3. Summary of data reported for summer 1998

The threshold for warning of the public (1h> 360 µg/m³) was exceeded at two stations in the Athens conurbation (both on more then one day), and on one station in France (Table 3).

Table 3: Observed exceedance of the threshold for warning of the public (1h concentration > 360 µg/m³) during summer 1998 (April-August).

During the summers of 1996 and 1997 the threshold for warning of the public was also exceeded in the Athens conurbation, although partly on different stations. The Athens episode will be described in more detail in Chapter 4. The French exceedance occurred in an isolated industrial area near Le Havre during a day with only a few other French stations reporting concentrations above 180 µg/m³.

Table 4 presents a general overview of the observed exceedances of the threshold for information of the public during April-August 1998 on a country by country basis. As the number of stations differs widely from country to country, the absolute number of exceedances is less suitable for comparison. As in the annual ozone report [De Leeuw and De Paus, 1998], the concept of 'occurrence of exceedances' is used here. Occurrence of exceedances is defined as the total number of exceedances summed over all stations divided by the number of stations.

Table 4: Summary of exceedances of the threshold for information of the public (1h concentration > 180 µg/m³) during summer 1998 (April-August) on a country by country basis.

^I Number of stations implemented in the framework of the Ozone Directive
^II The number of days on which at least one exceedance was observed
^III Left figure: averaged over all implemented stations, right figure: averaged over all stations which reported at least one exceedance.

Denmark, Finland, Ireland and Sweden did not observe exceedances of the population information threshold this summer. The same countries did not report any exceedances last year. In other countries, the number of days on which at least one exceedance was observed ranged from 7 in Belgium to 74 in Spain. 61 % of all stations reported one or more exceedance. On average 3.9 exceedances occurred this year on stations which recorded at least one exceedance. The average maximum hourly concentration during an exceedance of the threshold this year was 201 µg/m³.

Table 5 summarises the exceedances on a month by month basis. The relative abundance of observed exceedances in May can be attributed mainly to an early stable period of warm summer weather in Western Europe. April, June and July were on average unfavourable for the formation of ozone in Northern and Western Europe. August had the highest number of stations reporting exceedances and the highest occurrence, due to one period of favourable ozone formation conditions in Western Europe.

Table 5: Summary of exceedances of the threshold for information of the public (1h concentration > 180 µg/m³) during summer 1998 (April-August) on a month by month basis.

^I The theoretical maximum is 1066 stations (all stations which are assumed to be operational during summer 1998 and for which data was transmitted).
^II Left figure: averaged over all implemented stations, right figure: averaged over all stations which reported at least one exceedance.

In Figure 1 the number of days per month is presented on which at least at one station in a country an exceedance was recorded. No obvious monthly dependence is apparent. In the Northern and Western countries Austria, Germany, Luxembourg, Belgium and the Netherlands, August was the month with the highest number of days with exceedances observed. In the Mediterranean region most days with exceedances were observed in July, with the exception of Greece (August). In Portugal exceedances were only observed in August.

Figure 2 presents the occurrence of exceedances per country on a month by month basis. Again, there is no obvious dependence on the month.

Figure 1: Number of days on which at least one exceedance of the threshold value for information of the public (1h concentration > 180 µg/m³) was observed per country and per month during summer 1998. Finland, Denmark, Sweden and Ireland did not report exceedances.

Figure 2: Occurrence of exceedances of the threshold value for information of the public (1h concentration > 180 µg/m³) per country on a month by month basis during summer 1998. Finland, Denmark, Sweden and Ireland did not report exceedances.

The average occurrence of exceedances (in days) in each country of the threshold for information of the public by station type (rural, urban and street) is presented in Figure 3. Stations for which the type was not specified are not presented in this figure. The average occurrence rate is expected, according to ozone phenomenology, to decrease in general in the order rural-urban-street. For some countries, this decrease is apparent. In other countries, this relation is not visible or even contradicts.

Figure 3: Average occurrence of exceedances (in days) of the threshold for information of the public (1h concentration > 180 µg/m³) by station type (rural, urban and street) and country during summer 1998. Finland, Denmark, Sweden and Ireland did not report exceedances.

Figure 4 shows the frequency distribution of hourly ozone concentrations in excess of the threshold value using Box-Jenkins plots. For each Member State the Box-Jenkins plot indicates the minimum (here the minimum is 180 µg/m³), the maximum, the 25 percentile and the 75 percentile value of the exceedances. The figure shows that during 25% of all observed exceedances, the maximum hourly concentration recorded was just above the 180 µg/m³ threshold. 75% of all maximum exceedances recorded were below 214 µg/m³.

Figure 4: Frequency distribution of ozone concentrations in excess of the 180 µg/m³ threshold for hourly values (April-August 1998). Frequency distributions are presented as Box-Jenkins plots indicating the minimum, the 25-Percentile, the 75-Percentile and the maximum value.

3.1 Geographical distribution

Maps 2 and 3 show the geographical distribution of the number of days on which the threshold value for information of the public was exceeded for urban¹ and rural stations, respectively. Exceedance data for urban stations are presented as dots. The exceedance data for rural stations are interpolated using simple inverse distance weighting and a tentatively estimated 'radius of representativeness' of 100 km. Note that this radius might be different for the various regions in Europe.

The spatial pattern of exceedances observed this summer on urban stations and on stations of unspecified type is comparable to that observed during the previous summer. The number of exceedances observed in Northern and Western Europe increases from zero in Ireland, Denmark, Sweden and Finland to more than 10 on some stations in Southern Germany. The number of exceedances observed in the UK, Netherlands, Belgium and the Northern half of Germany is slightly lower than last year. No consistent spatial pattern is apparent in the Mediterranean region. Many stations did not report exceedances, while other stations reported more than 10 exceedances.

The spatial pattern of exceedances observed this summer on background stations (interpolated field) is also comparable to the field observed in the previous summer.

Map 2: Number of exceedances of the threshold value for the information of the public (1h > 180 µg/m³) observed at urban/street stations and stations of unspecified type. Summer 1998 (April-August).

Map 3: Number of exceedances of the threshold value for the information of the public (1h > 180 µg/m³) observed at background stations. Summer 1998 data (April-August), interpolated using inverse distance weighting, cut-off distance of 100 km.

3.2 Comparison with earlier years

Exceedances observed during the 1998 summer period were compared to exceedances observed during the same period in 1997, 1996 and 1995.²

Figure 5.a presents the average exceedance duration³ of the population information threshold, Figure 5.b the average occurrence and Figure 5.c presents the average maximum concentration observed during exceedances. The figures for 1997 are different from those presented in the 1997 summer report [Sluyter and van Zantvoort, 1997] because in the present report data from France and Italy could be included.⁴

Figure 5.a: Average duration in hours of exceedances during the summer period (April-August, French and Italian data were not included in 1995 and 1996).

Figure 5.b: Average occurrence (nr) of exceedances at stations which reported at least one exceedance during the summer period (April-August, French and Italian data were not included in 1995 and 1996).

Figure 5.c: Average maximum concentration (µg/m³) observed during exceedances during the summer period (April-August, French and Italian data were not included in 1995 and 1996)

It is difficult to assess a possible trend in the number, duration and severity of exceedances of the threshold for information of the public which in its turn could indicate a possible trend in precursor emissions, because:

• As indicated, French and Italian data were not included in the 1995 and 1996 data analysis. From a comparison for 1997 with the indicators calculated with and without the French data, it is expected that the average duration, occurrence and maximum would increase for 1995 and 1996 if Italian and French data were included.
• High ozone levels (in this case exceedances of the population information threshold) are mainly observed during periods with warm and sunny weather. Especially in western and northern Europe, the year-to-year variations in meteorological conditions favourable for high ozone levels are large. The resulting variations in exceedance statistics can obscure a possible trend due to changes in pre-cursor emissions. It is at the moment not possible to correct for this variability on a country by country basis nor for the complete EU territory;
• Exceedances statistics are available for only four years which is a very short time series to asses a possible trend;
• The number of stations implemented in the framework of the Ozone Directive increased with ca. 20% during the period 1995-1997 and remained at the 1997 level this year. The increased territorial coverage can have implications for the number of exceedances observed. Also, a changing ratio between the number of rural and urban/street stations can have implications for the number of observed exceedances since peak ozone levels will on average be lower in urban areas then in rural areas.

All three indicators are slightly higher than in the 1997 summer period but lower than in 1995 and 1996.

¹Exceedances reported from stations of unspecified type are also plotted in this map.

² 1995, 1996, 1997: Validated exceedance statistics as transmitted by Member States were used for this purpose.

³ Averaged over all stations which reported at least one exceedance.

⁴Data from France and Italy were not included in the calculations for 1995 and 1996; data was missing, incomplete or inconsistent.