## Info

Anthropogenic NMHC emission (%)

Figure 3. Peak O3 concentrations calculated various reductions in NO^ and NMHC emissions southern Taiwan. 100% represents current emissions.

Delta [O3+NO2] = 0.097±0.069*[Initial NOx] + 0.158±0.027*[Initial NMHCs] -

Figure 4. The total oxidant production rate, i.e. A[O3 + NO2 between 10 a.m. and 12 noon (black dots), is plotted in a three-dimensional space as a function of initial concentrations of NO^ and NMHCs. The shaded plane is a linear surface fit to the black dots. The equation listed represents the surface as a function of initial NO^ and NMHCs. R2 is the correlation coefficient of the fit. Lines between the black dots and the plane denote the deviation from the plane.

Figure 4. The total oxidant production rate, i.e. A[O3 + NO2 between 10 a.m. and 12 noon (black dots), is plotted in a three-dimensional space as a function of initial concentrations of NO^ and NMHCs. The shaded plane is a linear surface fit to the black dots. The equation listed represents the surface as a function of initial NO^ and NMHCs. R2 is the correlation coefficient of the fit. Lines between the black dots and the plane denote the deviation from the plane.

Results from the last section can be applied to formulating ozone pollution control strategies in a similar way to EKMA. Figure 4 shows A[O3 + NO2], or the Ox production rate, as a function of initial NOx and initial NMHC concentrations. A[O3 + NO2] is calculated from [O3 + NO2] at 11 a.m.-12 noon minus the values at 10-11 a.m. and 9-10 a.m. As with the earlier calculation, the initial concentration of NOx at each site is derived from the consumed NOx plus the concentration of NOx observed at 11 a.m.-12 noon. Initial NMHCs are calculated similarly. Consumed CO, CH4 and isoprene are also taken into account. In addition effects of NMHCs not resolved in the GC are included by multiplying the 56 NMHCs by a factor of 1.5.

Figure 4 is similar to the concept of EKMA. In fact, it is equivalent to a three-dimensional OBM ozone production rate isopleth diagram as a function of initial concentrations of ozone precursors. Initial concentrations are used because they are directly proportional to the emissions of ozone precursors.

In order to obtain a simple relationship between the Ox production rate and the precursors of O3, we have made a flat surface (linear) fit to data points of RCEC 2003. Multi-variate statistical regression is used to evaluate whether the correlations versus initial NOx and NMHCs have statistical significance as separate entities and uncertainty ranges associated with the slopes versus initial NOx and NMHCs. The correlation coefficient (R2) for the surface fit is 0.62. The slopes of A[O3 + NO2] versus initial NOx and NMHCs are 0.097 ± 0.069 and 0.158 ± 0.027, respectively. The statistical result of the F test shows that the regression surface is statistically significant and the value of the slope is significant against initial NMHCs but not against initial NOx. This lack of significance can weaken the conclusions in the following interpretation of Fig. 4.

The slightly greater slope against NMHCs suggests that NMHCs are slightly more effective than NOx in controlling the production rate of the total oxidant. Moreover, because NO2 is a part of the total oxidant Ox, reducing NOx emissions will reduce the titration of O3 and thus increase O3. For example, a 20 ppbv reduction in the initial concentration of NO would reduce the titration of O3 by 20 ppbv, which could be converted to about 5 ppbv/h of the ozone production rate over the morning hours (~4h). In comparison, the reduction of the Ox production rate for the 20 ppbv reduction in the initial concentration of NO is about 0.11-20 = 2.2 ppbv/h. This means that the production rate of O3 (not Ox) will increase by 2.8 ppbv/h, i.e. the O3 concentration will increase when the NOx initial value is decreased.

The effect of titration on O3 in KaoPing has been substantiated by an independent study (Chou et al., 2006) on the trends of urban O3 levels at three major metropolitan centers in Taiwan in 1993-2003. The study showed that urban O3 levels in Taiwan increased significantly between 1993 and 2003 due to reduced titration by NO when NOx emissions were reduced. Figure 5 illustrates this point for the KaoPing area. One can see that nearly 90% of the increase of O3 in 1993-2003 can be

## Post a comment