Hygroscopic aerosols

The swelling of hygroscopic particles with an increase in relative humidity generates haze particles that are significantly larger than their dry particle sizes. Some aerosols have a deliquescence relative humidity, over which they start picking up water spontaneously, while others like H2SO4 are hygroscopic, meaning that their size increases in a continuous manner over a broad range of relative humidity values (Seinfeld and Pandis, 1998, p. 507). Figure 1 shows the increase in size, normalized to its dry size, of two aerosols of different chemical compositions as a function of relative humidity. It is observed that the rate of size increase with increasing relative humidity of various aerosols may be very different for a given range of relative humidity values. In the example shown, size may increase by a factor larger than 2 when the relative humidity is increased toward saturation.

 

Fig. 1. The growth of aerosols of various chemical compositions as a function of relative humidity (adapted from Seinfeld and Pandis, 1998).

Hänel (1976) proposed simple models for the growth of hygroscopic aerosols as a function of relative humidity based on theoretical considerations and observational evidence. He has observed that the increase in particle size, as described by the ratio of the size of the particle over its dry size ( ), is a complicated function of relative humidity (RH), but can also be a function of
 the particle dry size ( ) as shown in Figure 2. From his experiments, Hänel has concluded that below a certain relative humidity the ratio of the particle size over its dry size ( ) is independent of the dry size of the particles. Also, for moderate and large relative humidities, the ratio  is smaller for those particles with smaller sizes in their dry state.

Fig. 2. Growth of urban aerosols of various dry sizes as a function of relative humidity. Based on measurements taken at Mainz, Germany in 1970 (adapted from Hänel (1976)).

 

Hänel (1976) proposed a simple model describing aerosol growth as a function of relative humidity. This model is expressed as:

 

                                                    (4)                                    

 

where RH is the relative humidity, and  and  are the densities of the dry particle and of pure water respectively.  is defined as the linear mass increase coefficient of the particle which depends on the aerosol composition. More sophisticated approaches exist (Feingold and Grund, 1994), but are not used here, since detailed modeling of aerosols is beyond the scope of the present paper. The relationship described by equation (3) is only used as an illustration of the types of relationships that can be used to characterize the aerosol hygroscopic growth. Relatively good fits of observed data have been obtained with this relationship (Hänel, 1976).

Apart from the change in size, hygroscopic aerosols experience a change in their refractive index as RH increases. Generally, as the water uptake by the particles gets more and more important, the real and imaginary parts of their refractive index tend to decrease (Figure 3), as the real part of the refractive index of pure water is lower than the one associated to dry particles and its imaginary part is zero. This would suggest a decrease in aerosol backscattering and absorption as RH increases. But as the scattering is a complex function of both refractive index and particle size, both effects need to be taken into account. In fact, variations in refractive index as RH increases are not large enough to counteract the variation of the particles’s cross-section due to size increase (r2 dependence). So the size dependence dominates, leading to an increase in backscattering as RH increases.

Fig. 3. Variations of the refractive index of urban aerosols as a function of relative humidity. Based on measurements taken at Mainz, Germany in 1970 (adapted from Hänel (1976)).


| Main page | Abstract | Introduction | Lidar remote sensing | Hygroscopic aerosols
| Lidar backscattering analysis I | Lidar backscattering analysis II | Lidar backscattering analysis III | Conclusions | References | Links |