration (NASA) wrote the following: Recent work has suggested that anthropogenic aerosols, especially sulfates, may exert a substantial radiative forcing of climate, comparable in magnitude, but opposite in sign, to the forcing from anthropogenic greenhouse gases. They may also play an important secondary role in climate as a source of cloud condensation nuclei. If these hypotheses are correct, they would have major implications on our perceptions of the entire climate change issue. In view of the potential significance of climate forcing by both manmade and natural aerosols and the relatively undeveloped ability to describe this forcing in climate models, we are writing to request that BASC advise the government on development of a strategy and potential program plan for a U.S. research effort. Given current agency priorities concerning climate and global change, we feel that the appropriate scientific focus for the U.S. program of aerosol research is on the climatic effects of aerosol particles. … Specifically, in our judgment, such a research program should be directed toward quantifying the radiative properties, sources and sinks of

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change tropospheric and stratospheric aerosols, and describing their forcing accurately in global models so that their climatic consequences may be evaluated. In our view, initial efforts should be directed at quantifying this forcing, with the objective of reaching an early understanding of its magnitude averaged over various geographical scales. It also seems prudent that actions be identified so that these can be initiated right away. Such actions might include the planning of airborne field campaigns in the near-term as well as initiation of the design of satellite systems for characterizing the aerosol forcing, particularly for tropospheric aerosols, in view of the long lead times for developing such systems and for securing flight opportunities. Similarly, BASC should identify measurement requirements, for aerosol properties and distributions, for research that needs to be initiated with minimum delay. This chapter is intended to provide an overview of effects on the Earth's radiative balance caused by changes in stratospheric and tropospheric aerosols (i.e., climate forcing by aerosols). After defining terminology, the nature of forcing by greenhouse gases is compared with that by aerosols. A description is presented of the evidence for climate forcing from secular increases in anthropogenic tropospheric aerosols. Whereas volcanic eruptions are decidedly nonanthropogenic, the aerosol they inject into the stratosphere affords an atmospheric experiment into the radiative effects of aerosols of unparalleled magnitude. Current knowledge about effects from episodic occurrences of stratospheric aerosol increases is then reviewed. Finally, the nature of climate forcing by different classes of anthropogenic aerosols is discussed. AEROSOL RADIATIVE FORCING OF CLIMATE Climate forcings are changes imposed on the planetary heat balance that alter global temperature (Hansen and Lacis, 1990; Hansen et al., 1993a). Although the term climate forcing is sometimes used to refer to changes in energy balance associated with internal fluctuations of the climate system (e.g., regional cloud or snow cover), we restrict the word forcing to describe anthropogenic or other externally imposed changes in energy balance. Such changes are measured in watts per square meter, and they allow direct comparison of forcing from different atmospheric constituents. The current, global mean forcing from anthropogenic increases in GHGs [including CO2, CH4, N2O, and chlorofluorocarbon (CFC) increases since ca. 1800] is estimated to be approximately +2.5 W m-2 (IPCC, 1995a). Substantial geographical variability in this GHG forcing exists mainly as a result of differences in the Earth's surface temperature. By comparison (as discussed below), the stratospheric aerosol from the eruption of Mt. Pinatubo caused a temporary global mean climate forcing (by scattering sunlight

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change back toward space) of about -4.0 W m-2 [i.e., exceeding GHG forcing in magnitude (opposite sign)]. Both models and measurements show a transient (1992-1994) response of about -0.5°C in global mean surface temperature attributed to the Mt. Pinatubo aerosol (McCormick et al., 1995). Although large geographical and temporal (diurnal to seasonal) variability existed from the volcanic aerosol, there was clear evidence of a transitory aerosol-induced effect on the radiative balance of the Earth resulting from the eruption of Mt. Pinatubo. While in the atmosphere, aerosol particles affect the energy balance of the Earth both directly [by reflecting and absorbing shortwave (solar) radiation and by absorbing and emitting some longwave (infrared) radiation] and indirectly [by influencing the properties and processes of clouds and possibly, although these effects are unknown, by modifying atmospheric dynamics and chemistry (e.g., by participating in the heterogeneous chemistry of reactive greenhouse gases such as O3)]. The direct effect can be observed visually as the sunlight reflected upward from haze when viewed from above (e.g., from a mountain or an airplane). The result of the process of scattering of sunlight is an increase in the amount of light reflected by the planet and, hence, a decrease in the amount of solar radiation reaching the ground. Possible meteorological consequences of the direct effect range from a global change in energy balance to changes in the rate of warming and drying of soil or the rate of evaporation of surface water. The indirect aerosol forcing of climate is a result of the observation that anthropogenic emissions cause increases in the number of particles that can nucleate cloud droplets. As a consequence, the number concentration of cloud droplets, which is governed, in part, by the number concentration of aerosol particles in the pre-cloud, is also increased. An increased number concentration of cloud droplets leads, in turn, to enhanced multiple scattering of light within clouds and to an increase in the optical depth and albedo of the cloud. The areal extent of the cloud may also increase as a result of several different mechanisms (e.g., larger droplet numbers yielding smaller droplets, slower coalescence, and thus longer droplet lifetimes, or organic material slowing the rate of evaporation of droplets). A key measure of aerosol influences on cloud droplet number concentrations is the number concentration of cloud condensation nuclei (CCN). These are particles that will activate to form cloud droplets at a given supersaturation of water. Other meteorological influences, such as changes in precipitation, might occur as a result of perturbations in the number concentration of aerosols; however, such effects have yet to be assessed quantitatively. Even though these indirect effects are not as theoretically tractable as the direct ones, they may be as important as (or more important than) direct effects in contributing to radiative forcing. For example, calculations suggest that a

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change global change of -/+15 percent in the droplet population of marine low (stratus) clouds would cause a change in energy balance of +/-1 W m-2 (Charlson et al., 1992a). Table 1.1 compares the elements of climate forcing by aerosols with that for GHGs. Two major aspects of the comparison of anthropogenic GHG and aerosol forcing are noteworthy: 1. The two forcings (GHG and aerosol) have very different spatial and temporal distributions: GHG forcing operates night and day, whether clear or cloudy, and is at a maximum in the hottest, driest places on Earth (most infrared radiation is trapped in the atmosphere by vapor, liquid, and solid forms of water). In contrast, forcing by anthropogenic aerosol occurs mostly by day, is strongest without clouds, and because of the relatively short residence time of aerosols, is concentrated near aerosol sources. 2. Most GHGs have lifetimes in air very long compared with lifetimes of aerosols. For example, the current, enhanced level of CO2 represents an accumulation of many decades of emission, whereas the current anthropogenic aerosol derives from emissions during only the previous few days. EVIDENCE FOR RADIATIVE FORCING BY ANTHROPOGENIC AEROSOLS The first question the panel addressed was: is anthropogenic aerosol forcing of climate a sufficiently important element of the overall climate system, such that its uncertainty is limiting our ability to quantitatively assess the effect of anthropogenic emissions on climate change? The panel finds that the evidence supporting an aerosol forcing effect on climate of a magnitude comparable, but opposite in sign, to that of GHGs over industrialized regions of the Northern Hemisphere is compelling. It is the opinion of the panel that the uncertainty in the magnitude of the effect of aerosols on climate is seriously hindering our ability to assess the effect of anthropogenic emissions on climate. There are a number of independent lines of evidence supporting the hypothesis that anthropogenic aerosols cause substantial climate forcing that is of a magnitude comparable to that of GHGs, but opposite in sign. The following lists some of this evidence, subdivided into direct and indirect forcings. Direct Forcing Observations show that stratospheric aerosol from the 1991 eruption of Mt. Pinatubo produced a peak global mean optical depth at 550-nm wavelength of 0.1 to 0.2, which resulted in a measured peak global forcing

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change TABLE 1.1 Comparison of Climate Forcing by Aerosols with Forcing by Greenhouse Gases (GHGs): Fundamental Differences in Approach to Determination and Nature of Forcing Factor Long-Lived GHGs (CO2, CH4, CFCs) Short-Lived GHGs (O3, HCFCs, VOCs) Aerosols Optical properties Infrared absorption is well quantified for all major and minor GHGs Infrared absorption is reasonably well quantified Refractive indices of pure substances are known, but sizedependent mixing of numerous species and the nature of mixing have optical effects difficult to quantify Important electromagnetic spectrum Almost entirely longwave (λ > lµm) For O3, solar and longwave are important For tropospheric aerosol, mainly solar; for stratospheric aerosol, solar and longwave contributions lead to stratospheric warming Amounts of material Well mixed; nearly uniform within the troposphere Highly variable in space and time. Concentrations may be estimated by chemical models, but with considerable uncertainty Pronounced spatial and temporal variations Determination of forcing Well-posed problem in radiative transfer; originally considered by Arrhenius (1896) Radiative aspects well posed. Global networks provide some data to test model predictions of geographical and altitudinal distributions Direct: Relatively well-posed problem, but dependent on empirical values for several key aerosol properties. Dependent on models for geographical/temporal variations of forcing Indirect: Depends on aerosol number distribution; no fundamental approaches are yet available. Inadequacy of mathematical descriptions of aerosols and clouds seriously restricts abilities to predict indirect forcing

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change Dependence of forcing on loading (at present) Varies as weak function (square root or logarithm) of concentration Generally nonlinear with concentration; for halocarbons, a linear dependence Direct: Almost linear in the concentration of particles Indirect: Undoubtedly nonlinear Nature of forcing Varies geographically, from ~0.6 W m-2 at South Pole to 3 W m-2 in the Sahara region. Forcing is exerted at the surface and in the troposphere. Operates night and day Strongly dependent on geographical, altitudinal, and temporal variation (e.g., O3). Forcing is exerted at the surface and troposphere. Operates night and day Tropospheric forcing varies strongly with location and season and occurs only during daytime; maxima of forcing occur near sources and at the Earth's surface. Stratospheric forcing includes some longwave effect but is dominated by shortwave radiation (daytime only); following major volcanic events, stratospheric mixing yields a forcing that is substantially global in nature       For nonabsorbing tropospheric aerosols, forcing is almost entirely at surface; for stratospheric aerosols, there is a small heating resulting in a transient warming of stratosphere NOTE: CFC = chlorofluorocarbon; HCFC = hydrogenated chlorofluorocarbon; VOC = volatile organic compound.

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change of about -4.0 W m-2 in 1992 (McCormick et al., 1995) and a temporary, calculated and observed cooling of the surface of approximately 0.5°C (Lacis, 1995). Optical depths from anthropogenic aerosol in industrial regions of the Northern Hemisphere are often greater than 0.1 or 0.2 and cover large portions of the Northern Hemisphere. Studies of visibility indicate increasing extinction from the 1940s to the 1970s as a result of anthropogenic aerosols in the eastern United States (Husar et al., 1981). Typical visually estimated extinction coefficients of (1-3) × 10-4 m-1 along with an estimated scale height of 2 km yield aerosol optical depths of 0.2-0.6 in agreement with observations. Regional-scale optical depth estimates coupled to a regional-scale atmospheric sulfur model for the eastern United States (Ball and Robinson, 1982) were suggested to produce an annual average loss of solar irradiance of 7.5 percent from sulfate and other anthropogenic aerosols relative to preindustrial times. There has been a corresponding, measured 3 to 4 percent loss of solar irradiance over Europe per decade during the past 40 years (Liepert et al., 1994). Measured light scattering by atmospheric aerosols is highly correlated with the measured masses of sulfate and organic compounds in the aerosol. A multiple regression analysis for the United States yielded a squared correlation coefficient of 0.95 (White, 1990). Maps of SO2 emission, sulfate aerosol concentration, acidic deposition, and extinction coefficient show geographical coherence over the eastern United States (Husar et al., 1981; Charlson et al., 1992b). Advanced Very High Resolution Radiometer (AVHRR) satellite imagery shows an enhanced aerosol optical depth in the Northern versus the Southern Hemisphere, with maxima in the vicinity of industrial regions (Durkee et al., 1991). In addition, the Stratospheric Aerosol and Gas Experiment (SAGE) data show a spring-and summertime enhancement of aerosol extinction by about a factor of 3 in the Northern Hemisphere midlatitude upper troposphere versus the Southern Hemisphere (Kent et al., 1991). As noted earlier in conjunction with Figure 1.1, most anthropogenic sulfate is produced in the Northern Hemisphere. Numerous quantitative estimates of the direct effect of anthropogenic sulfate alone have been made based on models of atmospheric radiative properties. Zero-, two-, and three-dimensional models of the sulfur cycle have been coupled to radiation transfer calculations, resulting in estimates from -0.3 to -1.3 W m-2 for the global average forcing. The two-and three-dimensional models suggest annual average maxima in the Northern Hemisphere midlatitudes from -4 to -11 W m-2. Uncertainties are estimated crudely as a factor of 2. Direct forcing by aerosols from biomass combustion has an even greater uncertainty, ranging from -0.05 to -0.6 W m-2 globally averaged. The Intergovernmental Panel on Climate Change (IPCC) has agreed on a ''best estimate" of -0.4 W m-2 for sulfate alone based on

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change Kiehl and Rodhe (1995), -0.2 W m-2 for the contribution from biomass burning and +0.1 W m-2 for soot based on Haywood and Shine (1995). A direct forcing of -0.5 W m-2 may have discernible climatic effects if it is regionally inhomogeneous. A more complete summary of direct forcing estimates is provided later in this report. It is currently estimated that about half of the background optical depth is the result of anthropogenic sulfate and organic aerosols (Andreae, 1995). As a result, based on a calculated global-scale cooling of 2-3°C from background aerosol (Coakley et al., 1983), a 1-1.5°C global-mean cooling can be estimated to be caused by anthropogenic aerosol. Indirect Forcing To quantify indirect climatic effects of aerosols requires relating increased mass concentrations of aerosol from anthropogenic sources to increased number concentrations of aerosol particles, to increased numbers of cloud condensation nuclei (CCN), to increased numbers of cloud droplets, to altered cloud radiative properties or lifetime. Increased droplet populations cause increased albedo for fixed liquid water path; therefore, increased numbers lead to negative forcing. Few studies exist that demonstrate the cause-and-effect relationships among these factors, although there are considerable data relating to the effect of anthropogenic emissions on CCN concentrations or cloud properties: It is well established that CCN concentrations are greater in anthropogenically influenced continental air masses than in the marine atmosphere; CCN concentrations in maritime air uninfluenced by anthropogenic emissions rarely exceed 100 per cubic centimeter (cm3), whereas concentrations in well-aged continental air generally exceed 1000 cm-3 (Pruppacher and Klett, 1978). Twomey et al. (1978) reported measurements of CCN concentrations exceeding 4500 cm-3 after relatively clean air (CCN levels as low as 50 cm-3) passed over an industrial area in southeastern Australia. Radke and Hobbs (1976) reported CCN concentrations of 1000 to 3500 cm-3 in air advecting off the eastern seaboard of the United States. Hudson (1991) reported CCN measurements along the west coast of the United States, indicating a background marine concentration of 20 to 40 cm-3, nonurban inland concentrations in Oregon of 100 to 200 cm-3, and urban concentrations in the vicinity of Santa Cruz, California, of 3000 to 5000 cm-3. Frisbie and Hudson (1993) report CCN concentrations of 500 and 5000 cm-3, respectively, upwind and downwind of Denver, Colorado. It has long been recognized that continental clouds tend to exhibit greater cloud drop number concentrations (CDNCs) than do marine clouds (Pruppacher and Klett, 1978). Numerous studies exist that link high CDNCs

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change are capable of obtaining globally relevant information on stratospheric aerosols. Quantitative consistency between satellite and aircraft/ground-based retrievals of stratospheric aerosol properties has been achieved (e.g., Russell and McCormick, 1989), providing robust estimates of transient stratospheric aerosol climate forcings. Observations following the major eruption of Mt. Pinatubo have offered a wealth of details regarding forcing and response characteristics associated with stratospheric aerosols (McCormick et al., 1995). Indeed, several aspects of present theoretical knowledge on the radiative and climatic effects of volcanic aerosols have been strengthened in the wake of these observations, as outlined below. Radiative forcing by stratospheric aerosols is governed essentially by the column burden of the particles and their sizes. Surface-troposphere forcing by stratospheric aerosols is less sensitive to aerosol composition and location within the stratosphere, but warming within the lower stratosphere does depend on altitude (Pollack and Ackerman, 1983; WMO, 1989). For particle effective radii of approximately 2 µm or less (typical for volcanic sulfate), cooling of the surface-troposphere system occurs (Lacis et al., 1992), with the sensitivity of the net radiative forcing estimated to be -2.5 to -3 W m-2 for an increase in midvisible optical depth of 0.1 (Harshvardhan, 1979; Lacis et al., 1992). Initially inhomogeneous, the forcing evolves spatially and temporally, consistent with microphysical mechanisms governing aerosol formation and removal, stratosphere-troposphere exchange, and global-scale transport. The eruption of Mt. Pinatubo in June 1991 yielded optical depth perturbations ranging from about 0.1 to 0.3, varying with location and time. The most optically thick portions of the aerosol were located between 20 and 25 km and were confined to 10°S-30°N during the early period (see Geophysical Research Letters 19, 149-218, 1992). Within two to three months, perturbed stratospheric optical depths were observed to at least 70°N, along with an enhancement in the Southern Hemisphere. Directly observed narrow and broadband total solar irradiance effects (IPCC, 1995a) indicate that reductions in surface solar flux ranged from <5 to 20 W m-2 in the diurnal mean, depending on prevailing aerosol optical depth. Satellite observations indicate a global mean decrease of about 5 W m-2 in the absorbed solar radiation in the period immediately following the eruption (IPCC, 1995a). Model computations of radiative forcing by Pinatubo aerosols are in broad agreement with the observations mentioned above. In one model study (Hansen et al., 1992), a global mean midvisible optical depth of 0.1 ten months after the eruption and decaying exponentially with a one-year time constant was considered. The corresponding global mean forcing was estimated to have a maximum value of about -4 W m-2. The volcanic aerosol forcing during the first two years after the eruption is estimated to

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change be greater than or comparable to GHG forcing over the past century and is substantially greater than the current decadal increase in GHG forcing (about 0.4 W m-2 per decade). Evidence for Climate Response to Stratospheric Aerosol Perturbation Stratospheric aerosol perturbations by volcanic eruptions tend to cool the surface-troposphere system. Time series of anomalies from satellite microwave measurements reveal that there was a nearly global, but nonuniform, tropospheric cooling following the Pinatubo eruption (Dutton and Christy, 1992), coinciding with the reduction in solar radiation reaching the troposphere. Cooling during the summer of 1992 was greatest in continental interiors (Hansen et al., 1993a). Although there are difficulties in isolating signals from volcanic versus other causes (Robock and Mao, 1994), attempts to identify patterns of climate response from large volcanic eruptions, by removing influences of El Niño/Southern Oscillation (ENSO) events, have shown that cooling following an eruption can last up to about two years, with an amplitude of approximately 0.1-0.2°C when averaged over the six largest eruptions of the past century. A general circulation model (GCM) investigation of the climatic impact from the 1991 Mt. Pinatubo eruption, using an initial optical depth of about twice El Chichon's and thereafter decaying with a one-year e-folding time (Hansen et al., 1992), predicted a transient cooling, maximum amplitude, and temporal evolution agreeing reasonably well with observations. The model-computed surface cooling ranged from 0.4 to 0.6°C, whereas the observed cooling estimated in 1992 (about one year after the eruption) ranged from 0.3 to 0.5°C. GCM investigations also revealed the possibility of dynamically induced responses in the tropospheric circulation patterns caused by volcanic perturbations (Graf et al., 1993). Heating of the tropical lower stratosphere resulting from an enhancement of the aerosol layer by volcanic injections leads to a transient, local temperature increase. This has been observed following eruptions of Mt. Agung, El Chichon, and most recently Mt. Pinatubo (IPCC, 1995a). Besides increasing temperature, the heating by aerosols could lead to anomalous upward motion in the lower stratosphere. This, in turn, would lead to adiabatic cooling and reductions in ozone concentrations (Kinne et al., 1992). Increases in stratospheric aerosols raise the potential for ozone destruction in the lower stratosphere via heterogeneous reactions on particles (Hoffman and Solomon, 1989). Observations in 1993-1994 after the Pinatubo eruption indicate that ozone values fell to unusually low levels. Changes in ozone concentrations are known to perturb the radiative balance (WMO, 1992).

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change Satellite observations (Minnis, 1994) suggest the potential for significant cloud modifications and hence an indirect radiative forcing from volcanic aerosols. Polarization lidar observations by Sassen (1992) in the northern midlatitudes following the Pinatubo eruption indicate that resulting changes in upper-tropospheric aerosols may have modified the microphysical and optical characteristics of upper-tropospheric clouds. Inferences from Stratospheric Aerosol Research Major volcanic eruptions provide a test of our ability to model climate change caused by transient, near-global, aerosol-induced perturbations. For the Pinatubo eruption, the fact that both computed forcing and modeled response of the climate system are in reasonable agreement with observations is encouraging. An important implication is that fundamental aspects of our knowledge on climate, at least for impacts from a moderately large and global aerosol forcing of limited duration (two to three years), appear to be sound. It is important to realize, however, that this volcanic aerosol test is an inappropriate analogue for a tropospheric aerosol forcing with a secular trend spanning several decades (Ramaswamy et al., 1995). In particular, the Pinatubo ''experiment" inadequately tests the role of the oceans in climate change and neglects land-sea contrasts, characteristic of forcing from tropospheric aerosols. In many respects, the episodic stratospheric aerosol forcing problem is far less complicated than the tropospheric aerosol climate problem. Nonetheless, the relatively better understanding of stratospheric aerosol climate effects does provide an extremely useful reference with which to compare and contrast forcings and responses from secular increases in tropospheric aerosols, since the optical depths in the two cases are comparable, albeit operating on different time and space scales. CLIMATE FORCING BY KEY TROPOSPHERIC AEROSOL TYPES Historically, many different classification schemes have been used to describe key aerosol types (e.g., see d'Almeida et al., 1991). An alternative to classifying aerosols by regional types, based on the notion of mass balances within global biogeochemical cycles, has been used increasingly for connecting atmospheric aerosol mass concentrations and burdens to the source strengths of aerosols or their gaseous precursors. Some of these aerosol types are defined simply by the chemical cycle of an element (sulfur, carbon, nitrogen, etc.), whereas others relate to a particular source (soil dust, sea salt, etc.). Because there is a wide variety of sources and molecular forms for carbonaceous aerosols, for example, there are several mass-balance

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change cycles to consider for this type of aerosol (photochemical oxidation products of natural and anthropogenic hydrocarbons; smoke from biomass combustion; and soot, particularly from fossil fuel combustion). Table 1.2 provides estimates of global source strengths and resultant optical depths for natural and anthropogenic aerosol types (Andreae, 1995). Whereas the anthropogenic aerosol mass flux is estimated to be only about 10 percent of the total, perhaps as much as about 50 percent of the global mean aerosol optical depth is anthropogenic; the reason follows from both shorter lifetimes and lower optical extinction efficiencies for soil dust and sea salt aerosols than for the major anthropogenic types (sulfates and smoke from biomass combustion). Because Table 1.2 displays the components of optical depth for each aerosol type, it is possible to estimate the increment of climate forcing from each type, as shown in Table 1.3. The estimated global mean optical depth figures are based on the "best-estimate" source strength given in Table 1.2, assumed average lifetimes, and mass extinction coefficients at ambient relative humidity for each of the aerosol types. No formal estimate of the overall uncertainty is given, nor is it available; however, despite the uncertainty that is implicitly present (e.g., in the range of flux estimates of Table 1.2), the singular message is clear. That is, given current best estimates of source strengths and aerosol properties, anthropogenic sulfates, organics, and soot are very likely to contribute a substantial fraction of the aerosol optical depth of the whole atmosphere. Table 1.3 summarizes estimates of climate forcing for sulfates, smoke from biomass combustion, and soot, with an additional entry for chemically undifferentiated "anthropogenic aerosols." With the exception of the early, low estimate by Bolin and Charlson (1976), which was based on the then very uncertain assumption—now known to be wrong—that anthropogenic sulfates are present over only 1 percent of the globe, all of the estimates of global mean forcing by sulfate alone fall in the range from -0.3 to -1.3 W m-2. The multiple-box models are based on only three chemical models (Langner and Rodhe, 1991; Pham, 1994; Taylor and Penner, 1994) and give global mean anthropogenic sulfate forcings ranging from -0.3 to -0.9 W m-2. Analysis of uncertainties of the estimates in Table 1.3 has thus far been cursory, ranging from a stated factor of 2 (Charlson et al., 1991) to approaches using the square root of sum of squares (Charlson et al., 1992a; Penner et al., 1994b). Consequently, given inadequate uncertainty analyses along with incomplete sensitivity tests, a single best estimate (and its uncertainty) of the global mean forcing is currently unavailable. Comparison of the calculated magnitudes of forcing by individual components in Table 1.3 suggests that the magnitude of the net forcing could be small if the lower magnitudes of sulfate and organic aerosol forcing are combined with the higher values for soot. It is clear that the magnitudes of the uncertainties at present are too large to allow a resolution of this situation and that more measurements

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change TABLE 1.2 Source Strength, Atmospheric Burden, Extinction Efficiency, and Optical Depth for Various Types of Aerosols   Flux (Tg/yr) Mass Extinction Coefficient (m2 g-1) Estimated Global Mean Optical Depth Source Low High Best     Natural Primary Soil dust (mineral aerosol) Sea salt (mass mean diameter 1,000 3,000 1,500 0.7a 0.023 = 5 μm, σg = 2 1,000 10,000 1,300 0.4b 0.003 Volcanic dust 4 10,000 33 2.0 0.001 Biological debris 26 80 50 2.0 0.002 Secondary Sulfates as (NH4)2SO4 from natural precursors 85 210 102 5.1 0.014 Organic matter from biogenic VOCs (as C) 40 200 55 5.0 0.011 Nitrates from NOx 15 50 22 2.0 0.001 Subtotal 2,200 23,500 3,062   0.055 Anthropogenic Primary Industrial dust, etc. 40 130 100 2.0 0.004 Soot (elemental carbon) 5 20 10 10.0 0.003 Secondary Sulfates as (NH4)2SO4 from SO2 120 250 140 5.1 0.019 Biomass burning (as C) 50 150 80 5.0 0.017 Nitrates from NOx 25 65 36 2.0 0.002 Organic from anthropogenic VOCs (as C) 5 25 10 5.0 0.002 Subtotal 300 650 376   0.047 Total 2,500 24,000 3,438   0.102 NOTE: VOCs = volatile organic compounds. a Another value for the soil mass extinction coefficient is 1.0 m2 g-1 (cf. Malm et al., 1994), and values well above 2 m2 g-1 have been inferred (e.g., Ouimette and Flagan, 1982). White has recommended a value around 0.7, that given in this table (White, 1990). b The sea salt mass extinction coefficient is based on a purely coarse mode sea salt, but recent studies have suggested many more submicron particles than had previously been believed (cf. O'Dowd and Smith, 1993; McInnes et al., 1994). The mass scattering efficacy could be higher than the value of 0.4 m2 g-1 cited. ADAPTED FROM: Andreae (1995); IPCC (1995); Kiehl and Rodhe (1995).

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change are needed to resolve the issue of the overall magnitude of anthropogenic aerosol forcing. Similarly, predictions for geographical and seasonal variations of forcing by anthropogenic sulfates differ among available models. These variations may be more important than differences in global mean values for estimates of meteorological effects. Despite these differences in predictions, all models show large negative forcings over the industrial regions of the United States, Europe, and Asia that are greater than calculated GHG forcings in these regions. Because there have been no estimates of uncertainties of predicted geographical variations, selection of a best or most probable estimate is impossible. Nonetheless, data exist for industrial regions that are consistent with predictions of large negative forcings (Liepert et al., 1994); it is therefore unlikely that predictions are so inaccurate that sulfate aerosol forcing can be neglected. Uncertainties in direct forcings by aerosol types other than sulfate are even less reliable; therefore, the same general conclusion applies—that selection of a "best estimate" or "central value" is fraught with unacceptably large uncertainties. For indirect forcing, the most obvious unifying theme among available predictions is author admission of model inadequacies. If the observed decreased droplet size in Northern Hemisphere clouds relative to those in the Southern Hemisphere given by Han et al. (1994) is caused by anthropogenic aerosol, the required approximately 10 to 50 percent estimated enhancement of CCN would correspond to forcings in the range of about -0.5 to -3 W m2 (Kaufman et al., 1991; Boucher and Rodhe, 1994; Jones et al., 1994). This estimate, along with the frequently demonstrated observation that anthropogenic pollution enhances CCN population (thereby causing larger numbers of smaller droplets), strongly suggests that some amount of negative forcing exists and that its magnitude may be significant (e.g., see Hobbs et al., 1974). Table 1.4 lists key anthropogenic aerosol types, their forcing mechanism(s), and brief assessments of current understanding. CONCLUSIONS It is our judgment that climate forcing by anthropogenic aerosols is likely to be of sufficient magnitude to necessitate its representation in models of climate change over the industrial period; and present estimates of anthropogenic aerosol forcing are sufficiently uncertain as to be inadequate to usefully represent this forcing in models of climate change over the industrial period.

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change TABLE 1.3 Estimates of Direct Climate Forcing (W m-2) by Anthropogenic Aerosols Aerosol Type Global Mean Forcing Regional Maximum Forcing Reference as Basis of Estimate Anthropogenic sulfate alone -0.1 to -0.2 -10 to -20 Bolin and Charlson (1976);a turbidity data   -1.6 — Charlson et al. (1990); single-box model   -0.6 -4 (eastern Charlson et al. (1991); global 3-D chemical/radiative model     Mediterranean Sea) Langner and Rodhe (1991); slow oxidation rate case     -2 (eastern U.S.)     -1.3 — Charlson et al. (1992a); single-box model   -0.3 -4.2 (Europe) Kiehl and Briegleb (1993); 3-D chemical model and GCM     -3 (eastern U.S.) Same as Kiehl and Rodhe (1995, Figure 6b); slow oxidation case of Langner and Rodhe (1991)   -0.66 -11 (central Europe) Kiehl and Rodhe (1995), Pham (1994); chemical model     -5 (eastern U.S.)     -0.45 (est.) -6 (est.) Kiehl and Rodhe (1995); standard oxidation case of Langner and Rodhe (1991)   -0.95 -4 (over Europe) Taylor and Penner (1994)     -3 (eastern U.S.)     -0.56 to -0.94 — Penner (1995); single-box model   -0.36 to -0.79 — Haywood and Shine (1995)

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change Organic, carbonaceous -0.1 — Penner et al. (1992); biomass combustion aerosols -0.5 — Penner (1995); single-box model Soot +0.35   Penner (1995)   +0.05 to +0.27   Haywood and Shine (1995) Chemically undifferentiated  -1 .4 to -2   Coakley et al. (1983) or mixed anthropogenic aerosols ca. -0.25 (sulfate plus soot) — Hansen et al. (1993b)   -0.5 (total, including biomass smoke)       ca. -2 (summer) -7 (summer 50°N) Grassl (1988); estimated, based on zonal-mean model   ca. -0.5 (winter) -1.4 (winter 30°N) No soot   ca. -0.7 (summer) -2.9 (summer) Grassl (1988); including 20% soot   ca. -0.2 (winter) -0.4 (winter)     -0.29 -3 (Europe) Haywood and Shine (1995). Langner and Rodhe (1991); slow oxidation case external mixture (7.5% soot)     -2 (eastern U.S.)     -0.15 -2.5 (Europe) Haywood and Shine (1995); internal mixture     -1.5 (eastern U.S.)   NOTE: 3-D = three-dimensional; est. = estimated. a Forcing was extracted from the Bolin and Charlson (1976) estimate of optical depth for sulfate by applying a sensitivity of forcing to optical depth of 30 W m-2 per unit optical depth (Charlson et al., 1991). Forcing by soot is taken to be a positive number that reduces the magnitude of negative forcing from sulfate because of the presence of soot.

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change TABLE 1.4 Key Anthropogenic Aerosol Types, Associated Forcing Mechanisms, and Status of Understanding Key Anthropogenic Aerosol Types Forcing Mechanism(s) Status of Understanding 1. Water-soluble inorganic species (e.g., sulfate, nitrate, ammonium) from atmospheric reactions of precursor gases (e.g., SO2, NOx, NH3 ); key sources include combustion of fossil fuels and smelting of sulfide ores a. Direct, clear-sky upscatter of solar radiation i. Fundamental scattering theory is fully understood for spherical particles; well in hand for nonspherical ones ii. In situ and remote measurement methods are available for quantifying the direct optical effect and the size-resolved chemical composition. Satellite measurements currently are limited; promising possibilities exist for large improvements iii. Coupling of local radiative properties of aerosol to energy balance is well understood iv. Relationship of aerosol optical and chemical properties to relative humidity is known in principle for pure compounds; the role of mixing with organic species is yet to be explored v. Quantitative connection has been attained relating SO2 source strength to geographically dependent sulfate aerosol concentration; uncertainty approximately a factor of 2 in the newest models. Less is known about other compounds   b. Indirect effect of CCN on cloud albedo i. Theory suggests increased CCN number concentration should increase cloud albedo, if all other factors are held constant ii. Empirical evidence suggests that anthropogenic aerosols increase cloud albedo iii. Relationship of aerosol mass concentration to CCN number concentration is practically an open question   c. Indirect effects of CCN on cloud droplet lifetime i. Problem almost entirely open

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change 2. Condensed organic species from atmospheric chemical reactions of reactive organic gases or from smoke produced by biomass combustion a i. Same as 1.a.i-iii ii. Connection to source strength of reactive organic gases is expected from laboratory work but not yet quantified in the field iii. Relationship to relative humidity remains to be explored   b i. Same as 1.b.i-iii ii. Role of water-soluble and partly soluble organics modifying cloud nucleating properties is an almost open question   c i. Role of organics in modifying cloud droplet lifetime is an open question 3. C(0), elemental or black carbon soot from incomplete combustion (e.g., diesel fuel) d. Absorption of solar radiation; change in vertical temperature profile i. Fundamental theory of light absorption by aerosols is well established ii. Radiative transfer theory for absorption is also well established iii. Data on presence or amount of light-absorbing particles (especially elemental and black carbon) are sparse; in situ methods are available iv. Connection of amount of absorbing aerosol to source field is possible in principle but remains to be done 4. Mineral dust, windblown soil, and desert dust a, b, c, d, and absorption/emission of longwave (IR) radiation i. Same as 1.a.i-iii ii. No estimates yet available for fraction of soil dust that is anthropogenic iii. Little information is available on cloud nucleating properties of soil dust. May be a problem of second-order importance because of large particle sizes and relatively small number concentrations

OCR for page 7
A Plan for a Research Program on Aerosol Radiative Forcing and Climate Change We recommend that the uncertainties in calculated aerosol forcing at the top of the atmosphere be reduced to within ±15 percent both globally and locally. This limit of uncertainty is equivalent to that required in estimating greenhouse gas forcing (IPCC, 1995a). Locally this would imply an uncertainty in forcing of less than 1.5 W m-2 (by assuming a local aerosol effect of -10 W m-2 in the diurnal mean). This report presents our opinions about needed scientific studies (measurements, observations, model developments); the technologies required (satellites, computers, aircraft, instruments); the necessary resources; and an implementation plan for a U.S. multiagency program to answer these scientific questions and, thereby, to improve climate models. Chapters 3 and 4 contain our recommendations for implementing a focused, integrated research program. First, however, Chapter 2 describes needed research.