Identify and quantify pollutants in the air

Metrology is necessary to ensure and maintain confidence in measurements. For the measurement to have meaning and for the results to be incontestable and to be comparable over time and between other places, each measurement must be connected to the national reference standard by an uninterrupted chain of comparison; This is called metrological traceability. Airparif is a link in this chain of traceability, it participates in the operation and consistency of air quality monitoring.


The measuring device

The number of measuring stations and their distribution across the territory comply with rules set out at European level. These rules are essential to ensure quality and comparable measurement across regions and countries. For Île-de-France, different types of stations and analyzes are spread over a 100 km radius around Paris and measure the quality of the air breathed by the population (more than 12 million inhabitants throughout the region) whatever their environment, near or far from sources of pollution. Airparif has around 70 measuring stations, including more than fifty permanent stations, real mini static laboratories.

 

  • Automatic and permanent stations 7 days a week, and hour by hour 

The automatic stations provide continuous measurement every hour, and allow monitoring of air quality in real time. The choice of the location of automatic stations and the pollutants measured there responds primarily to a public health concern. There are two typologies: bottom stations, far from sources of pollution, particularly traffic routes, and stations close to traffic. This classification is identical in France and in Europe.

Discover in pictures the Halles measuring station (Paris 1er:         

                                                                                                                   

  • Semi-permanent stations reinforcing the network near traffic                 

With a network of around twenty semi-permanent stations along the traffic routes, the automatic network is reinforced by measurements using diffusion tubes. This method guarantees regular but not continuous monitoring and makes it possible to calculate an annual average and estimate the situation with regard to regulatory values.                                                                      

  • Delayed analyzes                                                                                                                 

Certain pollutants (metals and certain compounds among Polycyclic Aromatic Hydrocarbons - PAHs) are not measured automatically and in real time by the stations. Their measurement requires sampling (on filter or diffusion tubes) and deferred analysis techniques in specific laboratories, or even specific measurement campaigns (dioxins, pesticides, etc.). Consult the map of stations in the Airparif measurement network.                           

                                                                      

From measurement to information

Each year, from automatic devices or manual measurements, nearly 6 million new pollution data converge on Airparif's central computer. To control this uninterrupted flow of information, it is necessary to store, validate and classify data in order to discern average data from peak data, compare them to current standards and provide information to everyone in the form of reports, graphs, pollution maps or air quality indices. 

 

1. Take 

Automatic analyzers draw air from outside the measuring station, at a sampling head, using a pump. On the other hand, diffusion tubes absorb pollutant passively (Fick's law).


2. Analyze

The concentration of pollutants is directly estimated by the automatic analyzer based on the optical, physical or chemical characteristics of the pollutants. Samples from filters or sorbents and passive tubes require laboratory analysis.


3. Transmit

The electrical signals provided by the automatic analyzers are converted into digital data by an acquisition chassis. Every quarter of an hour, it establishes quarter-hourly averages for all the analyzers based on measurements taken continuously. The data is then transmitted, several times a day, to Airparif servers and then stored in a database. In the event of a pollution episode, this data is automatically repatriated every quarter of an hour for all stations. As for laboratory analysis results, they are entered manually into the database.


4. Validate and use

From the "quarter-hour data" of the automatic stations, averages for each hour and for each day are calculated. The diffusion tubes provide weekly information of one sample on one day per week. All this data is then validated and then used by Airparif technicians and engineers. 
They feed other monitoring tools such as models to make maps, forecasts, evaluate the impact of reduction actions or the possible evolution of pollution levels in the years to come according to different hypotheses (link section modeling) . They also allow comparisons with the air quality criteria established by European directives or the recommendations of the World Health Organization.


5. Inform

Every day, data from the monitoring network makes it possible to calculate the air quality index broadcast, among others, on the Airparif website. Used by models, they make it possible to produce pollution maps and forecasts of air quality and pollution episodes. This is the information that is included by France 3 Île-de-France in its daily bulletin. 

The measuring stations also detect pollution episodes and Airparif relays this information to the authorities, the media and the public. 

Data from each measuring station are also available on Open Data


Calculating uncertainty

When a measurement is made, the result produced is associated with a certain degree of uncertainty. Indeed, depending on the measuring device used, the conditions in which the tests are carried out and a certain number of other parameters cause dispersion in the results produced. It is to quantify this variation that an uncertainty is defined for any measurement produced. 

Uncertainty defines the range within which the true value lies and with what probability. This uncertainty, assigned to the measurement result, characterizes the quality of the result produced: the lower the uncertainty associated with the result, the better this result will be.

The calculation of uncertainty takes into account the different components of the measurement and each of its stages. Errors can be due to sampling, the characteristics of the analyzers (linearity, reproducibility at zero and set point, response time, interferents), the acquisition of the measurement results provided by the analyzer, the rounding of the measurement result stored in the database, ambient conditions (temperature, humidity, pressure, etc.) as well as errors of the standards described in the calibration chain. The accuracy of the measurement result is linked to the difference between the standard used and the reading of this standard on the analyzer at the measurement site. This deviation is corrected on site by periodically and preventively adjusting the analyzer.