Aerosols and climate

B. Geerts


What are aerosols?

Aerosols are fine, airborne particles consisting at least in part of solid material. Density of the basic materials of aerosols range from 1.0 g/cm3 (for soot) to 2.6 (for minerals). The ocean is a major source of natural aerosols. Air-sea exchange of particulate matter contributes to the global cycles of carbon, nitrogen, and sulfur aerosols, such as dimethylsulfide (DMS) produced by phytoplankton. Ocean water and sea salt are transferred to the atmosphere through air bubbles at the sea surface. As this water evaporates, the salt is left suspended in the atmosphere. Four other significant sources of aerosols are terrestrial biomass burning, volcanic eruptions, windblown dust from arid and semi-arid regions, and pollution from industrial emissions (Fig 1).

Clean continental air often contains less than 3,000 particles per cubic centimetre (of which half are water-soluble), polluted continental air typically 50,000/cm3 (of which two-thirds are soot, and the rest mostly water-soluble). Urban air typically contains 160,000/cm3, mostly soot, and only 20% is water-soluble. Desert air has about 2,300/cm3 on average, almost all water-soluble. Clean marine air generally has about 1,500/cm3, about all water-soluble. The lowest sea-level values occur over the oceans near the subtropical highs (600/cm3 on average, but occasionally below 300/cm3). Arctic air has about 6600/cm3 (including 5,300 soot) and on the Antarctic plateau only 43/cm3 occur (about all sulphate) (1).

Fig 1. Main sources and types of aerosols that affect climate.

Concentrations of aerosols decrease about exponentially with height, thus

N(z) = N(0) exp (-z/H)

where N(z) is the concentration at height z (km) above ground level. The scale height H typically is 10 km for continental (incl desert) air in summer, 6 km for urban air, 2 km for continental air in winter, and 1 km over the marine subtropical high pressure regions. The higher values during the summer imply more deep-tropospheric stirring of the aerosols by thunderstorms. In places with a well-defined, long-lived inversion on top of a well-mixed planetary boundary layer, the aerosol concentration tends to be homogenous in the PBL and perhaps 10x lower above the inversion.

Aerosols can be long-lived and therefore be advected over a long distance. Four weeks of measurements of tropospheric aerosols at Mildura (in the Australian state of Victoria) showed that some came from Africa, about eight days upwind, and possibly South America, even further away. These aerosols may have come from gases generated by forest fires (2).


How aerosols affect climate

Aerosols play an important role in the global climate balance, and therefore they could be important in climate change. Natural variations of aerosols, especially due to episodic large eruptions of volcanoes, such as Mt. Pinatubo in 1991, are recognized as a significant climate forcing, that is, a factor that alters the Earth's radiation balance and thus tends to cause a global temperature change. In addition, there are several ways in which humans are altering atmospheric aerosols, not only near the ground (e.g. industrial emissions) but as high as the lower stratosphere (where they are continuously emitted by aircraft), and thus possibly affecting climate (e.g. through contrails) (3).

Aerosols force climate in two ways (4):

Greenhouse gases have a well-understood effect on the global radiative balance and surface temperatures, their concentration has little variability (except water vapour and ozone), and their long-term trends are well-known. Therefore, there is much confidence in the greenhouse gas component of the anticipated climate change during the next few decades. However climate forcings due to aerosols are not determined well, especially the indirect radiative forcing. Indeed, aerosols are one of the greatest sources of uncertainty in interpretation of climate change of the past century and in projection of future climate change.

The effect of aerosols on clouds is highly speculative. The theory is that the more aerosol, the smaller the cloud droplets tend to be, and clouds with more but smaller drops have a higher albedo. This would increase the planetary albedo, i.e. have a cooling effect. Twomey (5) proposed the first step in this theory, by showing empirically that the mean droplet mass in a cloud decreases in proportion with the number concentration of aerosols (N). In other words, the mean droplet radius (r) is proportional to:

r ~ N-0.33 or Dr/r = -(DN/N)/3

So if the aerosol concentration increases by 30% (DN/N = 0.3), the droplet radius decreases by 10%, and for the same total cloud water content, the number of droplets will increase by 30%.

While this relation has been corroborated surprisingly well in various field experiments, it is uncertain whether the cloud water content will be conserved. And the radiative forcing of clouds depends strongly on the heights of their bases and tops, which are unknown. Also, a variation in the drop size of a cloud will affect the way in which the cloud will evolve, produce rain or evaporate. In some circumstances aerosols may create clouds where none existed before, because they act as cloud condensation nuclei. Satellite imagery over ocean areas prone to stratus clouds (especially over high-pressure regions and/or low sea surface temperatures) reveal that ships often trigger lines of stratus along their track. These 'ship tracks' increase the albedo and cause net cooling. Another example is the production of DMS by phytoplankton (3). Some of this DMS seeps into the atmosphere where the sulphate aerosol enhances cloud formation. Accurate measurements at Cape Grim in northwest Tasmania have shown that DMS is the main agent of the nucleation of clouds over the southern oceans (6). The surface waters within the large ocean gyres are generally depleted of nutrients, especially iron. An experiment has shown that the seeding of iron dramatically increases the DMS production in the ocean. The fractions of aerosols that are manmade, biogenic, volcanic, or soil-based, and the chemical reactions of some aerosols, are poorly understood. The inclusion of the ice phase adds an extra step of complexity. In short, the effect of aerosols on clouds and thus on climate is very uncertain.

 The climate forcing of greenhouse gases, aerosols, and other variants, is commonly expressed in terms of the resulting net change to the radiation balance, expressed in W m-2. What really matters of course is how much warming or cooling a variant bears. For instance, the net warming of the Earth's atmosphere (as compared to an atmosphere without greenhouse gases, aerosols or clouds) is 33K at the surface (from -18 C to +15 C, Section 2.8). The reason for the use of radiation units rather than temperature units is that the effect of some perturbation (such as a volcanic eruption) on surface temperatures involves a cornucopia of complex feedback effects, such as atmospheric and ocean circulations. The radiative forcing is more 'raw', and allows more direct comparisons, e.g. between general circulation models (GCMs).

The aerosol concentrations depend on the wind, the land surface conditions (vegetation cover), sea surface temperature, and other climate factors. During an Ice Age the aerosol types and distributions were very different from those during an interglacial. It is possible that aerosol acts as a positive feedback in enhancing the difference between glacials and interglacials. For instance, during a glacial, the stronger winds over denuded periglacial plains might pick up more dust, contributinging to a higher planetary albedo, maintaining the cold. In other words, climate and aerosol are inter-related in a complex way.


Measuring aerosol concentrations

There are many surface stations that measure at least some component of aerosols (their size, type, concentration). These measurements tend to be biased by the local/regional industrial activities, vehicle transport, land surface cover and soil type, and winds. There are only few stations that are truly remote from sources of manmade aerosols, e.g. the Cape Grim Baseline Air Pollution Station in Tasmania, or the Mauna Loa Observatory in Hawaii. In short, in view of local sources, the high spatial/temporal variability of aerosols, and the complexity of aerosol properties, aerosols are much undersampled to really understand their distribution.

Aerosol could be sampled remotely, from satellite, but only indirectly, by means of a combination of various wavelengths of upwelling radiation (7). Even then numerous assumptions are required. At this time there is no satellite instrument that is specifically designed to measure aerosol. There are two instruments that can be used to estimate the spatial distribution of aerosol optical thickness, but not their vertical profile. The optical thickness is an integrated amount, and in the visible spectrum it is a measure of haziness.

One satellite instrument is the AVHRR (Advanced Very High Resolution Radiometer). A combination of spectral bands was used to estimate the global-mean aerosol optical thickness, shown in Fig 2 (8). An optical thickness of 1 means that no radiation (in this case light at 0.55 micron) directly penetrates through the atmosphere. The high values at low latitudes (esp. around Africa) may be largely due to fires in support of subsistence farming. The AVHRR-based technique performs poorly over land because of the high variability of radiative properties of the land surface. Even over the ocean the signal-to-noise ratio is not much more than one. The other instrument is TOMS (Total Ozone Mapping Spectrometer). TOMS data can be used to detect absorbing aerosols over land, but are insensitive to aerosols located below 1 km, including the planetary boundary layer in which most aerosols reside.

More accurate aerosol measurements will soon be taken by the MODIS instrument to be launched soon on the Terra satellite. MODIS (Moderate Resolution Imaging Spectrometer) will measure global variations in aerosol optical thickness over land and oceans, as well as global aerosol particle size distribution over oceans (as there is currently no simple way to derive aerosol particle size distribution over land). The algorithms for remote sensing of aerosols over oceans and land (MODIS Aerosol Product) are very different from one another because of differences in the spectral reflectance of water and land under the semi-transparent aerosol layer. MODIS's objective is to provide a comprehensive series of high-resolution global observations of the Earth's land, oceans, and atmosphere in the visible and infrared regions of the spectrum in such a way as to view the entire surface of the Earth every two days. MODIS will continue to take measurements in spectral regions that have been and are currently being measured by other satellite sensors, such as the AVHRR (9).

Recently (1998) the Global Aerosol Climatology Project was established to obtain a better insight of aerosol concentrations and types, by means of satellite and surface data, field experiments and numerical simulations. The movement and chemical evolution of aerosols can be simulated by means of GCMs that include cloud microphysics, but until recently these models have used crude assumptions, and they could be initialized better by means of satellite-retrieved data.



  1. Hess, M., P. Koepke and I. Schult 1998. Optical properties of aerosols and clouds. Bull. Amer. Meteor. Soc. 79, 831-44.
  2. Young, S. 1999. Aerosol over south-east Australia. Atmosphere (Newsletter of the Atmospheric Research Division of the Australian Commonwealth Scientific & Industrial Research Organisation), 6, 8-9.
  3. Charlson, R.J., S.E. Schwartz, J.M. Hales, R.D. Cess, J.A. Coakley, Jr., J.E. Hansen, and D.J. Hoffman 1992: Climate forcing by anthropogenic aerosols. Science, 255, 423-430.
  4. Hansen, J., Mki. Sato, A. Lacis, and R. Ruedy, 1997: The missing climate forcing. Phil.Trans. Royal Soc. London. B 352, 231-240.
  5. Twomey, S., M. Piepgrass, and T.L. Wolfe, 1984: As assessment of the impact of pollution on global cloud albedo. Tellus, 36B, 356-366.
  6. Pearman, G. 1996. Remark at a Climate Change conference at Canberra (Australian Academy of Science.)
  7. Rosenfeld, D. and I.M. Lensky, 1998: Satellite-based insights into precipitation formation processes in continental and maritime convective clouds. Bull. Amer. Meteor. Soc. 79, 2457-2476.
  8. Husar, R.B., J.M. Prospero, and L.L. Stowe, 1997. Characterization of tropospheric aerosols over the oceans with NOAA Advanced Very High Resolution Radiometer optical thickness operational product. J. Geophys. Res., 102, 16889-909.
  9. King, M. D., Menzel, W. P., Grant, P. S., Myers, J. S., Arnold, G. T., Platnick, S. E., Gumley, L. E., Tsay, S.-C., Moeller, C. C., Fitzgerald, M., Brown, K. S., & Osterwisch, F. G., 1996. Airborne scanning spectrometer for remote sensing of cloud, aerosol, water vapor and surface properties. J. Atmos. Ocean. Tech., 13, 777-794.