4.3 Results and discussion
4.3.4 Segregation effects
Table 4.1 lists intensities of segregation (in percentage) between selected pairs of chemical species averaged across the canyon and over the period of 180 to 240 min. It is interesting to note that intensities of segregation between A and B (where A=B) (as shown along the diagonal line in Table 4.1) are positive, with the largest value of 28.49 % between NO and NO, and the smallest value of 0.36 % between OH and OH. This is attributed to the fact
110
that the auto-covariance of any chemical species is always positive if the chemical species is not homogenously distributed within the canyon. Intensities of segregation between A and B (where A=B) may reflect the spatial variability of the chemical species within the canyon relative to its mean concentration.
It is found that there are positive values for intensities of segregation between NO, NO2
and VOCs, indicating that ‘emitted chemical species’ have similar correlations and are driven by the dynamical processes acting upon emissions. The highest value is found to be 22.32 %, which is the intensity of segregation between NO and VOCs. These emitted chemical species are carried by the canyon vortices and removed from the canyon roof level to the background atmosphere. Positive values of intensities of segregation between O3, OH and HO2 are also clearly observed, but the magnitudes are lower around 3% (e.g.
2.87 % for the intensity of segregation between O3 and HO2). This can be explained by considering that O3, OH and HO2 are ‘entrained chemical species’ with higher levels in the background environment than those inside the street canyon and thereby exhibiting similar behaviour. This indicates that segregation effect would enhance the rate of a reaction between pairs of species with similar origins (either ‘emitted chemical species’ or
‘entrained chemical species’).
It is also noted that negative values are found for intensities of segregation between emitted chemical species (i.e. NO, NO2 and VOCs) and entrained chemical species (O3, OH and HO2). This is attributed to the opposite origin of those chemical species, i.e. one is emitted from the street canyon while the other is entrained from the background environment. Negative correlations between those species are therefore expected, giving the negative values for intensities of segregation between them. As shown in Table 4.1, these pairs of both emitted and entrained chemical species generally undergo the chemical
111
reactions within the canyon. The average chemical reaction rates across the canyon domain are expected to be reduced due to the incomplete mixing in such an environment, which plays a key role in determining the net chemical processing in the street canyon.
Segregation effects are relatively larger between O3 and emitted species (i.e. -11.09 % for NO, -5.10 % for NO2, -8.91 % for VOCs) than those between OH (or HO2) and emitted species. It is expected that the NO and O3 titration to generate NO2 within the street canyon is reduced by 11.09 % due to the segregation effect compared with a well-mixed system. It is also noted that intensity of segregation between VOCs and OH is -2.37 %, indicating that the canyon-averaged reaction rate between VOCs and OH will be retarded by -2.37 % due to incomplete mixing, thereby leading to a reduction in the additional conversion rate of NO to NO2 by the VOCs oxidation chemistry. This negative intensity of segregation between VOCs and OH (about -3.4 %, slightly higher than -2.37% in this study) was also found by (Krol et al., 2000) in which a LES model in a convective atmospheric boundary layer was conducted. Auger and Legras (2007) suggested that due to the nonlinear nature of chemical processes, even a small value for intensity of segregation (e.g. 1 %) may lead to significant effects on the mean concentrations, especially while the pollutant residence time is short. This further indicates that segregation effects are very important and should be highlighted for any incomplete mixing environment (e.g. the street canyon), in which the interactions between the dynamics and nonlinear chemistry take place.
112
Table 4.1 Intensities of segregation (in percentage) between pairs of chemical species averaged among the canyon and over the period of 180 to 240 min. Values shown in parentheses and bold denote those pairs of chemical species that react directly with each other.
O3 NO NO2 VOCs HO2 OH
O3 6.34
NO (-11.09) (28.49)
NO2 (-5.10) 11.18 4.73
VOCs (-8.91) 22.32 8.86 17.51
HO2 (2.87) (-5.67) (-2.44) (-4.51) (1.39)
OH (1.25) (-3.03) (-1.17) (-2.37) (0.66) 0.36
Figure 4.9 depicts spatial variations of intensities of segregation (in percentage) between (a) O3 and NO, and (b) OH and VOCs within the street canyon (z/W≤2). It is found that there are very large negative segregation effects close to the emission source towards the windward wall at the street level. The highest negative values of intensities of segregation could be about -90 % between O3 and NO and about -20 % between OH and VOCs. This can be explained by the large spatial variability in these regions which are directly determined by emitted species. These large negative values indicate that the associated chemical reactions near emissions are significantly reduced due to the non-uniform emissions. Large negative segregation effects were also observed at the canyon roof level towards the windward wall, i.e. about -60 % between O3 and NO and about -8 % between OH and VOCs. This is attributed to that these regions are places where the background atmosphere (e.g. O3 and OH) is entrained into the street canyon and then interact with the emitted species from the street canyon. The large spatial variability in these species is also expected. Towards the leeward wall near the canyon roof level, there is a rapid decrease in
113
the intensities of segregation for both pair of chemical species, indicating that there are much greater mixing for emitted species in these regions. It is also noted that intensities of segregation are separated by the two vortices formed in the street canyon and then increase both upwards to the canyon roof level and downwards to the street ground.
(a) Is (O3+NO) (b) Is (OH+VOCs)
Figure 4.9 Spatial variations of intensities of segregation (in percentage) between (a) O3 and NO; (b) OH and VOCs.