Assessment of Representativeness of Air Quality Monitoring Stations

Geneva, 11.6.2007

Wolfgang Spangl

26.10. 2006 | Folie 1

Service contract to the Commission for the Development of the methodologies to determine representativeness and classification of air quality monitoring stations

Contractor to DG ENV: Umweltbundesamt Austria Subcontracts with TNO Central Institute for Meteorology and Geodynamics, Vienna 26.10. 2006 | Folie 2

Motivation for Assessment of Representativeness

   Information about air quality is required for the whole territory (e.g. in the AQ framework Directive 96/62/EC) But: Monitoring gives information for distinct point locations (i.e. monitoring stations) Methods to get spatial information about air quality are required 26.10. 2006 | Folie 3

Spatial Information - Modelling

Spatial information about air quality can be achieved by

modelling

. However, models have some disadvantages:     Expensive High demand on input data (emission, meteorology, land use, …..) Limited resolution (regional, urban, street, …) Limited representation of „reality“ due to various uncertainties 26.10. 2006 | Folie 4

Spatial Information – Representativeness of monitoring data

Spatial information about air quality can be achieved by extending measured data to the “representative area” of a monitoring station.

To delimitate the representative area requires: 1.

A definition of „representativeness“ 2.

3.

Setting quantitative criteria for representativeness The identification of the area which fulfills the criteria for representativeness, based e.g. on a combination of model results, emissions, land use data, etc.

26.10. 2006 | Folie 5

“Representative area” of monitoring stations

Be aware that both   Definition and Quantitative criteria are deliaberate.

Other definitions and criteria are possible.

Definition and criteria esentially influence the outcome of representativeness assessment.

26.10. 2006 | Folie 6

Definition of Representativeness

The area of representativeness is defined by the criteria: 1 . The concentration within a certain range. The concentration is assessed according to limit and target values of EC legislation, related to annual means or annual exceedance numbers.

2. Similar concentrations shall be determined by similar reasons:  Emissions  Dispersion conditions due to buildings, topography and climate  Atmospheric transformation and transport 26.10. 2006 | Folie 7

Definition of Representativeness

Statistic parameters related to EC AQ regulations to determine representativeness:  PM10: Annual mean, 93.2-percentile of daily mean values (equivalent to 35 days per year above 50 µg/m³)  NO 2 : Annual mean  Ozone: 90.4-percentile of daily maximum 8 hour mean values (equivalent to 25 days per year above 120 µg/m³) 26.10. 2006 | Folie 8

Definition of Representativeness

The concentration in the area of representativeness of a certain AQ MS shall be within a range of 10% of the total concentration range observed in Europe.

 PM10: Annual mean: ±5 µg/m³, 93.2-percentile of daily mean values: ±8 µg/m³   NO 2 : Annual mean: ±5 µg/m³, which shall also be applied to NOx Ozone: 90.4-percentile of daily maximum 8 hour mean values: ±9 µg/m³ 26.10. 2006 | Folie 9

Definition of Representativeness

The spatial variation of parimary pollutants (incl. NO 2 ) is higher than for partly secondary pollutants (PM10) and secondary pollutants (Ozone). Therefore the representative areas, applying these criteria, are, on general, smaller for NO 2 and larger for Ozone.

26.10. 2006 | Folie 10

Similar reasons for similar concentrations:

Emissions

Classification according to emissions of   Local road traffic Domestic heating  Industry 3 classes each  A monitoring site is representative for areas falling into the same class as the respective monitoring site 26.10. 2006 | Folie 11

   

Similar reasons for similar concentrations: Dispersion and atmospheric transport

Local dispersion due to building structure and street geometry: street canyon; detached buildings; flat terrain; exposed Regional dispersion due to topography (10km): Flat terrain; hilly terrain; valley; basin; …… Large-scale (100km) Regions with different topography and climate: Alps (north/south), Po Valley, Pre-Alpine Lowlands, Pannonian Plane, ….

Maximum extension of representative area related to chemical (trans)formation of NO 100 km in Central Europe) 2 , O 3 , PM10, … (depending on average wind speed, radius approx 26.10. 2006 | Folie 12

Large-scale Regions with different topography and climate

26.10. 2006 | Folie 13

Methods to determine the area of Representativeness

Sources of spatial information:  Modelling  Measurement (regular monitoring networks, temporal measurements)  Surrogate information : emission inventories; surrogate for emission: land use (e.g. CORINE landcover), TeleAtlas roads), population distribution Model results and (additional) measurement data are used to derive relations between surrogate information and concentrations 26.10. 2006 | Folie 14

Area of Representativeness – based on model results and land use data

Klagenfurt (Austria), NO 2 Representative area of two monitoring stations:  urban background, annual mean 27 µg/m³ (yellow)  kerb side, annual mean 43 µg/m³ (red) 26.10. 2006 | Folie 15

Area of Representativeness – based on surrogate information

Illmitz (rural background), Ozone The area of representativeness is part of the Pannonian Plane ( dark green ) and excludes:  Area above 300m ( brown )  Large cities ( blue )  Major roads ( red ) 26.10. 2006 | Folie 16