The indirect radiative effect of aerosols
Contributed by K. Noone, Department of Meteorology, Stockholm University, Sweden
Determining the radiative effects of aerosols is currently one of the most active areas in climate research. Aerosols influence the Earth’s radiative balance directly by scattering incoming shortwave radiation back to space, or indirectly through their influence on cloud properties. The indirect effect is considered to be one of the largest uncertainties in current global climate models (GCMs). Correctly predicting the indirect radiative effect of aerosols requires that we understand the interactions between aerosols and clouds that determine cloud properties.
There has been a large amount of very interesting new work done recently on the subject of indirect radiative effects—far too much to attempt to present a comprehensive review here. Rather, the purpose of this note is to present a short tutorial on the subject, and to highlight a few of the original papers and recent advances.
Different facets of the indirect effect
There are several indirect ways that aerosols influence the radiative balance of the Earth. Aerosols contribute to determining the reflectivity of clouds—cloud albedo. They also affect precipitation processes. Aerosols influence the initial droplet size distributions produced close to cloud base, and can subsequently influence the effectiveness of coalescence at a later stage of cloud development through changing the spread in drop sizes. Aerosols also influence freezing processes in mixed-phase clouds. Both these mechanisms (coalescence and freezing) influence precipitation development. Precipitation is a key component in determining the lifetime and extent of clouds. It also is a key component in the atmospheric energy balance through the redistribution of latent heat. Finally, it is possible that aerosols can influence the dynamical processes in the atmosphere that drive cloud formation and development.
These phenomena are illustrated in Figure 1, a cartoon of aerosol influences on cloud properties. The left panel represents a clean (background) case with a given cloud albedo, extent, precipitation intensity and dynamic
The Albedo Effect
Perhaps the first to be recognized (and receive the most attention) is the albedo or “Twomey” effect—the increase in cloud albedo due to an increase in aerosol concentration. For a dynamic forcing that creates a cloud with a given vertical extent and liquid water content, an increase in aerosol concentration going into the cloud can result in the formation of a larger number of smaller droplets as compared to an unperturbed cloud. The end result is in an increase in cloud albedo [Twomey, 1974, 1977a].
An early observation that anthropogenically-produced aerosols could change cloud properties came when “anomalous cloud lines”—bright curvilinear structures in low clouds—were seen in early TIROS satellite imagery [Conover, 1966]. The suspicion was that these areas of increased cloud albedo were caused by aerosol emissions from ships [Conover, 1966; Twomey et al., 1968]. These features have subsequently become known as ship tracks and serve as a textbook example of the effect pollution aerosols can have on the albedo of warm clouds.
Experiments in marine stratocumulus clouds have confirmed the general conceptual model of this effect [Coakley et al., 1987; King et al., 1993; Radke et al., 1989]. Several recent large process-oriented field experiments have been dedicated to investigating the relationship between aerosols and cloud albedo [Brenguier et al., 2000; Durkee et al., 2000].
The albedo effect does not necessarily imply an increase in cloud reflectivity. The reflectance (R) of a cloud is a function of several variables: R = R(t, w0 g) where t is the optical depth of the cloud, w0 is the single scattering albedo of the droplets, and g is the asymmetry parameter (describing the fraction of energy scattered into the forward direction). An aerosol/cloud interaction that can change any of these parameters can also affect cloud reflectivity. For instance, w0 is approximately unity for water droplets at visible wavelengths. If a significant amount of absorbing aerosol were to be incorporated into cloud droplets, it could reduce w0 and decrease R. On the whole, however, increasing aerosol concentrations are expected to increase cloud albedo for all but the thickest clouds [Twomey, 1977b].
Precipitation: Cloud lifetime, extent and energy redistribution
The same processes that increase cloud albedo in low-level clouds (a production of more and smaller droplets) tend to decrease the efficiency with which precipitation is formed. Albrecht  proposed that a decrease in drizzle production in these kinds of clouds could increase both the cloud liquid water content (and thus liquid water path) and the fractional cloudiness. Aerosol-induced precipitation suppression has been observed both with in situ measurements [Ferek et al., 2000], and in satellite observations [Rosenfeld, 2000] showing that the effect does in fact occur in the atmosphere. Some time ago, Warner  deduced that aerosol emissions from sugar cane fires were affecting precipitation by examining rainfall records from western Australia.
If precipitation is suppressed, water that would have been removed from the atmosphere remains aloft and can be transported to other locations before it is deposited to the surface. The same is true for the energy associated with this water—the latent heat released on condensation in clouds and the energy required for evaporation of water from the surface. This redistribution of water and latent heat due to precipitation suppression may have the potential to influence circulation patterns.
Another kind of indirect effect may arise due to the presence of absorbing aerosols. It has been hypothesized that heating of the boundary layer by absorbing aerosols may influence cumulus cloud formation by stabilizing the layer and reducing relative humidity [Ackerman et al., 2000]. GCM calculations indicate that while this effect may reduce cloudiness in some highly polluted locations, the increase in cloud albedo and lifetime are still predicted to dominate on the global scale [Lohmann and Feichter, 2000]. Field experiments in polluted winter fogs showed that despite the presence of considerable amounts of absorbing aerosol, fog still persisted for several days [Hallberg et al., 1992; Noone et al., 1992].
These results are not necessarily contradictory—what they do show is that this subject is not yet very well understood. The number and complexity of the steps in the causal relationship between aerosols and precipitation makes quantifying this indirect radiative effect very challenging. Even more challenging is understanding and quantifying the consequences of the effect in terms of circulation patterns and energy redistribution. Clearly, there is still plenty of work to be done in this area.
The previous discussion has implicitly been related to clouds in (or at least originating in) the lower troposphere—e.g., marine stratocumulus and tropical cumulus clouds. Aerosols affect cirrus cloud properties as well. Here, our knowledge of aerosol indirect effects is even more rudimentary than for the case of warm clouds. In fact, it is difficult to even come up with a sign for the effect, much less quantify its nature and magnitude.
The difficulty arises because cirrus clouds can act very differently in terms of their radiative properties depending on their optical thickness. Optically thin cirrus clouds let most of the incoming solar radiation pass through, but tend to absorb outgoing infrared radiation emitted from the Earth’s surface. They re-radiate this energy—but at a much lower temperature. In this way, they behave similarly to greenhouse gases, and can have a net warming effect. Optically thick cirrus clouds, on the other hand, reflect much of the incoming solar radiation. They too absorb and re-radiate outgoing infrared radiation, but if the loss of solar radiation dominates the energy balance, the net effect is cooling. An additional complication arises from the fact that cirrus clouds contain non-spherical crystals, the optical properties of which are more difficult to describe compared with homogeneous liquid spheres.
Of particular interest in this regard is the effect of aircraft emissions on cirrus cloud properties. Several field campaigns and modeling efforts have been undertaken recently to examine how aircraft emissions influence cirrus clouds [e.g., Brasseur et al., 1998; Schumann et al., 2000; Toon and Miake-Lye, 1998]. In one study, Ström and Ohlsson  used measurements of absorbing aerosol material inside cirrus cloud elements to show that crystal concentrations were enhanced in areas affected by aircraft exhaust. Like ship tracks, this is a case where a clear causal link between a specific aerosol emission and a cloud effect was identified. Unfortunately, our knowledge of the microphysical properties and chemical composition of the particles emitted by jet engines is extremely limited. Without this information, understanding and predicting the effects of aircraft emissions on cirrus cloudiness will remain an uncertain endeavor. As with the case for precipitation effects, a great deal of work remains to be done to determine the indirect radiative effects of aerosols on cirrus clouds.
The process that ultimately drives the indirect radiative effect of aerosols—that of cloud droplet (or crystal) formation—happens on time scales of seconds and spatial scales of micrometers. Yet the end result of this process— the clouds themselves—is a global scale phenomenon. One of the largest challenges in understanding and predicting the indirect radiative effect of aerosols is bridging the gap between these spatial and temporal scales in terms of both our understanding of the processes involved, and in our descriptions of the processes in global-scale models. Making progress in this area requires that modelers and experimentalists work together and communicate effectively with each other. IGAC has played a tremendously important role in bringing these communities together to make real advances in the area of atmospheric chemistry. It can also serve as a unifying forum for the exchange of information and ideas in the area of climate research as well.