Characterization and Parameterization of CCN Activity

The CCN activity of an aerosol particle depends on its size and chemical composition. According to Raoult's law, the ability of the water-soluble fraction to absorb water and reduce its activity is determined primarily by the number of ions or molecules that can go into a solution per unit volume of the particulate material (Raoult efficiency of the solute).

For particles with a fully known composition, the effective number of solute molecules or ions per unit volume can be derived from the basic physico-chemical properties of the components. The relevant properties, such as dissociation constants and activity coefficients, are, however, not well known for many atmospherically relevant substances (e.g., as secondary organic aerosol components). Ambient particles contain a vast number of compounds many of which are unknown, not individually measurable, and without available thermodynamic information.

Consequently, empirical techniques have been developed and used to characterize the hygroscopicity of ambient aerosol particles. They can be broadly classified in two categories:

• Techniques that measure the water uptake and hygroscopic growth of aerosol particles in the subsaturated regime (RH < 100%), which can be extrapolated to CCN activation under supersaturated conditions.

• Techniques that directly measure the activation of CCN and formation of cloud droplets in the supersaturated regime (RH > 100 %).

Most information can be obtained with online and size-resolved techniques, and the most widely used instruments are hygroscopicty tandem differential mobility analyzers (HTDMA) for subsaturated conditions and various types of CCN counters (mostly thermal gradient diffusion chambers and flow tubes) for supersaturated conditions. HTDMAs are usually more robust and simple to construct and operate than CCN counters, but obviously CCN counters provide more direct insight into CCN activation. The reliability of CCN measurement results has been a subject of continuing debate. Recently, CCN counters with high and well-documented measurement precision and accuracy have become commercially available and are now widely used in laboratory and field studies (McFiggans et al. 2006; Rose et al. 2008a,b).

Measurement results on the hygroscopic growth and CCN activation of atmospheric aerosol particles have often been reported in terms of an (equivalent) soluble fraction, which can be defined as the volume fraction of a model salt (e.g., (NH4)2SO4 or NaCl) in a dry particle consisting of the model salt and an insoluble core, such that the model particle would exhibit the same hygroscopicity and CCN activity as the actual particle (Andreae and Rosenfeld 2008a,b).

Petters and Kreidenweis (2007) have proposed an effective hygroscopicity parameter, k, which relates the volume of water taken up by a particle to the activity of water in the formed aqueous droplet. Measured or calculated values of k and similar parameters (e.g., "ion densities" as defined and used by Rissler et al. 2006 and Wex et al. 2007) enable efficient calculation of critical supersaturations or critical dry particle diameters according to the Köhler theory (effective hygroscopicity parameter models; Rose et al. 2008a).

A simple analytical approximation model enables the calculation of the critical supersaturation S as a function of k and of the dry particle mass or volume equivalent diameter Ds, respectively (Rose et al. 2008a)

Here the Kelvin effect is described by the parameter A as a function of the surface tension of the aqueous solution droplet (aso), the absolute temperature (T), the ideal gas constant (R), and the density and molar mass of water (p , M ):

Under the assumption that the surface tension of the aqueous solution droplets formed upon CCN activation equals the surface tension of water, the Kelvin parameter can be approximated by A ~ (0.66 x 10-6 K m)/T .

Under the assumption of linear additivity of the influence of individual chemical components on the activity of water in a multicomponent aqueous solution (Zdanovski-Stokes-Robinson or ZSR assumption), the effective hygro-scopicity parameter of an aerosol particle (k) can be calculated by linear combination of the effective hygroscopicity parameters of the individual chemical components (k) weighted by their volume fraction (e.) in the dry particle:

Insoluble components can be described by k = 0, and limited solubility can be taken into account as described by Petters and Kreidenweis (2008) and Kreidenweis et al. (this volume). For fully soluble components, k is directly related to the molar mass (M), density (pj), and van't Hoff factor (ij) or stoichiometric dissociation number and osmotic coefficient (v. ) of the dry component and of water (subscript w), respectively (Rose et al. 2008):

Accordingly, k and k can be regarded as an "effective Raoult parameter"; that is, as an effective or equivalent molar density of soluble ions or molecules in the dry particle or dry particle component respectively, normalized by the molar density of water molecules in liquid water (~55 mol l-1). For compounds with low molecular mass (including most inorganic salts like sulfates, nitrates, chlorides as well as mono- or dicarboxylic acids, monosaccharides, etc.), the effective molar density of ions or molecules is usually close to the actual molar density of ions/molecules (i.e., the osmotic coefficients are close to unity). For macromolecular organic compounds like proteins, however, the osmotic coefficient can increase up to >100, and the effective Raoult parameter can only be regarded as an equivalent parameter (Mikhailov et al. 2004). The assumption of a constant value of k for a soluble particle component is essentially equivalent to the assumption of a concentration-independent van't Hoff factor or osmotic coefficient, respectively.

Note, however, that Petters and Kreidenweis (2007) have suggested folding surface tension effects into the hygroscopicity parameter K as well; that is, to assume that the surface tension of aqueous solution droplets formed upon CCN activation equals the surface tension of water (a , = a = 0.072 J m 2 at

298K) both upon deriving k values from experimental data as well as upon applying them for model predictions. To avoid confusion, we suggest that a

fw rw w

distinction be made between "effective Raoult parameters," as defined in the above equations and in the basic equations of Petters and Kreidenweis (2007), and "effective hygroscopicity parameters" that have been derived and should be applied under the assumption of asol = aw. For most practical applications, this discrimination will be of minor importance, and in many cases it may also not be possible to determine independently effective Raoult parameters (water activity reduction) and Kelvin parameters (surface tension reduction) for atmospheric aerosol particles. Neverthless, establishing unambiguous terms and definitions should be helpful for consistent and efficient communication between different scientific communities (aerosol and cloud; field, laboratory, and model).

Characteristic values of k are in the range of ~1.3 for NaCl, ~0.6 for (NH4)2SO4, ~0.2 for levoglucosan and various organic acids, and ~0.1 for secondary organic aerosols. For biomass-burning particles, k values range from about 0.01 for very fresh smoke containing mostly soot particles to 0.55 for aerosols from grass burning. The available data suggest that after short aging on a timescale of hours, most pyrogenic aerosols will have k values in the range of 0.1 to 0.3 (Petters and Kreidenweis 2007; Andreae and Rosenfeld 2008; Rose et al. 2008b; Kreidenweis et al., this volume). Externally mixed aerosol populations can be described by separate k values, and the effects of internal mixing (coagulation/condensation) can be described according to Equation 3.5. Based on aerosol and CCN field measurements and laboratory experiments, it will be possible to establish an inventory of effective hygroscopicity and CCN activity of particles from various sources and to describe efficiently the influence of atmospheric aging.

A summary of k values derived from various sources is presented in Figure 3.13. The data plotted in this figure show that continental aerosols fall in a narrow range of k values around 0.3, consistent with the suggestion by Dusek et al. (2006) that, for such aerosols, the composition can be treated to a good approximation as invariant, and that the CCN activity of particles is controlled mostly by particle size. In that study, the observed range of k was 0.15-0.30. Although Hudson (2007) correctly points out that field data cover a larger range of k values, his polluted continental data also give an average of k = 0.33±0.15. As expected, his clean marine data indicate higher values, with k = 0.87±0.24 (colored bands in Figure 3.13). Measurements at the coast of Puerto Rico also showed CCN activation diameters corresponding to k values in the range 0.6±0.2 (Allan et al. 2007). A large number of field data has been compiled by Kandler and Schütz (2007) and expressed in the form of soluble fractions. When converted to k values, these data are also consistent with urban and continental values of k around 0.2-0.3 and marine values around 0.6. Recent measurements of CCN, as a function of particle size and water-vapor supersaturation at background and polluted continental locations in Europe, Asia, and America, confirm fairly uniform average hygroscopic properties and CCN activities of continental aerosols (k = 0.3±0.1). Higher variability was

s ro

Figure 3.13 Average relations between critical supersaturation and aerosol dry diameter (Andreae and Rosenfeld 2008). The colored bands reflect polluted continental and clean marine data from Hudson (2007); the colored dots with colored borders are from Dusek et al. (2006) and Andreae and Rosenfeld (2008). The colored dots with gray borders have been recalculated from Kandler und Schütz (2007). The lines representing constant effective hygroscopicity parameters k are from Petters and Kreidenweis (2007).

observed only at low supersaturation (< 0.1%) and for freshly emitted or newly formed aerosols dominated by organic components (Rose et al. 2007, 2008b).

Overall, k values of 0.3±0.1 and 0.7±0.2 appear to be representative for average continental and marine aerosols, respectively. The assumption of fairly uniform effective hygroscopicity parameters is also supported by studies indicating that the influence of aerosol chemical composition is limited by the dynamics of cloud droplet growth. Moreover, the influence of the composition-dependent hygroscopicity parameter is, according to Köhler theory, weaker by the power of 3 than that of particle diameter (see Equation 3.3: k-1/2 vs. D-m). Accordingly, Dusek et al. (2006) and Andreae and Rosenfeld (2008) have argued and presented evidence that particle size is the dominant property in controlling the variability in CCN activity of atmospheric aerosols. Moreover, Andreae (2008) found a surprisingly close correlation between the average CCN concentrations observed in field measurements around the globe and the corresponding aerosol optical thickness (AOT) data from collocated or nearby remote sensing instruments (mostly AERONET sun photometers). Aerosol particle number concentrations and size distributions obtained by remote sensing and their implications for the abundance of CCN on global and regional scales have also been addressed by Kinne (see Part 1).

Hygroscopicity (k)

Hygroscopicity (k)

mtg r ÖOO\ ; \ •

A

- \pc

O Marine

Continental/urban

Mod. aged biomass

Very fresh biomass

.., .s mW.

Nevertheless, more detailed investigations and representations of aerosol particle hygroscopicity and composition are desirable for a full elucidation of aerosol-cloud interactions, especially for low water-vapor supersaturations, low aerosol concentrations, and organic components (McFiggans et al. 2006; Rose et al. 2007, 2008b; Kreidenweis et al., this volume).

0 0

Post a comment