Air Pollution as a Climate Forcing: A Workshop

Day 2 Presentations

Aerosol Forcings Inferred from INDOEX and Estimated Regional Effects

A. Jayaraman
Physical Research Laboratory, Ahmedabad, India

Introduction. Prior to the onset of the summer monsoon the surface wind over the Indian subcontinent is predominantly north-easterly, flowing from the continent towards the Indian ocean, transporting large amount of aerosols and precursor gases. The Indian Ocean Experiment (INDOEX) was a major field experiment (1) conducted from 1996 to 1999 with an objective to assess the climate forcing caused by both the natural and anthropogenic aerosols over the tropical Indian Ocean. On a global average, the climatic cooling effect of aerosols is shown (2) to compete with the warming caused by the increase in the greenhouse gases concentration. But, over regions of large aerosol concentration, such as over the ocean region surrounding the peninsular India, the aerosol effect could be much larger than the greenhouse effect and the forcing sign could be positive or negative, depending on whether the particles are absorbing or scattering. INDOEX has shown (3) that during winter months anthropogenic haze spreads over much of the Asian region and the northern Indian Ocean. As this happens, prior to the onset of the summer monsoon it is important to examine the impact of this continental size haze on the radiation budget and the climate system.

The Study Area. During winter months the polluted air brought from the northern land mass covers almost the entire region north of the inter-tropical convergence zone (ITCZ) bound by the land masses of Arabia, Africa in the west and the southeast Asian regions, Myanmar, Thailand, Malaysia in the East. The tropical Indian Ocean region, including the Arabian Sea and the Bay of Bengal is polluted by the intense influx of anthropogenic aerosols, trace gases and their reaction products (4). At the south of the ITCZ the air is relatively pristine and is very little influenced by the continents. During the northern winter months the ITCZ is generally located between the equator and 10°S latitude. The study region is over the tropical Indian Ocean from 15°N to 20°S and 60°E to 80°E.

Chart of aerosol optical depth.

Figure 1: Latitudinal variation in the aerosol optical depth measured by the Indian R/V Sagar Kanya, during January to March, 1996-99.

Aerosol Characteristics over the Study Region. For aerosol radiative forcing studies we are concerned with the total column concentration of aerosols, distributed from the surface to the top of the atmosphere. This is inferred by measuring the attenuation of the incoming solar radiation. The aerosol optical depth, δλ at a wavelength λ is defined such that, the surface reaching direct solar radiation intensity decreases by exp (-δλ/cos θ) due to aerosol scattering and absorption, where θ is the solar zenith angle measured from the vertical. Figure 1 shows the aerosol optical depth measured at λ = 500 nm during the ORV Sagar Kanya cruises conducted during January to March months from 1996 to 1999.

High δλ observed near the Indian coast (10° -18°N) is a common feature in all years except that differences exist in the absolute values. The steep latitude gradient shows the increasing concentration of the sub-micron size particles as one approaches the continent. Typically a δλ value of 0.2 or less for the visible region is considered to represent the background value and any value above 0.4 represents polluted air. Measurements show that the 1996 and 1998 aerosol optical depth values are lower compared to the 1997 and 1999 values. Also the amplitude of the annual variation is larger for the lower wavelength optical depth compared to higher wavelengths. Anthropogenically produced sub-micron particles brought from the continents contribute significantly to the lower wavelength optical depth over the ocean and the observed year to year variation is predominantly due to variations in meteorological conditions for the transport and dispersal of the pollutants. In the present study we consider only the average aerosol loading around the continent for the radiative forcing estimation.

Another important feature found over the Arabian Sea is that the concentration of the nucleation mode particles (size less than 0.1 μm) is very high, in the range of 20 to 50 μg/m3. As the precursor gases for the formation of these sub-micron particles come mostly from the continents4, the newly nucleated particles are also found in great abundance near the coastal region. In contrast the concentration of the nucleation mode particles south of the ITCZ is less than 10 μg/m. Mass concentration values are converted to number density distribution taking appropriate aerosol density and corrected for the ambient relative humidity variation. Figure 2 shows the average aerosol size distribution obtained for the Arabian Sea for the winter months and corrected to 50% RH. More details on the measurements are available elsewhere (5). The standard deviation of the mean is shown as vertical bars.

Three distinctive modes are seen in the distribution that could be individually fitted using log-normal distribution curves as shown in the figure. The major difference between the aerosol size distribution obtained over the Arabian Sea and that for a "standard ocean atmosphere" available in the literature (6) is that the number of nucleation mode particles is more than two orders of magnitude higher over the Arabian Sea and the concentration is comparable to that commonly found in polluted urban locations. Particle concentration obtained in the accumulation mode (0.1 to 1 μm) is more or less similar to model values available in the literature for other ocean regions while the coarse mode concentration is again higher due to the direct influence of primary particles brought from the continent. This indicates the strong influence of the continental aerosols in the marine boundary layer over the Arabian Sea. Aerosols that are important to the radiative forcing are mainly the nucleation and accumulation mode particles.

-- FIGURE 2 -- Figure 2. Average aerosol size distribution obtained from the measured aerosol mass concentration over the Arabian Sea and corrected for 50% relative humidity.

By combining two independent measurements viz., the scattering coefficient using nephelometer and absorption coefficient using Particle/Soot Absorption Photometer (PSAP) the aerosol single scattering albedo, ω is estimated, which is the ratio of the scattering coefficient to the total extinction (scattering + absorption) coefficient. The ω value is found to be in the range of 0.8 to 0.9 (with an uncertainty of about ±4%) for the polluted region and about 0.95 or more for the pristine ocean region, south of the ITCZ. Low ω values are found more close to the Indian coastal region, where the measured columnar aerosol optical depth values are also high. For a ω value of 0.85, and aerosol optical depth in the range 0.4 to 0.7 in the visible region the columnar aerosol absorption amounts to 0.06 to 0.11. This is much higher than the combined absorption by all known molecular gases, in the visible region of the solar spectrum. Carbonaceous aerosols (soot particles) from fossil fuel and biomass burning are the main contributors to the aerosol absorption (7).

Aerosol Radiative Forcing. A forcing efficiency term (8) is defined as the ratio of the change in the surface reaching solar radiation intensity, ∆F(down) in W/m2 or the upscattered solar radiation flux ∆F(up), leaving the atmosphere to the change in columnar aerosol optical depth, ∆δ. The forcing efficiencies at the surface (∆F(down/∆δ) and at the top of the atmosphere (∆F(up)/∆δ) are estimated separately and the difference between the two terms is defined as the net aerosol absorption efficiency within the atmosphere. The magnitude of the forcing at any location and time depends on the amount of aerosols, their optical properties, underlying surface albedo and the solar zenith angle. The direct solar visible (<780 nm) flux values, measured using calibrated pyrheliometer, are compared with the independently measured instantaneous aerosol optical depth normalized for the solar zenith angle variation. The data are treated separately for the coastal region, open Arabian Sea region and the pristine Indian Ocean region south of the ITCZ.

Bar chart of aerosol forcing over three regions.

Figure 3: Aerosol forcings estimated over the three regions by multiplying the forcing efficiency with the average aerosol optical depth for the three regions.

We estimate the diurnally averaged net surface forcing efficiency for the three regions as, 62.7, 70.7 and 67.5 -W/m2, and the net surface forcing (after multiplying by the average aerosol optical depth for that region) as 27, -12.7 and -2.0 W/m2 respectively, with an absolute uncertainty of about 0.5 W/m2 (Figure 3). The minus sign indicates a reduction in the flux value with increase in the aerosol optical depth. From the estimated surface radiative forcing, the top of the atmosphere (TOA) radiative forcing efficiency (∆F(up)/∆γ) is derived from the scattering efficiency of the atmosphere and the upscatter fraction (fraction of the scattered solar radiation that escapes to space). The downward scattering efficiency, e is defined (8) as the ratio of the increase in the diffuse sky radiation to the decrease in the direct solar radiation. The obtained e values are 0.38, 0.47 and 0.63 respectively for the coastal region, the open Arabian Sea and the pristine ocean region south of the ITCZ. Radiative transfer models (9) can be used to calculate the upscatter fraction ß' for different aerosol size and sun position. In general a value of 0.3 is found appropriate for ß' for realistic aerosol atmosphere. Using the scattering efficiency values and the upscatter fraction, the radiative forcing at the top of the atmosphere due to aerosols is estimated as 6.9, 4.8 and 1.2 -W/m2 respectively for the coastal region, the open Arabian Sea and the pristine ocean region south of the ITCZ. The difference between the forcing at the surface and the forcing at the TOA yields the net solar radiation absorption within the atmosphere and the values for the three regions are, 20.1, 7.9 and 0.8 -W/m2. The atmospheric solar heating by about 20 W/m2 caused by aerosol absorption over the west coast of India is about a factor of eight greater than the enhanced greenhouse warming caused by all the known greenhouse gases in the visible region of the solar spectrum. In summary it can be said that the surface forcing is about three times more than the TOA forcing in the polluted regions while the difference between the surface and TOA forcing is much less for the pristine region. The values of the single scattering albedo, ω of the aerosols found near the coastal region are very low, between 0.8 and 0.9 compared to the pristine ocean value of more than 0.95. Another study (3) shows that the column average of ω lies generally in the range between 0.86 and 0.89 for the polluted regions. These ω values represent a highly absorbing aerosol mixture.

More or less similar aerosol characteristics over a very wide area indicate that the NE monsoon flow is responsible for the dispersion and mixing of continental aerosols to a very large region over the ocean surface. It should be noted that the above estimations are for clear sky conditions as the radiation and aerosol optical depth data are screened for cloud contamination. In the presence of clouds, it is not sufficient to consider only the direct radiative effects of aerosols but also the indirect effect caused due to the cloud reflection. For example computations (3) show that during overcast sky condition the TOA forcing nearly doubles if the aerosol layer is above the clouds rather than below, because the elevated aerosol layer absorbs more solar radiation reflected by the clouds.

Estimated Regional Effects. A preliminary assessment of the observed aerosol forcing using Community Climate Model (10) shows (3) a low level air temperature increase by about 0.5 to 1.5 K over the region where the aerosol forcing was imposed. The warming extends beyond the high forcing region from the southern Indian Ocean region to East Asia. This low level heating resulted in an enhanced convection and strengthening of rainfall along the ITCZ by as much as 15 to 30% and a subsequent reduction in the north (Arabian Sea) and the south (3). A decrease in low level wind decreases the surface evaporation over the northern Indian Ocean by about 10 to 20 W/m2. In addition the aerosols are also found to perturb the inter-hemispheric heating gradients. The gradient that is one of the driving forces of the thermohaline circulation is found to increase by as about 50%. Another important impact could be on the ecosystem as about 70% of the forcing is concentrated in the photosynthetic part of the solar spectrum. Also, burning of low clouds by soot absorption1 (1,12) and a suppression in precipitation could significantly reduce the wet removal of aerosols resulting in a positive feedback of enhanced aerosols lifetime and increase in their amount.

Acknowledgements. The results presented in this study have come from the internationally coordinated Indian Ocean Experiment Program funded in India by the Department of Space, Council of Scientific and Industrial Research, Department of Ocean Development, Department of Science and Technology and India Meteorological Department. The author would like to thank Prof. V. Ramanathan, SIO, UCSD, La Jolla, USA for his guidance and for providing the various aerosol and radiation measuring instruments deployed onboard ORV Sagar Kanya.


  • 1. Ramanathan, V. et al., Indian Ocean Experiment: A multi-agency proposal for field experiment in the Indian Ocean, C4 pub. 162, 1996, Scripps Institution of Oceanography, La Jolla, Calif., pp83.
  • 2. Kiehl J.T., B.P. Briegleb, Science, 1993, 260, 311-314.
  • 3. Ramanathan, V. et al., J. Geophys. Res., 106, 28371-28398, 2001.
  • 4. Lelieveld J., et al., Science, 291, 1031-1036, 2001.
  • 5. Jayaraman, A., S.K. Satheesh, A. P. Mitra and V. Ramanathan, Current Science, 80, 128-137, 2001.
  • 6. Jaenicke, R., in Aerosols and cloud climate interactions, P.V. Hobbs (Ed.) Academic Press, New York, 1993, 1-31.
  • 7. Penner J. et al., Chap V, Climate Change 2001, Report of the IPCC WG1, Cambridge Univ. Press, pp 289-348.
  • 8. Jayaraman A., D. Lubin, S. Ramachandran, V. Ramanathan, E. Woodbridge, W.D. Collins and K.S. Zalpuri, J. Geophys. Res., 1998, 103, 13827-13836.
  • 9. Wiscomb, W., Grams, G., J. Atmos. Sci., 1976, 33, 2440-2445.
  • 10. Kiehl J.T. et al., J. Clim., 11, 1131-1149, 1998.
  • 11. Hansen J., M. Sato and R. Ruedy, J. Geophys. Res., 102, 6831-6834, 1997.
  • 12. Ackermann A.S. et al., Science 288, 1042-1047, 2000.

Workshop Homepage * Background
Summaries: Overview, Gases, Aerosols, Tech., Health, Agri./Eco.
Abstracts: Day 1, Day 2, Day 3, Day 4, Day 5 * Participants