Elsevier

Atmospheric Environment

Volume 43, Issue 25, August 2009, Pages 3858-3866
Atmospheric Environment

Meteorological variability in NO2 and PM10 concentrations in the Netherlands and its relation with EU limit values

https://doi.org/10.1016/j.atmosenv.2009.05.009Get rights and content

Abstract

The extent of the exceedance of the EU limit values for nitrogen dioxide (NO2) and particulate matter (PM10) concentrations within the Netherlands is expected to decrease significantly, in the coming years. Whether limit values will actually be exceeded, in the next decade, depends not only on European, national and local policies, but also on the effects of inevitable interannual meteorological fluctuations. An analysis of model calculations and measurements yields variations (1 sigma) in the annual average concentration of about 5% for NO2 and 9% for PM10, due to meteorological fluctuations. These deviations from long-term average concentrations affect assessments of future levels, set against limit values. For instance, an NO2 concentration of 39 μg m−3, estimated for a given year with long-term average meteorology, indicates that it is likely (chance >66%) that the limit value of 40 μg m−3 will not be exceeded in that particular year. At the same time, the estimation also indicates, for example, that this situation is unlikely (change <33%) to continue for three years in a row. However, with an estimated concentration of 38 μg m−3, it is likely that the limit value will not be exceeded for three years in a row. The limit value for the daily average PM10 concentration is equivalent to an annual average of about 32 μg m−3. This threshold is unlikely to be exceeded for three years in a row, when an annual average concentration of 29 μg m−3 is estimated. Interannual variations in concentrations of NO2 and PM10 are linked to large-scale meteorological fluctuations. Therefore, similar results can be expected for other European countries.

Introduction

Since mid-2008, the new EU Directive on Air Quality (EU, 2008) has been in force within Europe, with limit values for several air-quality components. The most stringent limit values for the Netherlands are 1) 40 μg m−3 for the annual average NO2 concentration, to be met by 2010, and 2) 50 μg m−3 for the daily average PM10 concentration, not to be exceeded more often than 35 times a year, from 2005 onwards. The EU directive offers the possibility to derogate from these limit values for several years. Ultimately, including the allowed derogation, the limit values for NO2 and PM10 concentrations must be adhered to, by 2015 and 2011, respectively.

In Europe, the attention directed at air quality is a result of not only the adverse effects on human health from high concentrations of air pollutants in the ambient air (Brunekreef and Holgate, 2002, Gauderman et al., 2007, WHO, 2004, WHO, 2005), but also of the EU regulations which aim to reduce emissions (EU, 2001) and to improve air quality (EU, 2008). Limit value exceedances are not unique to the Netherlands. Air-quality limit values are currently being exceeded at many locations around Europe (Mol et al., 2008). The Netherlands is unique, however, in its implementation of the EU Directive on Air Quality (EU, 2008) into national legislation (Backes et al., 2005). The national law states that new building projects (roads, factories, etc.) cannot be started until an assessment of the effects on the future air quality is presented, in which is shown that, once the project has been completed, the air quality will comply with the standards set by the EU directive. The implementation of this law has resulted in delays or even a freeze, imposed by the Council of State (the Dutch court in such cases) of a large number of building projects. This has drawn extra attention to the need for an improvement in the air quality in the Netherlands, to comply with the EU directive, as much as possible. Moreover, Dutch citizens can now demand that the authorities take actions to improve the air quality, as was recently acknowledged by a verdict of the European Court of Justice.

Measurements and model calculations are used for assessing current and making estimates of future concentrations of NO2 and PM10, in the Netherlands (Velders et al., 2008). Both measured and modelled concentrations are inhibited by an uncertainty range of up to about 15–20% (Velders and Diederen, 2009). Although this uncertainty range still meets the requirements of the EU directive, it affects the conclusions that can be drawn when assessing compliances with EU limit values. Velders and Diederen (2009) compared projections of NO2 and PM10 concentrations with the limit values, taking these uncertainties into account. Because of the uncertainties, they do not make absolute statements on whether concentrations will fall below the limit values, in time, but discuss the likelihood of meeting these limit values.

Apart from uncertainties associated with measurements, model calculations, and emissions, concentrations are also affected by inevitable interannual meteorological fluctuations. Assessments of future concentrations are made in the Netherlands using long-term average meteorological data. Since PM10 and NO2 concentrations must be brought and then kept below their limit values for all years onwards of, ultimately 2011 and 2015, respectively, it is important to know the magnitude of the effect of the meteorological fluctuations on the concentrations. This magnitude will affect the likelihood of meeting the EU limit values in future years, and can be used for assessing by how much the concentrations must be brought below the limit value to reduce the risk of exceedances.

Air-quality assessments based on measurements and modelling have been published for several countries and models (Cuvelier et al., 2007, Jimenez-Guerrero et al., 2008, Monteiro et al., 2007, Stedman et al., 2007, Vautard et al., 2008), but few have quantified the effects of meteorological fluctuations on annual average surface concentrations or air pollutants (Andersson et al., 2007). It is well established that meteorological fluctuations, to a large degree, determine surface concentrations of air pollutants in European countries (see, for example, EMEP, 2008, Holst et al., 2008, Minguzzi et al., 2005, Rost et al., 2009, Solberg et al., 2008). In the Netherlands, concentrations of air pollutants are mostly affected by the large-scale synopses. With westerly winds, relatively clean moist air masses are transported from over the Atlantic to the Netherlands. Southerly winds carry polluted air masses to the Netherlands, while easterly winds carry polluted dry air masses to the Netherlands in combination with high temperatures. Because the Netherlands is small in size, emissions from sources within the country and from neighbouring countries, to a large degree, determine the surface concentrations of air pollutants (see Section 2.1).

Over the last decades, many studies have identified the characteristics of air pollution concentrations, in terms of sources and sinks in relation to meteorological processes. In that respect, PM10 and NO2 are different. Where sources of NO2 are largely anthropogenic and rather well defined, those of PM10, characteristically, are much more distributed, and often disperse. NO2 concentrations in the Netherlands are, on average, for more than 60% determined by national anthropogenic emissions; for PM10 this is less than 20%. PM10 concentrations are the result of many different emissions and processes, and the contribution from long-range transport plays a more important role than for NO2. The effect of meteorology differs per particle fraction (Buijsman et al., 2005). In the Netherlands, about half the PM10 concentrations are of (semi) natural origin: sea salt, mineral dust and water. About a third of PM10 concentrations consist of particles which are chemically formed in the air from gaseous precursors, mostly of anthropogenic, but also of biogenic origin. The rest, around 20%, is due to the emission of primary particles directly emitted from mostly anthropogenic sources. Measurements at street, urban and rural locations show that dynamic behaviour of NO2 and PM10 in the Netherlands is similar to other locations in Europe (Mol et al., 2008). Model studies confirm that air-quality dynamics which determine the concentrations of NO2 and PM10 in the Netherlands are intrinsically the same for each type of location in Europe, although concentration levels may differ, due to other sources and source strengths (EMEP, 2008).

The meteorological variability in NO2 and PM10 concentrations is studied here by using two approaches. The first approach is based on an analysis of measured concentrations of NO2 (22 years of data) and PM10 (16 years of data) at rural background, city background, and street locations, in the Netherlands. The second approach uses model calculations governed by meteorological fields between 1981 and 2007, using the same emission data for all years. In Section 2, a description is given of the measurements and emission data used in approach 1 and the model calculations used in approach 2. The effects of the meteorological variability on the NO2 and PM10 concentrations are presented in Sections 3 Variability in NO, 4 Variability in PM, respectively. A discussion on and conclusion of the results from both approaches and the effect of the meteorological variability on the likelihood of meeting the EU limit values in the Netherlands is given in Section 5.

Section snippets

Approach 1: trends in measured concentrations and emissions

In the first approach, to determine the effects of the meteorological variability on annual average NO2 and PM10 concentrations, measurements for the Netherlands are analysed. For each location, a linear trend is removed from the time series of measured concentrations and the remaining variation in the concentrations is ascribed to meteorological fluctuations. Statistical quantities are derived from these variations to yield information on the effect of meteorological fluctuations on the

Variability in NO2 concentrations

The effects of the meteorological fluctuations on the measured and modelled annual average NO2 concentration are shown in Fig. 3 and Table 1. Analyses show that the deviations of the NO2 (and PM10) concentrations from the linear trend are, to a large degree, distributed normally. Random errors in the annual average measured concentration of 2% have been subtracted from the derived variability (see Section 2.1). Both the measured and modelled concentrations show an interannual variability. The

Variability in PM10 concentrations

The effects of the meteorological fluctuations on the measured and modelled annual average PM10 concentration are shown in Fig. 4 and Table 2. This study considers only the annual average PM10 concentration, since that is calculated by the OPS model. An empirical relation is used for relating these to the limit value for the daily average PM10 concentration. Observations in the Netherlands (Matthijsen and Visser, 2006) show a strong correlation between the number of days on which the daily

Discussion on the effects on limit values

Measurements and model calculations are used for assessing current and future concentrations of NO2 and PM10, within the Netherlands. Projections of concentrations are derived from scenarios for emissions and long-term average meteorology (Velders and Diederen, 2009). These estimates are used for assessing whether limit values will be met in time, in the Netherlands. In busy streets and along motorways, concentrations of NO2 and PM10 are expected to remain close to the EU limit values, over the

Acknowledgements

We thank Hub Diederen, Robert Koelemeijer and Hans Visser for fruitful discussions and Annemieke Righart for editorial comments.

References (39)

  • J. Van de Kassteele et al.

    Uncertainty assessment of local NO2 concentrations derived from error-in-variable external drift kriging and its relationship to the 2010 air quality standard

    Atmospheric Environment

    (2006)
  • C. Andersson et al.

    Interannual variation and trends in air pollution over Europe due to climate variability during 1958–2001 simulated with a regional CTM coupled to the ERA40 reanalysis

    Tellus

    (2007)
  • C.W. Backes et al.

    Transformation of the first Daughter Directive on air quality in several EU Member States and its application in practice

    European Environmental Law Review

    (2005)
  • F.T. Blank

    Measurement Uncertainty in the National Air Quality Monitoring Network (LML)

    (2001)
  • E. Buijsman et al.

    Particulate Matter: a Closer Look

    (2005)
  • EMEP

    Transboundary Acidification, Eutrophication and Ground Level Ozone in Europe in 2006

    (2008)
  • EU

    Directive 2001/81/EC of the European Parliament and the Council of 23 October 2001 on the National Emissions Ceilings for Certain Atmospheric Pollutants

    (2001)
  • EU

    Directive 2008/50/EC of the European Parliament and of the Council of 21 May 2008 on Ambient Air Quality and Cleaner Air for Europe

    (2008)
  • R. Gense et al.

    Latest Insights into Direct NO2 Emission from Road Transport, the Current State of Knowledge

    (2007)
  • Cited by (16)

    • Effects of European emission reductions on air quality in the Netherlands and the associated health effects

      2020, Atmospheric Environment
      Citation Excerpt :

      The air quality limit values addressed here are for NO2 the annual average concentration of 40 μg m−3, for PM10 the daily average concentration of 50 μg m−3 which may not be exceeded more than 35 times a year, and for PM2.5 the annual average concentration of 25 μg m−3. The PM10 daily limit value is found to be equivalent to an annual average concentration of about 32 μg m−3 (Matthijsen and Visser, 2006; Velders and Matthijsen, 2009). The modelled concentrations are therefore compared with this lower concentration.

    • Ensemble classification for identifying neighbourhood sources of fugitive dust and associations with observed PM<inf>10</inf>

      2017, Atmospheric Environment
      Citation Excerpt :

      Particulate matter (PM) is a highly erratic pollutant in urban landscapes due to its formation from both mechanical and chemical processes, sensitivity to meteorological conditions, volatile residence times as a result of sedimentation and increased likelihood of impaction given the larger built-up footprint in urban areas (Beelen et al., 2009; Velders and Matthijsen, 2009; Zwack et al., 2011; Barmpadimos et al., 2011).

    • Temporal stability of land use regression models for traffic-related air pollution

      2013, Atmospheric Environment
      Citation Excerpt :

      Further, a decrease in the explanatory power of the models in hindcasting scenarios was observed in all of the studies. There was also an overall decrease in measured NO2 concentrations in the Netherlands (Eeftens et al., 2011; Velders and Matthijsen, 2009) and in Rome between the two measurement periods, although the magnitude of the decrease was smaller than in Vancouver. This result supports our conclusion that hindcasting is less accurate than forecasting in areas where ambient concentrations are decreasing over time, due to the necessity of extrapolating beyond the range of the measured concentrations.

    View all citing articles on Scopus
    View full text