ANNEX V

Table 1 – Uncertainty for measurement and modelling of long-term (annual) mean concentrations Air pollutant | Maximum uncertainty of fixed measurements | Maximum uncertainty of indicative measurements (1) | Maximum ratio of uncertainty of modelling applications and objective estimation over uncertainty of fixed measurements Absolute value | Relative value | Absolute value | Relative value | Maximum ratio PM2,5 | 3,0 μg/m3 | 30 % | 4,0 μg/m3 | 40 % | 1,7 PM10 | 4,0 μg/m3 | 20 % | 6,0 μg/m3 | 30 % | 1,3 SO2 / NO2 / NOx | 6,0 μg/m3 | 30 % | 8,0 μg/m3 | 40 % | 1,4 Benzene | 0,85 μg/m3 | 25 % | 1,2 μg/m3 | 35 % | 1,7 Lead | 0,125 μg/m3 | 25 % | 0,175 μg/m3 | 35 % | 1,7 Arsenic | 2,4 ng/m3 | 40 % | 3,0 ng/m3 | 50 % | 1,1 Cadmium | 2,0 ng/m3 | 40 % | 2,5 ng/m3 | 50 % | 1,1 Nickel | 8,0 ng/m3 | 40 % | 10,0 ng/m3 | 50 % | 1,1 Benzo(a)pyrene | 0,5 ng/m3 | 50 % | 0,6 ng/m3 | 60 % | 1,1 Table 2 – Uncertainty for measurement and modelling of short-term (24-hour, 8-hour and hourly) mean concentrations Air pollutant | Maximum uncertainty of fixed measurements | Maximum uncertainty of indicative measurements (2) | Maximum ratio of uncertainty of modelling applications and objective estimation over uncertainty of fixed measurements Absolute value | Relative value | Absolute value | Relative value | Maximum ratio PM2,5 (24-hour) | 6,3 μg/m3 | 25 % | 8,8 μg/m3 | 35 % | 2,5 PM10 (24-hour) | 11,3 μg/m3 | 25 % | 22,5 μg/m3 | 50 % | 2,2 NO2 (24-hour) | 7,5 μg/m3 | 15 % | 12,5 μg/m3 | 25 % | 3,2 NO2 (hourly) | 30 μg/m3 | 15 % | 50 μg/m3 | 25 % | 3,2 SO2 (24-hour) | 7,5 μg/m3 | 15 % | 12,5 μg/m3 | 25 % | 3,2 SO2 (hourly) | 52,5 μg/m3 | 15 % | 87,5 μg/m3 | 25 % | 3,2 CO (24-hour) | 0,6 mg/m3 | 15 % | 1,0 mg/m3 | 25 % | 3,2 CO (8-hour) | 1,0 mg/m3 | 10 % | 2,0 mg/m3 | 20 % | 4,9 Ozone (8-hour mean) | 18 μg/m3 | 15 % | 30 μg/m3 | 25 % | 2,2 When assessing compliance with the data quality objectives in Tables 1 and 2 of this Point, the uncertainty for measurements (expressed at a 95 % confidence level) of the assessment methods shall be calculated in line with the respective EN standard of each pollutant. For methods where no standard is available, the uncertainty of the assessment method shall be evaluated in accordance with the principles of the Joint Committee for Guides in Metrology (JCGM) 100:2008 ‘Evaluation of measurement data – Guide to the Expression of Uncertainty in Measurement’ and the methodology in Part 5 of ISO 5725:1998. For indicative measurements, in the absence of a relevant EN standard, uncertainty shall be calculated according to the guidance on the demonstration of equivalence referred to in Point B of Annex VI. The percentages for uncertainty in Tables 1 and 2 of this Point apply for all limit values and target values that are calculated by simple averaging of individual measurements such as hourly mean, daily mean or yearly mean values without considering the additional uncertainty for the calculation of the number of exceedances. The uncertainty shall be interpreted as being applicable in the region of the appropriate limit values or target values. The uncertainty calculation does not apply to AOT40 and values that include more than 1 year, more than one sampling point (e.g. AEI) or more than one component. They are also not applicable for alert thresholds, information thresholds and critical levels for the protection of vegetation and natural ecosystems. Before 2030, the relative values for maximum uncertainties in Tables 1 and 2 shall apply for all pollutants, except PM2,5 and NO2/NOx in Table 1, for which the maximum uncertainties of fixed measurements shall be 25 % and 15 %, respectively. From 2030, the uncertainty of measurement data used for ambient air quality assessment shall not exceed the absolute value or the relative value, whichever is higher, expressed in this Point. The maximum uncertainty of modelling applications is set to the uncertainty for fixed measurements multiplied by the applicable maximum ratio. The modelling quality objective (i.e. a modelling quality indicator less or equal to 1) shall be verified at least at 90 % of the available sampling points, over the assessment area and period considered. At a given sampling point, the modelling quality indicator shall be calculated as the ratio of the root mean square error(s) between modelling results and measurements over the square root of the quadratic sum(s) of the modelling application and measurement uncertainties, over an entire assessment period. Note that the sum will reduce to a single value when annual means are considered. All fixed measurements that meet the data quality objectives (i.e. uncertainty of measurement and data coverage of measurement as specified in this Point and Point B, respectively) located in the modelling application assessment area shall be used for the evaluation of uncertainty of the modelling application. Note that the maximum ratio shall be interpreted as being applicable over the entire concentration range. For short-term mean concentrations, the maximum uncertainty of measurement data used to assess the modelling quality objective shall be the absolute uncertainty calculated using the relative value expressed in this Point, above the limit value and shall decrease linearly from the absolute value at the limit value, to a threshold at zero concentration (3). Both the short-term and long-term modelling quality objectives shall be fulfilled. For modelling of annual mean concentrations of benzene, arsenic, cadmium, lead, nickel and benzo(a)pyrene, the maximum uncertainty of measurement data used for assessing the modelling quality objective shall not exceed the relative value expressed in this Point. For modelling of annual mean concentrations of PM10, PM2,5 and nitrogen dioxide, the maximum uncertainty of measurement data used for assessing the modelling quality objective shall not exceed either the absolute value or the relative value expressed in this Point. Where an air quality model is used for assessment, references to descriptions of the modelling application and information on the calculation of the modelling quality objective shall be compiled. The uncertainty of objective estimation shall not exceed the uncertainty for indicative measurements by more than the applicable maximum ratio and shall not exceed 85 %. The uncertainty for objective estimation is defined as the maximum deviation of the measured and calculated concentration levels, over the period considered, by the limit value or target value, without taking into account the timing of the events. ‘Data coverage’ refers to the proportion of the calendar year for which valid measurement data are available, expressed as a percentage. Air pollutant | Minimum data coverage Fixed measurements (4) | Indicative measurements (5) Annual means | 1-hour, 8-hour or 24-hour means | Annual means | 1-hour, 8-hour or 24-hour means SO2, NO2, NOx, CO | 85 % | 85 % | 13 % | 50 % O3 and related NO and NO2 | 85 % | 85 % | 13 % | 50 % PM10, PM2,5 | 85 % | 85 % | 13 % | 50 % Benzene | 85 % | — | 13 % | — Benzo(a)pyrene, polycyclic aromatic hydrocarbons (PAH), total gaseous mercury, particulate and gaseous divalent mercury | 30 % | — | 13 % | — As, Cd, Ni, Pb | 45 % | — | 13 % | — BC, ammonia, UFP, size distribution of UFP | 80 % | — | 13 % | — Nitric acid, levoglucosan, organic carbon (OC), elemental carbon (EC), chemical composition of PM2,5, PM oxidative potential | 45 % | — | 13 % | — Total Deposition | — | — | 30 % | — Fixed measurements of SO2, NO2, CO, O3, PM10, PM2,5 and benzene are to be carried out continuously during the full calendar year. For the other cases, measurements are to be evenly distributed over the calendar year (or over the April to September period for indicative measurements of O3). In order to comply with those requirements and to ensure that any potential losses of data do not skew results, the minimum data coverage requirements shall be met for specific periods (quarter, month, weekday) of the whole year depending on the pollutant and measurement method or measurement frequency. For the assessment of annual mean values via indicative measurements and, via fixed measurements for pollutants with a minimum data coverage below 80 %, Member States may apply random measurements instead of continuous measurements if they can demonstrate that the uncertainty, including the uncertainty due to random sampling, meets the required data quality objectives and minimum data coverage for indicative measurements. Such random sampling shall be evenly distributed over the year in order to avoid skewing of results. The uncertainty due to random sampling may be determined by the procedure laid down in ISO 11222 (2002) ‘Air Quality – Determination of the Uncertainty of the Time Average of Air Quality Measurements’. Normal maintenance of instrumentation shall not take place during pollution peak periods. Minimum 24-hour sampling is required for the measurement of benzo(a)pyrene and other polycyclic aromatic hydrocarbons. Individual samples taken over a period of up to 1 month may be combined and analysed as a composite sample, provided the method ensures that the samples are stable for that period. The three congeners benzo(b)fluoranthene, benzo(j)fluoranthene, benzo(k)fluoranthene can be difficult to resolve analytically. In such cases, they can be reported as a sum together. Sampling shall be spread evenly over the weekdays and the year. For the measurement of deposition rates, monthly or weekly samples throughout the year are recommended. Furthermore, those provisions on individual samples shall also apply to arsenic, cadmium, lead, nickel and total gaseous mercury. Moreover, subsampling of PM10 filters for metals for subsequent analysis is allowed, providing there is evidence that the subsample is representative of the whole and that the detection sensitivity is not compromised when compared with the relevant data quality objectives. As an alternative to daily sampling, weekly sampling for metals in PM10 is allowed provided that the collection characteristics are not compromised. For total deposition, Member States may use wet-only sampling, instead of bulk sampling, if they can demonstrate that the difference between them is within 10 %. Deposition rates shall generally be given as μg/m2 per day. The following criteria shall be used for checking validity when aggregating data in order to calculate statistical parameters: Parameter | Required proportion of valid data 1-hour means | 75 % (i.e. 45 minutes) 8-hour means | 75 % of values (i.e. 6 hours) 24-hour means | 75 % of the 1-hour means (i.e. at least 18 hourly values during the day) Maximum daily 8-hour mean | 75 % of the hourly running 8-hour means (i.e. at least 18 eight-hour values during the day) An assessment of compliance with the relevant limit values and target values shall be carried out regardless of whether the data quality objectives for data coverage are achieved, provided the available data allows for a conclusive assessment. In cases relating to the short-term limit values and target values, measurements that only cover a fraction of the calendar year, and that have not delivered sufficient valid data as required by Point B, may still constitute non-compliance. Where this is the case, and there are no clear grounds to doubt the quality of the valid data acquired, this shall be considered an exceedance of the limit value or target value and be reported as such. The following information shall be compiled for zones where air quality modelling applications or objective estimation are used: a description of assessment activities carried out; the specific methods used, with references to descriptions of the method; the sources of data and information; a description of results, including uncertainties and, in particular, the extent of any area or, if relevant, the length of road within the zone over which concentrations exceed any limit value, target value or long-term objective, and of any area within which concentrations exceed the assessment threshold; the population potentially exposed to levels in excess of any limit value for protection of human health. | 1. | To ensure accuracy of measurements and compliance with the data quality objectives laid down in Point A of this Annex, the appropriate competent authorities and bodies designated pursuant to Article 5 shall ensure the following: (a) that all measurements undertaken in relation to the assessment of ambient air quality pursuant to Article 8 are traceable in accordance with the requirements set out in the harmonised standard for testing and calibration laboratories; (b) that institutions operating networks and individual sampling points have an established quality assurance and quality control system which provides for regular maintenance and technical checks to assure the continued accuracy of measuring devices and that they remain operational; the quality system shall be reviewed as necessary and at least every 5 years by the relevant national reference laboratory; (c) that a quality assurance/quality control process is established for the process of data collection and reporting and that organisations appointed for this task actively participate in the related Union-wide quality assurance programmes; (d) that the national reference laboratories are appointed by the appropriate competent authority or body designated pursuant to Article 5 of this Directive and are accredited for the reference methods referred to in Annex VI to this Directive, at least for those pollutants for which concentrations are above the assessment threshold, according to the relevant harmonised standard for testing and calibration laboratories, the reference to which has been published in the Official Journal of the European Union pursuant to Article 2, point 9, of Regulation (EC) No 765/2008 of the European Parliament and of the Council (6) setting out the requirements for accreditation and market surveillance; those laboratories shall also be responsible for the coordination in Member State’s territory of the Union-wide quality assurance programmes to be organised by the Commission’s Joint Research Centre (JRC) and shall also be responsible for coordinating, at national level, the appropriate use of reference methods, and the demonstration of equivalence of non-reference methods; national reference laboratories organising intercomparison at national level shall also be accredited according to the relevant harmonised standard for proficiency testing; (e) that the national reference laboratories take part at least every 3 years in the Union-wide quality assurance programmes organised by the JRC for at least those pollutants for which concentrations are above the assessment threshold; participation for other pollutants is recommended; if that participation produces unsatisfactory results, then the national laboratory shall demonstrate at the next participation in the intercomparison satisfactory remediation measures, and provide a report to the JRC on these measures; (f) that the national reference laboratories support the work done by the European network of National Reference Laboratories set up by the JRC; (g) that the European network of National Reference Laboratories be responsible for the periodic review, at least every 5 years, of the measurement uncertainties of fixed measurements and of indicative measurements listed in Point A, Tables 1 and 2, of this Annex and subsequent proposal of any necessary changes to the Commission. | 2. | All reported data under Article 23 shall be deemed to be valid except data flagged as provisional. To promote and support the harmonised use of scientifically sound air quality modelling approaches by the competent authorities with an emphasis on model application, the appropriate competent authorities and bodies designated pursuant to Article 5 shall ensure the following: that the designated reference institutions participate in the European network of air quality modelling set up by the JRC; that best practices in air quality modelling identified by the network through scientific consensus are adopted in relevant applications of air quality modelling for the purposes of fulfilling legal requirements pursuant to Union legislation, without prejudice to model adaptations necessitated by singular circumstances; that the quality of relevant applications of air quality modelling is periodically checked and improved through intercomparison exercises organised by the JRC; that the European network of air quality modelling be responsible for the periodic review, at least every 5 years, of the maximum ratio of modelling uncertainties listed in Point A, Tables 1 and 2, of this Annex and subsequent proposal of any necessary changes to the Commission.

← ntr2-L_202402881EN.004101-E0002 · All articles · Data quality objectives →

Source: EUR-Lex (Cellar) · retrieved 2026-10-16 · Text as adopted (Official Journal); later amendments are not incorporated in this text.