Air Pollution as a Climate Forcing: A Workshop

Day 2 Presentations

How Have Accumulation Mode Aerosol Number Concentrations Changed Since Pre-industrial Times?

Julian Wilson, Frank Raes and Frank Dentener
Institute for Environment and Sustainability, DG Joint Research Centre, European Commission, Ispra (Va), Italy

You may download a MS PowerPoint version (3.0 MB) of this presentation.


The large uncertainties in current estimates of the radiative effects of aerosols reflect, among other factors, the use of empirical relationships to derive the radiative effects from the calculated mass concentrations of the aerosol species considered. Global models, which resolve the aerosol number size distribution and the state of mixing of the various aerosol chemical species, allow these effects to be calculated from first principles, i.e. Mie theory for scattering and absorption of light, and Kohler theory for the activation of aerosol particles into cloud droplets.

We have developed such a model (Wilson et al. 2001). It treats the primary emissions of sea-salt particles and black carbon (BC) particles from biomass and fossil fuel burning. It furthermore treats the emissions of biogenic DMS, SO2 from volcanoes and SO2 from fossil fuel use and industrial processes, and the transformation of these gases into sulfate. Organic carbon (OC) emissions are scaled to BC emissions in the case of biomass burning, and to sulfate production in the case of fossil fuel burning. Condensation, coagulation and cloud processing, allow the various chemical species to mix internally. The aerosol size distribution is described by several lognormal modes, spanning the size range from a few nanometers to a few micrometers. The model is consistent with the available observations. It produces, within a factor of 2, both observed zonal average marine aerosol number concentrations and observed sulfate mass/accumulation mode number concentration ratios from the N. Atlantic. It does less well at reproducing number concentrations at individual sites, and consistently over-predicts the number concentration of particles with diameter < 80 nm. (See Wilson et al. 2001, for details)

In the present study we focus on accumulation mode particles (AMP), which have a diameter between about 80 and 800 nm, and which interact most effectively with radiation and clouds. In particular, we look at the effect of the increases in SO2 and BC emissions between pre-industrial times and the present on the number concentration of AMPs. Pre-industrial times are simulated by setting all fossil fuel SO2 and BC emissions to zero, and by halving the present day BC emissions from biomass burning. Figure 1 shows the percentage increases for January and July

Maps of January/July AMP concentrations. See caption and text for detai ls.

Figure 1: Percentage change of January and July monthly average accumulation mode particle (AMP) concentrations (=sulfate accumulation mode + mixed BCf/OCf/sulfate + mixed BCb/OCb/sulfate + sea salt) at the surface predicted by M3+TM2.

While there is an increase in total annual average AMP number burden of 280%, from pre-industrial times, this increase is not distributed uniformly. The industrial increase in AMPs is seen primarily over the continents. Over the oceans, the increase is significant (i.e. > 25%) in areas of continental outflow. There are also areas where the present day number concentration of AMPs is significantly lower (i.e. < -25%) than before.

To understand these results it is important to realise that by increasing SO2 and BC emissions, AMP concentrations can be effected in several ways.

  1. Increasing SO2 will lead to a larger production of sulfate in the gas phase, which may enhance nucleation and subsequent growth of pure sulfate secondary AMP.
  2. Simultaneous emissions of BC however provide more aerosol surface on which sulfate will condense rather than nucleate and form new particles. When this surface is sufficiently large, nucleation and growth of secondary AMPs can be completely suppressed, and AMPs are formed by condensation of sulfate and OC on the primary BC particles. Hence, the change in the AMP number concentration is controlled either
    • a. by the growth rate of primary BC particles into AMP, where this is limited by the availability of sulfate and OC in the gas phase, or
    • b. by the change in emissions of primary BC particles where there is sufficient sulfate and OC for all BC particles to become AMPs.

In order to see which of these processes controls the increase in AMPs due to simultaneous SO2 and BC emissions we have calculated the following ratio:

R = (∆NAMP - ∆NBC) / NAMP,pre-industrial

which compares the increase in the AMP number concentration (∆NAMP) with the increase in mixed sulfate/OC/BC number concentration (∆NBC) (all primary BC particles, that are partly coated with the OC/sulfate)

  • The more positive R, the more process 1. prevails, and the more the increase in AMP is due to enhanced nucleation and growth of new secondary particles.
  • The more negative R, the more process 2.a. prevails, and the change in AMPs is controlled by the condensation of sulfate and OC on primary BC particles
  • R will be around zero in areas where process 2.b. prevails: all primary BC particles rapidly grow to AMPs, and the increase in the latter is controlled by the emissions of primary BC. (R will also be zero in remote areas where there is little change in either AMPs or BC particles).
Maps of January/July AMP controls. See caption and text for details.

Figure 2: Indication of areas where the January and July increases in AMPs between pre-industrial times and present are controlled by enhanced nucleation of sulfate particles (yellow areas) or are controlled by condensation of sulfate on newly emitted BC particles (dark blue areas). See text for details.

Values of R are plotted in Figure 2 for January and July. The large increases in AMPs over central Europe are explained by the increase in BC particles, however these particles are limited in their growth and do not all become AMPs. The increases over Eastern Siberia in January, and over Central Siberia in July are due to enhanced nucleation and further growth of sulfate particles. The increases of AMPs in the biomass burning plumes are controlled by the emissions of primary BC. (The model, in fact, applies faster growth rates to BC particles from biomass burning, than to BC particles from fossil fuel burning.) The areas where present day AMP concentrations are lower than in pre-industrial times, are explained by the increase in BC particles being sufficient to quench the pre-industrial nucleation cycle, but the increase in sulfate is not enough to grow these BC particles into AMPs.

Apart from explaining the changes in AMPs, the model also gives insight into the state of mixing of the various chemical species, which is important for calculating their radiative effects.

Maps of high January/July AMP concentrations. See caption and text for details.

Figure 3: Areas where over 75% of the January and July AMP number concentration population is one species.

Figure 3 shows the areas where more than 75% of the present day number concentration of AMPs are either sea-salt AMPs, pure sulfate AMPs, or mixed sulfate/OC/BC AMPs. Seasalt particles dominate over the oceans in the mid-latitudes in winter because of the local high wind-speeds. Pure sulfate particles dominate over the other parts of the remote oceans, where no significant changes in AMPs have occurred. Over the continents, most of the AMPs are mixed sulfate/OC/BC.

>4n class="hed6">Conclusion. The increase in SO2 and BC emissions between pre-industrial times and present has lead to changes in the importance of various AMP formation processes and eventually in changes to the concentrations of AMPs. The present day AMPs comprise more internally mixed particles, i.e. sulfate/OC/BC particles, than in pre-industrial times. The production of such AMPs can be either limited by the availability of sulfate (as over some industrial areas such as Central Europe) or by the emissions of primary BC particles (like in biomass burning plumes). Emissions of BC particles in terms of both number and size, and an understanding of their interaction with other aerosol species (sulfate, organic carbon) are needed to fully quantify the number and state of mixing of AMPs and, hence, their radiative effects.

Finally, the model simulations show that in the anthropogenic source regions the changes in AMP number concentration are driven by a combination of emissions of primary BC aerosols, and by the availability of condensing gases such as sulfate and organic carbon. It is therefore unlikely that empirical relationships between aerosol mass and AMPs number adequately describe the aerosol microphysics needed to calculate the direct and indirect aerosol effects in the past and in the future.


  • J. Wilson, C. Cuvelier and F. Raes, A modeling study of global mixed aerosol fields, J. Geophys. Res. 106, 34,081-34,108, 2001

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