5.2 Chemical results

**Figure 13:** Modeled ozone concentrations in the lowest layer of the model: **(a)** on 29 July 1993, 15:00 CEST; **(b)** the daily maximum of 29 July 1993; **(c)** on 30 July, 15:00 CEST.
$\begin{figure} \begin{center} \leavevmode \epsfysize=19cm \epsffile{oz.map.eps} \end{center} \end{figure}$

**Figure 14:** **(a)** Modeled time of the ozone maximum on 29 July (white areas did not show a clear ozone maximum); **(b)** modeled concentration of NO ₂ (ppb) in the lowest layer of the model on 29 July, 8:00 CEST and **(c)** on 29 July, 16:00 CEST.
$\begin{figure} \begin{center} \leavevmode \epsfysize=19cm \epsffile{nox.map.eps} \end{center} \end{figure}$

**Figure 15:** Comparison of modeled ozone concentrations in the lowest model layer and point measurements near the ground.
$\begin{figure} \begin{center} \leavevmode \epsfysize=19cm \epsffile{o3.comp.eps} \end{center} \end{figure}$

**Figure 16:** Comparison of modeled NO ₂ concentrations in the lowest model layer and point measurements near the ground.
$\begin{figure} \begin{center} \leavevmode \epsfysize=19cm \epsffile{no2.comp.eps} \end{center} \end{figure}$

**Figure 17:** Comparison of modeled NO concentrations in the lowest model layer and point measurements near the ground.
$\begin{figure} \begin{center} \leavevmode \epsfysize=19cm \epsffile{no.comp.eps} \end{center} \end{figure}$

**Figure 18:** Comparison of aircraft measurements and model results for ozone and NO ₂ along the flight path in the afternoon of 29 July 1993. The lowest panel in the figure shows the flight path. Gaps in the model results indicate that the aircraft was outside the model domain at that time.
$\begin{figure} \begin{center} \leavevmode \epsfxsize=\textwidth \epsffile{fly.eps} \end{center} \end{figure}$

Figures 13a and 13c show the spatial distribution of ozone concentrations in the lowest layer in the afternoon of 29 and 30 July, respectively. Figure 13b shows the maximum concentration of 29 July. Plumes the cities of Bern, Biel, Zürich, and Luzern/Zug can be clearly identified. Topography has a significant influence on the distribution of ozone. High concentrations are found in deep valleys and over lakes, where the wind speed is generally lower such that the pollutants have more time to react and to build up ozone before they are diluted by transport and turbulence. This effect can be seen especially in the region of Luzern. Figure 15 shows the comparison of measured and modeled ozone values. The ozone concentrations are well predicted during the day. During the night predicted concentrations are often to high. This again, can be explained by the fact that the lowest layer of the model has a depth of 50m, while measurements were made near the ground, typically at 2 m, which is especially important during the night, when vertical mixing is unefficient and strong vertical concentration gradients occur. The slight overestimation of ozone during the afternoon on 30 July may be due to clouds which were not considered by the model.

NO and NO ₂ are mainly observed near the emission sources (Figures 14b and 14c). Their concentrations at the surface are therefore more difficult to predict, since they are often strongly influenced by very local effects which can not be represented in a model with the spatial resolution of Metphomod. Considering this, Figures 16 and 17, show a good agreement between modeled and measured values, especially at sites which are far from major sources.

Figure 14a shows the time when peak ozone concentrations occur. In the main cities, the maximum occurs around noon, in the plumes in downwind direction the maximum occurs later. Some relationship between the distance from the urban center and the time of the maximum appears to exist.

Figure 15 also includes the values of a similar numerical study carried out with CIT (McRae et al.; 1982) by Kübler et al. (1995). A comparison between the two models shows that the present study generally performs better in reproducing the measured values. We think, that it may be due to the more realistic meteorological data produced by Metphomod which show a strong wind gradient near the ground. The wind interpolation method used in CIT which considered ground stations measurements and radio sounding data probably underestimated the prevailing wind speed.

A comparison of ozone and NO ₂ with aircraft data is shown in figure 18. The plume of the city of Zürich is reproduced by the model. Its horizontal width, however, seems to be underestimated, and peak concentrations are overestimated by model. The model also doesn't reflect the clear minimum in the Emmental region (a rural region between Bern and Luzern) observed by the aircraft. Other features are well reproduced: The peak NO ₂ concentrations are predicted in the right place, and the effects of the boundary layer height agree well, as seen in the two vertical soundings around 13:45 and 15:30.

Although some features that appear in the measurements are not correctly predicted by the model, the overall performance of the model is considered to be good.