«Sef van den Elshout October 2014 AQI & AQ Communication D2.3/v07 PartII Work Package 2, Task 2.2 Deliverable D2.3 Version number V07 Ver Date Author ...»
4.2. Policy indicators The year average AQI-related products are particularly useful for monitoring the evolution of air quality over years or for assessing the impact of air quality policy over time. It summarizes for the lay public and for public authorities (often no air quality specialists) the status of the air quality and the results of measures taken. For this purpose it is quite relevant to not only use the results of the monitoring sites but to generalise the results to make them more representative.
This can be done by weighting the results of individual monitoring sites as was proposed by Žujić et al (2009) for hourly/daily observations. The better method in my view is to categorise the results according to land use, to exposure type, etc. Combined with the number of inhabitants in each category a measure of ‘environmental pressure’ or ‘environmental comfort’ can be designed that could be used for trend monitoring. The most detailed way of doing this is by using modelling as a sophisticated kind of interpolation between the monitoring sites.
There are many ways of devising such an indicator and again one is not necessarily better than another. Bell et al (2011) present a few examples and stress the point that the design of the indicator has a major impact on the outcome and the relative ranking of the cities that were studied. The main point is that the same method has to be applied each year in the same way, after all these year average AQI-s are relative indicators as well. Their main use is in comparing situations (over time).
Policy indicators and the year average AQI-s are relevant to the authorities making policy but are also an excellent communication tool towards the public, showing that the relevant authorities are accountable. A selection of easy to understand indicators can be made for this purpose.
4.2.1. The EEA Environmental indicators The EU’s environemental agency, EEA, has developed a large set of indicators that is annually monitored and for which the memberstates obligatory deliver annual data. This covers all kinds
of environementla domains,including air quality. The air quality indicators can be found at:
They cover emissions but also exposure data. See for example: www.eea.europa.eu/dataand-maps/indicators/exceedance-of-air-quality-limit-1/exceedance-of-air-quality-limit-5.
The data is annually published by EEA (see www.eea.europa.eu/publications/air-quality-ineurope-2013) but every indicator also has its webpage (as the “exceedance of air quality limit” example shows) elaborating how the data are collected, time series of results, etc. Often even access to underlying raw data is possible.
The indicators have been developed and refined over the past decades and are now cornerstones of EU policy development/monitoring. Check this part of the EEA website for useful examples of indicators with a longer (longer than hours/days) averaging times.
Page 28 of 43 AQ Communication D2.3/V7 part II Above we discussed ways concentration or AQI maps can be population weighted. There are two schools of thought. One is, as described above, trying to relate as good as possible spatial concentration information to population numbers. The second approach in this is taken by EEA, and it is probably relevant to most of the medium sized EU cities, assuming that one concentration figure for a city is good enough because people don’t always stay at home but rather move around. In fact if one bases exposure only on residential address the true exposure could indeed be underestimated as during the day people tend to move to industrial areas or the city centre for their employement, school or other services.
4.3. AQI maps AQI maps can be made in near real-time by interpolating between monitoring sites or by running models. The AQI maps are increasingly common. In the past it was not always evident that the departments responsible for monitoring also had the modelling expertise to make realtime maps and the work was also computationally demanding, making it uncertain that an hourly updated map could be produced. Computational problems are generally solved by the ever increasing computer power, and many websites and apps have map presentations.
If the maps are meant to warn the public of adverse air quality it suffices to make a map showing the AQI only. If the presentation is also used for raising awareness one would like to see at least also which pollutant determines the index at a given time and space. This implies providing two types of information in a single map. This can be achieved by combining index colours which different hatchings for each pollutant. If this is done with transparent colours (to be able to see the geographic features on the base map) the result is likely to be a rather messy image that is hard to interpret. In one EU project it appeared that the AQI map often had the same colour throughout the area. It so happened that areas with different pollution problems still produced the same AQI as the highest pollutant determines the AQI. They dropped the idea of making AQI maps. In the US they present an AQI map and one can toggle between the AQI and the PM2.5 and O3 iAQI maps. This is a nice way of solving this problem and it works well with these two pollutants (a sensible selection both from a health and a communication perspective).
Modelled maps have one problem that is awkward for AQI-s aimed at alerting the public. The chemical transport models that are used to calculate these maps tend to work with fairly large gridcells averaging all pollution within the grid. Typical cell sizes in operational applications range from 5x5 to 20x20 km in the EU. The smaller the cell the more demanding the computations are and smaller cells are only viable if spatially detailed emissions are available.
The larger cells are more common. Averaging emissions over an area of 20x20 km can easily obscure locally relevant sources and even make cities below one million in habitants disappear in the general background. In several EU applications the few cities with sizes from 5-10 million can be seen and others can’t. In reality also the smaller cities experience higher concentrations of primary pollutants than the surrounding countryside. Grid averaging, by definition, reduces the spatial variability and hides/underestimates areas with high concentrations. There are ways
to prevent this but practical applications are not common yet. So the warning here is that modelled AQI maps often paint a better picture than the true situation: they underestimate high concentrations (and inflate low concentrations - but this is less important). They are most suited for situations where large scale atmospheric phenomena determine the AQI (O3, PM2.5 episodes).
Figure 3: www.obsairve.eu website/app, showing air quality in Europe.
The NO2 map shows elevated concentrations in the very big cities (London, Paris, Madrid), the industrial zone in northern Italy is visible but smaller with a million inhabitants or less, that should nevertheless show higher NO2 concentrations than their surroundings are not visible.
Page 30 of 43 AQ Communication D2.3/V7 part II
5. Literature Bell, Michelle L., Luis A. Cifuentes, Devra L.Davis, Erin Cushing, Adriana Gusman Telles and Nelson Gouveia. 2011. Environmental health indicators and a case study of air pollution in Latin American cities. Environmental Research 111, 57-66.
Burden, F.R., Ellis, P.S., 1996. Air pollution indices for Victoria. Clean Air 30, 26-30.
Cairncross, E.K., John, J. and Zunckel, M. 2007. A Novel Air Pollution Index Based on the Relative Risk of Daily Mortality Associated with Short-term Exposure to Common Air Pollutants. Atmos. Environ. 41: 8442-8454.
Chen Renjie, Wang Xi, Meng Xia, Hua Jing, Zhou Zhijun, Chen Bingheng, et al. 2013.
Communicating air pollution-related health risks to the public: an application of the Air Quality Health Index in Shanghai, China. Environ Int. 2013; 51:168–73.
COMEAP. 2011. http://comeap.org.uk/images/stories/Documents/Reports/comeap review of the UK air quality index.pdf. (Accessed 23-7-2012).
Cuilenburg, J.J. van, O. Scholten en G.W. Noomen, Communicatiewetenschap, Muiderberg:
Dimitriou K, Paschalidou AK, Kassomenos PA. 2013. Assessing air quality with regards to its effect on human health in the European Union through air quality indices. Ecol Indic 2013; 27:108–15.
Dusseldorp, A., P. Fischer, M. Dijkema and M. Strak. 2014. Luchtkwaliteitsindex.
Aanbevelingen voor de samenstelling en duiding. RIVM Briefrapport 2014-0050/2014.
Elshout, Sef van den, Karine Léger and Hermann Heich. 2013. CAQI Common Air Quality Index
- Update with PM2.5 and sensitivity analysis, Sci Total Environ, http://dx.doi.org/10.1016/j.scitotenv.2013.10.060 Elshout, S. van den. 2012. www.airqualitynow.eu/download/CITEAIR-Comparing_Urban_ Air_Quality_across_ Borders.pdf. (Accessed 13-7-2012) Elshout, S van den, K. Léger and F. Nussio. 2008. Comparing urban air quality in Europe in real time, a review of existing air quality indices and the proposal of a common alternative.
Environ. Int. 2008;34(5):720-6.
Garcia, J., Colossio, J. and Jamet, P., 2002. Air-quality indices. Elaboration, uses and international comparisons. Les presses de l’Ecole des Mines, Paris.
Johnson, B.B. 2003. Communicating air quality information: Experimental evaluation of alternative formats. Risk Analysis 23, 91-103.
Ruggieri, M. and Plaia A. 2012. An aggregate AQI: Comparing different standardizations and introducing a variability index. Sci Total Environ. 2012 Mar 15;420:263-72.
Shooter, David and Peter Brimblecombe. 2009. Air quality indexing. International Journal of Environment and Pollution. Volume 36, Number 1-3 / 2009.
Sicard P, Talbota C, Lesnea O, Mangina A, Alexandreb N, Collompb R. 2012. The aggregate risk index: an intuitive tool providing the health risks of air pollution to health care community and public. Atmos Environ 2012; 2012:11–6.
Stieb, D. M., Burnett, R. T., Smith-Doiron, M., Brion, O., Shin, H. H. Economou, V., 2008. A new multipollutant, no-threshold air quality health index based on short-term associations observed in daily time-series analyses. J Air Waste Manage Assoc 58, 435-450.
WHO Regional Office for Europe (2013). Review of evidence on health aspects of air pollution – REVIHAAP Project. Technical Report. www.euro.who.int/__data/assets/pdf_file/0004/ 193108/REVIHAAP-Final-technical-report-final-version.pdf?ua=1 Wikipedia. 2013. http://en.wikipedia.org/wiki/Air_quality_index (Accessed 8-4-2013).
Wong,Tze Wai, Wilson Wai San Tama, Ignatius Tak Sun Yua, Alexis Kai Hon Lau, Sik Wing Pang, Andromeda H.S. Wong, 2013. Developing a risk-based air quality health index. Atmos.
Env. Vol 76, 2013, Pages 52–58.DOI: 10.1016/j.atmosenv.2012.06.071.
Žujić, Aleksandra M., Bojan B. Radak, Anka J. Filipović and Dragan A. Marković. 2009.
Extending the use of air quality indices to reflect effective population exposure.
Environmental Monitoring and Assessment. Vol. 156, 2009.
The UK AQI uses triggers for the components with moving average concentrations to assure that warnings are better timed. For a full description see the COMEAP rapport (http://comeap.org.uk/images/stories/Documents/Reports/comeap review of the UK air quality index.pdf; Accessed 23-7-2012). The next figure (from annex 9 of the report) provides
the explanation of how the triggers work:
The moving average reacts quite late to the rising hourly values and the public might be informed rather late.
Don’t wait until the 8-hour average reaches a value of 80 µg/m3 but use a trigger at 82
µg/m3. If the hourly concentration reaches 82 and if the subsequent hour is even higher:
issue a warning.
Note that this system works during rising concentrations but not during falling concentrations.
Simply switching to hourly values and an hourly concentration grid is an easier solution that works both ways. Nevertheless this is an important step forward to improve real-time information.
Secondly the Type 1.b (in the main text) AQI-s where pollutant interaction is taken into account.
Lately many examples of sophisticated health based indices have been published. However few of them are actually in use. Apart from Canada, described here, Hong Kong is the only place where such kind of index was implemented. Examples from South Africa and Canada are given.
- The DAPPS uses relative risk (RR) values for individual pollutants, multiplies each concentration with their RR-1 and adds up the total to achieve a sum of RR-s. The total RR determines the degree of hazard the prevailing air presents. For a full discussion see the paper by Cairncross et al (2007).8 Here the sub-index bands are reproduced that allow a comparison between the DAPPS and the other indices (at iAQI level).
Total Risk = Σ((RRj-1)*Cj) where RRj is the RR of pollutant j and Cj is the concentration of pollutant j at the averaging time corresponding to RRj.
Free download at: http://www.ehrn.co.za/publications/download/04.pdf (Accessed 6-6-2013)
The summing of the individual RR-s (implicitly) assumes that the effects of each individual pollutant is independent of that of the others, which is not true. One could say that whereas the ‘highest iAQI determines the AQI’ concept is likely to underestimate the impact, this approach somewhat overestimates the likely health impact. (Unless the RR-s are derived from multi-pollutant models.) Also note that the amount of pollutants included determines to some extent the outcome. For consistent application of the DAPPS everyone should monitor the same set of pollutants.