BelAQI index

What is an air quality index?

To assess the air quality on a short time scale (daily/hourly), an air quality index is used. This index combines the concentration of various ambient air pollutants into one representative number. As such, technical measurement results are translated into an easily understandable assessment of the air quality.

BELAQI-index

In the context of the European "JOAQUIN"-project the Joaquin-index was developed, which has been used since one year in the Netherlands (see https://www.luchtmeetnet.nl/). The BELAQI-index is a version of the Joaquin-index adapted to the Belgian situation. These "100%" health-based indexes make use of "dose-response" relationships. As starting point for the Joaquin-index, concentration scales are determined based on the short term health impact of PM_2.5. This is particulate matter with an aerodynamic diameter less than 2.5 µm and is generally considered as the air pollutant with the biggest health impact.

The concentration scales of the other pollutants (nitrogen dioxide or NO₂, ozone or O₃, particulate matter with a diameter less than 10 µm or PM₁₀) are determined based on comparable health impact. To do so, the “Relative Risk” (RR) as formulated in the Health Risks of Air Pollution In Europe (HRAPIE) report of the World Health Organisation (WHO) are used. A RR represents the increase in health impact caused by a concentration increase of 10 µg/m³.

The increase of air pollution affects the total daily mortality but also causes an increase in, for example, hospitalisation cases related to heart-, vascular and lung diseases. Nevertheless, to determine the impact of air pollution the RR’s related to an increase in total daily mortality are used. This is because no data for hospitalisation increase by NO₂ pollution are available. As such, to include the RR’s related to hospitalisation additional assumptions are required, which would complicate the impact assessment.

The following RR’s (per 10 µg/m³ increase) from the HRAPIE report are used for increase in total daily mortality:

This means that for each 10 µg/m³ rise in PM_2.5 concentration the daily mortality undergoes a statistically significant increase of 1.23 % (likewise, for each 10 µg/m³ rise in the maximal 1-hour mean NO₂ concentration the mortality increases by 0.27 %). These RR’s for total mortality are determined based on epidemiological research that compares the daily variation in air pollution with the daily mortality rate.

The starting point of the BELAQI-index scale is PM_2.5 and the same methodology is used as was the case with the Joaquin-index. Based on the guideline values and interim targets of the WHO for PM_2.5, four “health”-concentration classes are made:

- Class A with daily mean PM_2.5 concentrations smaller than the yearly mean WHO guideline value: - Class B with daily mean PM_2.5 concentrations between the yearly and daily mean WHO guideline values: > 10 µg/m³ - - Class C with daily mean PM_2.5 concentrations between the daily mean WHO guideline value and the WHO interim target II: > 25 µg/m³ - - Class D with daily mean PM_2.5 concentrations larger than the WHO interim target II: > 50 µg/m³.

In order to work with 10 index classes these four classes are further subdivided, as shown for the BELAQI-index in Figure 1:

Figure 1: PM_2.5 classification according to the WHO values (blue: class A, green: class B, orange: class C, red: class D).

The classification (Figure 1) is slightly different from the classification used to determine the Joaquin-index.

Table 1 presents the index-concentration scale and the corresponding increase in daily mortality based on the RR for PM_2.5:

Table 1: BelAQI index scale, associated daily mean PM_2.5 concentrations and corresponding increase in daily mortality.

Using the RR’s for NO₂ and O₃ from the HRAPIE report, the concentration scales that cause an equal increase in mortality are calculated, resulting in the index-concentration scales presented in Table 2:

Table 2: BelAQI index scale and associated maximal 8-hour O₃ and 1-hour NO₂ mean concentrations based on an equal increase in mortality compared to the daily mean PM_2.5 concentrations.

With respect to O₃, the relation between concentration and health effects is determined based on the “maximal 8-hour mean” concentration of a given day. As European information and alert thresholds are formulated on a 1-hourly scale, it was decided to use hourly mean concentrations for the O₃ sub-index scale. This is scientifically sound due to the strong correlation between the maximal 8-hour and the maximal of a given day (R² =0,97 for the daily maxima of a “virtual” or “mean” Belgian O₃ station during the period 2013 – 2015). A maximal 1-hour mean O₃ concentration corresponds to 1.11 times the maximal 8-hour mean O₃ concentration (1.11 being the median).

With respect to PM₁₀, no RR for daily mortality is defined in the HRAPIE report. However, it was decided to also develop a PM₁₀ sub-index scale, as the information threshold for particulate matter (50 µg/m³, implemented the 1^st of November 2016) as well as the smog alert threshold (70 µg/m³, in use since 2006) are based on PM₁₀ concentrations. To determine a (sub-)index scale for PM₁₀, the PM_2.5/PM₁₀-ratio is used. This ratio amounts to 0.63 and is the median of the daily mean PM_2.5/PM₁₀-concentrations of a “virtual” or “mean” Belgian O₃ station during the period 2013 – 2015. Using this ratio, the PM₁₀ concentration scales can be determined based on the already existing PM_2.5 scales.

In Table 3, the concentration scales for PM₁₀ (calculated via the PM_2.5/PM₁₀-ratio), PM_2.5, O₃ (recalculated from maximal 8-hour to maximal 1-hour mean concentrations) and NO₂ are given.

Table 3: BelAQI index scale with PM_2.5 daily mean concentrations and corresponding PM₁₀ daily mean (based on the PM_2.5/PM₁₀-ratio), O₃ maximal 1-hour mean (recalculated from maximal 8-hour mean) and NO₂ maximal 1-hour mean concentrations (based on equal increase in mortality compared to PM_2.5 daily mean concentrations).

In order to develop a health impact classification, the concentration scales (Table 3) are rounded while taking into account the NO₂ threshold of 200 µg/m³, the NO₂ phase 1 threshold of 150 µg/m³ in Brussels, the European O₃ information and alert thresholds of 180 and 240 µg/m³ respectively, the PM₁₀ information threshold of 50 µg/m³ in Belgium, the smog alert threshold of 70 µg/m³ and the phase II threshold of 100 µg/m³ of the regional smog action plan of Brussels. The result is shown in Table 4.

Table 4: BelAQI index scale, health impact classification and associated concentration scales for PM₁₀ daily mean, PM_2.5 daily mean and O₃ and NO₂ daily maximal 1-hour mean.

Remark: to calculate the actual (hour per hour varying) sub-indexes and the global index, the concentration scales of Table 4 are applied to the latest hourly mean O₃ and NO₂ concentrations and the running 24-hourly mean PM_2.5 and PM₁₀ concentrations.

Global air quality index and classification

To determine the global air quality index, the sub-index for each pollutant is calculated based on the index scale in Table 4. The highest sub-index defines the global air quality index.

10 qualifications are used, going from “excellent” to “horrible”. The classification of the index is established in function of the severity of the health impact (mortality risk). Meanwhile, the information and alert thresholds for PM₁₀ and O₃ are taken into account. If the particulate matter (PM₁₀) concentrations exceed the smog information threshold, the air quality is qualified as “poor” (6). In the case of an exceedance of the PM₁₀ alert threshold, the index level amounts to eight and the air quality is described as “bad”. With respect to O₃, the (sub-)index in case of European information threshold exceedance is one class higher compared to PM₁₀, resulting in the air quality label “very poor” (7). If the O₃ alert threshold is exceeded, the air quality is qualified as “bad” (8).

The BELAQI-index gives no indication of the long term air quality. The index is meant to assess short term effects of the daily variation of air pollution. It is also important to note that an index evaluates the air quality in a qualitative way (and thus not quantitatively). Consequently, it is not recommended to use trends or statistics of air quality based on an air quality index. This is also the reason that the calculated indexes are not used in the yearly air quality reports.