Ten-year measurements of gaseous pollutants in urban air by an open-path analyzer

Ten-year measurements of gaseous pollutants in urban air by an open-path analyzer

ARTICLE IN PRESS Atmospheric Environment 42 (2008) 4138–4148 www.elsevier.com/locate/atmosenv Ten-year measurements of gaseous pollutants in urban a...

610KB Sizes 3 Downloads 46 Views

ARTICLE IN PRESS

Atmospheric Environment 42 (2008) 4138–4148 www.elsevier.com/locate/atmosenv

Ten-year measurements of gaseous pollutants in urban air by an open-path analyzer P. Avino, M. Manigrasso Laboratorio Inquinamento Chimico dell’Aria, DIPIA-ISPESL, via Urbana 167, 00184 Rome, Italy Received 19 November 2007; received in revised form 14 January 2008; accepted 14 January 2008

Abstract Object of this study is the versatility of a remote-sensing measurement system through experimental survey of the levels of primary and secondary pollutants in urban areas. For the first time a long-range measurement of different pollutant for a long period are reported in the urban area of Rome. Over 10 years of data collection, levels of NO2, O3, SO2, benzene and toluene are measured and discussed by means of a differential optical absorption spectroscopy (DOAS) system installed at ground level. The period is interesting because it represents one of the most important life-changes in Europe and in Italy: in fact, during the 1990s some significant decrees introduced the catalytic pots and limited the amount of benzene and sulfur dioxide in the fuel. The paper will describe how the main primary and secondary pollutants are changed during all the period: particularly, benzene and SO2 are not over a big social problem for the human health whereas the secondary pollutants, ozone and nitrogen dioxide, have the same trends during the period, meaning a no contribution to their evolution from the atmospheric processes. The DOAS technique is resulted very useful and versatile: comparisons with other traditional analyzers (i.e., GC-PID for benzene and toluene investigations) demonstrate how it allows to have information over all the territory saving the qualitative trends and giving a similar quantitative result. In this respect the authors show that the difference is due to the different sampling ways (the traditional analyzers are punctual whereas the DOAS is a remote-sensing technique covering a wide area) and does not influence the atmospheric process interpretation. Finally, it should be considered that the pollutant trends have been interpreted using the concentration measurements of natural radioactivity as tracer of the dynamic properties of the atmospheric boundary layer. r 2008 Elsevier Ltd. All rights reserved. Keywords: DOAS; Air pollution; Primary pollutant; Radon; Photochemical smog

1. Introduction In order to improve the air quality in Europe, the European Commission issued some Directives with Corresponding author. Tel.: +39 06 4714242; fax: +39 06 4744017. E-mail address: [email protected] (P. Avino).

1352-2310/$ - see front matter r 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.atmosenv.2008.01.024

regard to acceptable levels of a range of gaseous pollutants species as SO2, NO, NO2, O3 and benzene (Council Directive 96/62/EC, 1996). In particular, measurements of ozone and benzene represent nowadays a very important target in the atmospheric studies, due to the different sources and health effects of both gaseous species. Benzene being recognized as carcinogenic for human (IARC

ARTICLE IN PRESS P. Avino, M. Manigrasso / Atmospheric Environment 42 (2008) 4138–4148

Monograph, 1987) and due to its chemical stability, has constituted an element of constant hygienicsanitary interest since many years. Ozone has important adverse health (Mustafa, 1990; Pryor et al., 1995; Cotgreave, 1996; Hazuchaa and Lefohn, 2007) and environmental effects such as respiratory irritation: chemically, it is the primary constituent of smog, it is not emitted directly into the air by specific sources but created by photochemical reactions involving sunlight, nitrogen oxides and volatile organic compounds (VOCs) (Morales et al., 2004). Among the different methods used to analyze these compounds, the techniques involving openpath spectroscopy are ideal for their advantages over classical methods and point-source analyzers (Avino et al., 2004). The differential optical absorption spectroscopy (DOAS) is one of the most versatile optical techniques for the determination of air pollutants. DOAS is a useful measuring technique for routine work to control air quality and pollution emissions, but also for applications in air pollution studies (Platt et al., 1979; Platt and Perner, 1980; Karlsson, 1990). The DOAS is a remote-sensing analytical technique finding strong consensus as demonstrated by numerous papers published during these last years. The DOAS analytical method is based on the UV and vis light absorption (from 240 to 340 nm) of species with fine vibrational structures like SO2, NO2, ozone, nitrous acid, formaldehyde, benzene and toluene; the Lambert–Beer’s law regulates the relationship between the absorbed light intensity and the concentration of each species. This technique furnishes the concentrations related to a portion of atmospheric environment varying also up to some kilometers. In fact, these ‘‘integrated’’ concentrations represent the average area pollution level better than the measurements obtained by traditional analyzers, which are considered ‘‘punctual’’. The constituents investigated with DOAS technique include sulfur dioxide (SO2), nitrogen dioxide (NO2), nitrous acid (HNO2), formaldehyde (HCHO), ozone (O3), ammonia, mercury vapor (actually, these last two pollutants can be detected using a long optical path length) and aromatic hydrocarbons (benzene and toluene). This papers reports measurements of benzene, toluene, SO2, NO2 and O3 carried out in downtown Rome in the decade from 1991 to 2000 by means of

4139

a DOAS system installed at about 10 m above the ground level. All the measurements are interpreted using the planetary boundary layer (PBL) evolution assessed by means of a radon analyzer (Avino et al., 2003). 2. Experimental The DOAS system (mod. AR 500, Opsis, Sweden) has been installed at the ISPESL monitoring station located in downtown Rome (near St. Maria Maggiore Cathedral) at ground level in an urban zone characterized by high density of anthropogenic activities (autovehicular traffic, domestic heating, etc.). The DOAS system consists of an emitter (a xenon lamp at high pressure), a receiver, a spectrophotometer equipped with an optical fiber and a computer for the system management (data elaboration and data storage). The absorption spectra of each monitored chemical species are acquired at their relative typical wavelength ranges; subsequently, the interferences are eliminated by comparison with the reference spectra. The distance between emitter and receiver is about 280 m: this parameter is important because it influences the sensitivity of the measures. The absorbance of light from the emitter is continuously measured within the wavelength range 240–350 nm to determine several compounds. The aromatic hydrocarbons are detected in the wavelength range between 250 and 290 nm, where the major interfering gases are oxygen, ozone and sulfur dioxide; around 100 spectra per second are collected in this wavelength range and stored in a register with 1000 channels with a resolution of better than 0.05 nm. The concentrations of air pollutants are automatically calculated from the absorbance values, in according with Lambert–Beer’s law (Platt et al., 1979; Platt and Perner, 1980; Brocco et al., 1992). From the ambient air recording, a differential absorption spectrum for the atmosphere gases is produced in two steps. First, the ambient air recording is divided by a system reference spectrum, i.e. a spectrum which is pre-recorded in the total absence of gaseous absorption. The result is a total absorption spectrum of the ambient air. Second, a fifth-order grade polynomial function is fitted to the spectrum to eliminate all broadband influences in the atmosphere. From this quotient the logarithm is calculated to obtain a differential absorption spectrum. The concentrations are calculated by

ARTICLE IN PRESS P. Avino, M. Manigrasso / Atmospheric Environment 42 (2008) 4138–4148

4140

fitting a theoretical spectrum, based on the prestored differential cross section Da(i) onto the measured spectrum A(i). The differential cross sections Da(i), i.e. the reference spectrum, are produced in a laboratory with the same equipment, by using pure standard gas. The unknown concentration C is then varied until the equation DaðiÞLC ¼ AðiÞ

(1)

achieves the best solution. This is done by minimizing the following expression, using least square fit: X ½AðiÞ  LCDaðiÞ2 , (2) i

where L is the optical path length. The channels (i) used in the evaluation are chosen for two criteria: a longer number of channel will improve the statistical error and the intervals should be chosen in order to minimize influences from interfering compounds, noise and other disturbance. When several substances are interfering, prerecorded differential cross section curves Daj(i) for each one of the components are used in the deconvolution of the measured spectrum. Expressions (1) and (2) above will be rewritten as ½Da1 ðiÞC 1 þ Da2 ðiÞC 2 þ Da3 ðiÞC 3 þ   L ¼ AðiÞ (3) and XX 2 ½AðiÞ  LC j Daj ðiÞ . i

(4)

j

A factor CjDaj is included in the analysis for each one of the additional substances. For aromatic hydrocarbon measurements only oxygen, ozone and sulfur dioxide are taken into consideration (Hallastadius et al. 1991). Spectra are required on an average time of 7 min for the system located in downtown Rome. The specification provided by the manufacturers of the instrument used are:

  

Minimum detectable concentrations over 280 m ¼ 5 mg m3. Zero point stability ¼ +10 mg m3 per month. Linearity ¼ +10 mg m3 in the range 0–100 mg m3.

The cross sensitivities on benzene and toluene measurements, evaluated in accordance with the U.K. National Physical Laboratory methods (Partdridge

et al., 1995), are o2% of the concentration of each interfering gas (ozone, SO2 and NO2 were tested). The Radon concentrations were measured by means of a SM200 Atmospheric Stability Monitor (Opsis, Sweden) (Avino et al., 2003).

3. Results and discussion 3.1. Planetary boundary layer evolution The temporal trend of pollutant concentration can be interpreted using the measurements of the concentrations of natural radioactivity as tracer of the dynamic properties of the atmospheric boundary layer (Avino et al., 2003). The atmospheric primary pollution level is in fact determined by both the intensity of its emissions and the dynamic properties of the atmospheric boundary layer, causing pollutant dilution or accumulation. In this respect, radon and its short-lived decay products can be considered as a tracer of the mixing properties of the lower boundary layer, since radon is emitted at a rate that is spatially and temporally constant (Porstendorfer et al., 1991; Febo et al., 1996), consequently its concentration is mainly determined by the dilution properties of the atmospheric boundary layer. The temporal evolution of Radon concentration can be described by the following equation: qC ¼ aFR  bfC R g þ Adv; qt

(5)

where a is a parameter that links the mixing properties of the boundary layer with the source intensity, FR is the emission rate of Radon, b{CR} is a term taking into account the vertical mixing properties of the atmosphere, Adv is the advective term. Atmospheric stability conditions maximize aFR (b{CR} is negligible) whereas atmospheric instability makes maximum the contribution of the term b{CR}.

3.2. Primary pollutants The average concentrations of primary pollutants can be regarded as closely related to the intensity of their emissions, so that their trends throughout a wide temporal window of measurements directly reflect important changes in the life style and the

ARTICLE IN PRESS P. Avino, M. Manigrasso / Atmospheric Environment 42 (2008) 4138–4148

pollutant dispersion. Further, it should be noted that the drastic decreasing in concentrations levels occurred in 1994, 3 years before that the outcome of the law on the limit of the 1% v/v of benzene in gasoline. The annual average toluene concentrations (Fig. 3) passed from 108 mg m3 in 1991 to 54 mg m3 in 2000, following the same decreasing trend as benzene one, with a toluene/benzene ratio of about 5, in agreement with the values of 3–5 reported in literature for urban area pollution (Field et al., 1992). The two pollutants are well correlated due to their common origin related to the autovehicular traffic, as shown in Fig. 4 (R2 ¼ 0.984), reporting typical daily trends in the urban area of Rome during a cold period. In particular, looking at the Fig. 4 it can be possible to extrapolate some interesting considerations. First of all, the benzene and toluene concentrations in urban area during a winter period are perfectly coherent stressing the same origin sources for both pollutants. The high hourly levels reached by benzene (up to 80 mg m3) during the evening of the 3rd depend strictly on the stability conditions in that period. In fact, during the next days remixing conditions are present and the benzene levels go down to below 10 mg m3 until the afternoon of the 7th when new stability conditions are present again. This boundary layer evolution is typical in a Mediterranean area, especially in the urban area of Rome where cyclic stability/instability meteo conditions are present. Fig. 4 (middle) shows the trends of benzene calculated by two different systems: one by the on-line DOAS and the second by a off-line GCPID (Syntech-Spectra, The Netherlands) analyzer sampling every 15 min. It should be noted the

20

4

15

3

10

2

5

1

0

Benzene (% v/v)

Gasoline (Mt)

enactment of legislative decisions meant to intervene directly on the source of pollution. Even if the gasoline consumption increased in the decade 1990–2000 (De Lauretis et al., 2003), we have assisted to a 60% reduction of benzene emissions in Italy (De Lauretis et al., 2003) as shown in Fig. 1. Such reduction was promoted by the growing percentage of cars with catalytic converter and by the Italian Law 413/97 (Law n, 1997) that fixed the limit of 1% v/v of benzene in gasoline. The same law prescribes that gasoline dispensing facilities must be equipped with vapor recovery equipment to control emissions of VOCs during the filling of on-road motor vehicle fuel tanks. In an urban context dominated by the autovehicular traffic pollution, as downtown Rome, a direct consequence of such legislative decisions was the steady reduction of the benzene emissions. In this context can be placed the sequence of 10 years of measurements taken at the ISPESL monitoring station and can be understood the overall decline of benzene annual average concentrations (Fig. 2). Taking in account this consideration and the frame of measurements determined in Rome, it can be understood the overall decline of benzene and toluene annual average concentrations (Figs. 2 and 3). Analyzing Figs. 2 and 3, we observe how the benzene and toluene levels vary. In particular, the annual benzene trend, starting from 25 mg m3 in 1991, reaches 8 mg m3 in 2000: maximum relative levels of benzene are found during the last months of each year (from 30 mg m3 in December 1991 to 13 mg m3 in December 2000) when meteorological conditions of high pressure are unfavorable for the

4141

0 1990

1991

1992

1993

1994

1995 Year

1996

1997

1998

1999

2000

Fig. 1. Gasoline consumption in Italy (bar) and relative benzene concentration allowed (line) (De Lauretis et al., 2003).

ARTICLE IN PRESS P. Avino, M. Manigrasso / Atmospheric Environment 42 (2008) 4138–4148

4142

40 Annual average

35 Benzene (μg/m3)

30 25 20 15 10 5 0 01/91

01/92

01/93

01/94

01/95

01/96 Year

01/97

01/98

01/99

01/00

Fig. 2. Monthly (bar) and annual (line) average levels of benzene concentrations (mg m3) measured by a DOAS system in downtown Rome.

200 Annual average

180

Toluene (μg/m3)

160 140 120 100 80 60 40 20 0 01/91

01/92

01/93

01/94

01/95

01/96 Year

01/97

01/98

01/99

01/00

Fig. 3. Monthly (bar) and annual (line) average levels of toluene concentrations (mg m3) measured by a DOAS system in downtown Rome.

behaviors not depending on the sampling or detection systems: in fact, the trends appear similar even if the DOAS technique is a remote-sensing analysis involving a wide area portion whereas the GC-PID system is a punctual analyzer determining the benzene and toluene levels at the relative sampling point. According to these definitions we are expected to find the same benzene modulation but lower values in the remote-sensing measurements as confirmed in Fig. 4 (middle). Finally, Fig. 4 (below) reports the correlation curve between benzene data collected by DOAS and GC-PID systems in the same period: R2 is very good (0.984) and the relative equation is y ¼ 0.427x0.047 meaning the strict coherence among the measurements performed by different apparatus.

In the same decade, other important European Directives, 93/12/EEC (Council Directive 93/12/ EEC, 1993) and 98/70/EC (Council Directive 98/70/ EC, 1998), have imposed and have scheduled for the next years considerable reductions of the sulfur content in fuels (Fig. 5). Consistently, the annual average concentrations of SO2 measured at the ISPESL Pilot Station are characterized by a clear decreasing trend (Fig. 6). In fact, by analyzing Fig. 6, we can understand that the annual levels vary from 15 mg m3 in 1991 to 4 mg m3 in 2000. As in the case of the previously analyzed pollutant, the maximum relative levels are found during the coldest periods (every January from 36 mg m3 in 1992 to 10 mg m3 in 2000).

ARTICLE IN PRESS P. Avino, M. Manigrasso / Atmospheric Environment 42 (2008) 4138–4148

4143

300

90

250

70 60

200

50 40

150

30

100

20

Toluene (μg/m3)

Benzene (μg/m3)

80

50

10 0

03/02

04/02

05/02

06/02 date

07/02

08/02

0 09/02

04/02

05/02

06/02 date

07/02

08/02

09/02

140

Benzene (μg/m3)

120 100 80 60 40 20 0 03/02 35 30

BTX

25 20

y = 0.427x - 0.047

15

R2 = 0.984

10 5 0 0

10

20

30

40 DOAS

50

60

70

80

Fig. 4. (above) Typical winter (3–9 February 1997) trends of benzene and toluene in downtown Rome; (middle) comparison trends of benzene determined by GC-spectra analyzer (line) and DOAS system (dot line); (below) correlation curve between GC-spectra and DOAS data.

This occurrence is due to the contribution of domestic heating, whereas the benzene, toluene and SO2 levels decrease clearly during summer period due to the meteorological conditions characterized by long periods of solar radiation, favoring the pollutant dispersion due to convective mixing (Avino et al., 2003). 3.3. Secondary pollutants For secondary pollutants, such as ozone and nitrogen dioxide, no clear trend can be recognized

looking at their average annual concentrations (Figs. 7 and 8), because they are not directly related to the pollution sources. Their pattern of variation is determined both by the atmospheric stability conditions and by the pool of reactions involving O3, NOx, VOCs and UV radiation. As regards the month average concentrations (Fig. 7), in summer ozone concentrations are higher than in the rest of the year, due to the higher photochemical activity in this period. Nitrogen dioxide exhibits a complementary trend (Fig. 8) to the O3 one. NO2 is produced through the oxidation

ARTICLE IN PRESS P. Avino, M. Manigrasso / Atmospheric Environment 42 (2008) 4138–4148

4144

3500 Gasoline Diesel fuel

3000

Sulphur content (ppm)

3000 2500 2000 2000 1500 1000 1000

1000 500 500

500

350 150

50 50

10 10

0 1990-1993

1994-1995

1996-1999

2000-2004

2005-2008

2009-

Years Fig. 5. Sulphur content (ppm) allowed in gasoline and diesel fuels (European Directives).

45 Annual average

40

SO2 (μg/m3)

35 30 25 20 15 10 5 0 01/91

01/92

01/93

01/94

01/95

01/96

01/97

01/98

01/99

01/00

Year Fig. 6. Monthly (bar) and annual (line) average levels (mg m3) of SO2 concentrations measured by a DOAS system in downtown Rome.

of NO by O3 and the radical RO2 and HO2. Ozone is formed through the reaction of O2 with atomic oxygen deriving from the photolysis of NO2. No O3 accumulation would occur if the formation of NO2 were due only to the reaction of O3 with NO. In this case NO2 and O3 show a complementary pattern of variation: maximum values of O3 corresponds to minimum values of NO2 and viceversa. However, in the case of intense episodes of pollution, in presence of reactive hydrocarbon and of OH radicals, the RO2 and HO2 radical that are formed oxidize NO and cause O3 accumulation (Avino, 2004). The occurrence of radical oxidative processes in atmosphere can be represented through

the variable Ox, defined as the sum of O3 and NO2 concentrations. During periods of low photochemical activity and of high advection, Ox is characterized by a trend fairly constant with oscillation around the background ozone value. In this case, since ozone is mainly formed through the photolysis of NO2 and decomposed by reaction with NO, the two pollutants show a complementary pattern of variation (Fig. 9). When the advective transport mechanism is negligible and during atmospheric stability periods, radical oxidative processes may play an important role. Ox trend is very well structured (Fig. 10). It displays minimum values at night, almost due to

ARTICLE IN PRESS P. Avino, M. Manigrasso / Atmospheric Environment 42 (2008) 4138–4148

4145

80 Annual average

70

Ozone (μg/m3)

60 50 40 30 20 10 0 01/91

01/92

01/93

01/94

01/95

01/96 Year

01/97

01/98

01/99

01/00

Fig. 7. Annual (line) and monthly (bar) average levels (mg m3) of ozone concentrations measured by a DOAS system in downtown Rome.

80 Annual average

70

NO2 (μg/m3)

60 50 40 30 20 10 0 01/91

01/92

01/93

01/94

01/95

01/96

01/97

01/98

01/99

01/00

Year

Fig. 8. Annual (line) and monthly (bar) average levels (mg m3) of NO2 concentrations measured by a DOAS system in downtown Rome.

NO2, because the ozone concentrations are very low. During the day, peak values above the background value are indicative of radical oxidative activity. As a consequence, O3 and NO2 present a no more specular trend of variation. The radon concentration trend (Fig. 10) displays a peculiar modulation characterized by maximum values at night, due to nocturnal atmospheric stability, alternating with minimum values during the day, due to convective mixing. It is interesting to note that such pattern is complementary to the Ox one: minimum radon concentration value (high degree of mixing of the low atmospheric layer) correspond to maximum Ox values (high solar

radiation and radical oxidative activity) and viceversa. During nighttime, ozone level within the mixing height are relatively low due to the reaction of O3 with NO. However, since such mechanism of destruction of O3 operates in the presence of emission sources of NOx, if transport of ozone occurs away from NOx sources, O3 decomposition is no more possible. During the daytime, when the solar radiation is maximum and the convective mixing mechanism is effective, ozone is homogeneously distributed through the mixing layer. At night, due to the decline of the mixing height, NOx emissions at the ground level can no more mix with

ARTICLE IN PRESS P. Avino, M. Manigrasso / Atmospheric Environment 42 (2008) 4138–4148

4146

120 Ox O3 NO2

100

(μg/m3)

80 60 40 20 0 5/11

6/11

7/11

8/11 date

9/11

10/11

11/11

6/11

7/11

8/11 date

9/11

10/11

11/11

90 80

β count (a.u)

70 60 50 40 30 20 10 0 5/11

Fig. 9. Typical daily trends of ozone, NO2 and Ox (above) and of radon concentration (below) during a cold period (2–11 November 1996).

the layer above, consequently destruction of O3 in the higher level is not possible and such level can act as a reservoir of ozone. As a consequence, ozone peaks during nocturnal hours can be measured in the presence of a vertical component of the wind. This is the case of Fig. 11 reporting the average hourly concentrations of O3 in winter period near an industrial site and displaying nighttimes ozone peaks of the same order of magnitude of those diurnal. 4. Conclusions The DOAS technique is an advanced measurement system useful for integrating information coming from monitoring networks: furthermore, it allows investigating and understanding the complex mechanisms of formation and transformation of atmospheric pollutants through the contemporary

measurements of gaseous primary and secondary species. In this way the representation of the air pollution is more correct than using automatic analyzers. The results obtained take into account the spatialtemporal variations of the pollutant concentrations due to both the atmospheric dishomogeneity and the contributions of sources not directly present in the considered area. Data referring to 10-year time span from 1991–2000 in downtown Rome, have clearly shown a steadily decreasing trend for some important primary pollutants, such as benzene and SO2. The reduction of their atmospheric concentrations was strictly related to some normative decisions that limited the concentration of benzene and sulfur contents in fuel. Non-clear temporal trends have been observed for the secondary pollutants, ozone and dioxide nitrogen, in the investigated period.

ARTICLE IN PRESS P. Avino, M. Manigrasso / Atmospheric Environment 42 (2008) 4138–4148

4147

14 12

Ox (ppb)

10 8 6 4

Radon (counts)

2 0 1/8

2/8

3/8 date

2000 1800 1600 1400 1200 1000 800 600 400 200 0 1/8

2/8

3/8 Date

4/8

4/8

5/8

Fig. 10. Typical daily trends of the variable Ox (above) and of radon concentration (below) in a summer period (1–4 August 1997).

100

Ozone (ug/m3)

80

60

40

20

0 4

5

6

7

8

9

10

11 12 13 14 15

16 17 18 19 20

21 22 23 24 25 26 27 28 29

Date Fig. 11. Hourly average concentrations of ozone measured by a DOAS system in February 2000.

Their evolution has been deeply studied trough the knowledge of the variable Ox (O3+NO2) as representative of oxidative radicalic phenomena occurring in atmosphere. The radon short-lived

product measurements has provided a useful description of the dynamic evolution of the low boundary layer for interpreting the daily and seasonal trends of the pollutants investigated.

ARTICLE IN PRESS 4148

P. Avino, M. Manigrasso / Atmospheric Environment 42 (2008) 4138–4148

Acknowledgments This work was supported by ISPESL/DIPIA/P14 Grant. The authors wish to thank Drs. L. Lepore and D. Brocco for your kind help in the result discussion and the anonymous reviewers for having improved the paper. References Avino, P., 2004. Mechanism of smog photochemical formation in the urban area of Rome. In: Zerefos, CS. (Ed.), Proc XX Quadrennial Ozone Symposium, vol. 2. Kos, Greece, pp. 936–937. Avino, P., Brocco, D., Lepore, L., Pareti, S., 2003. Interpretation of atmospheric pollution phenomena in relationship with the vertical atmospheric remixing by means of natural radioactivity measurements (Radon). Annali di Chimica 93, 589–594. Avino, P., Brocco, D., Lepore, L., Russo, MV., Ventrone, I., 2004. Remote sensing measurements for evaluation of air quality in an urban area. Annali di Chimica 94, 707–714. Brocco, D., Fratarcangeli, R., Lepore, L., Ventrone, I., 1992. Spettrometria di Assorbimento Ottico Differenziale per la valutazione della qualita` dell’aria. Acqua e Aria 7, 617–627. Cotgreave, IA., 1996. Absorption and metabolic fate of ozone— the molecular basis of ozone-induced toxicity. Scandinavian Journal of Work, Environment and Health 22 (Suppl. 3), 15–26. Council Directive 93/12/EEC of 23 March 1993 relating to the sulphur content of certain liquid fuels. Council Directive 96/62/EC of 27 September 1996 on ambient air quality assessment and management. Council Directive 98/70/EC of the European Parliament and of the Council of 13 October 1998 relating to the quality of petrol and diesel fuels and amending Council Directive 93/12/ EEC. De Lauretis, R., Ilacqua, M., Romano, D., 2003. Emissioni di benzene in Italia dal 1990 al 2000. Rapporti X/2003. APATDipartimento Stato dell’Ambiente, Controlli e Sistemi Informativi-Unita` Interdipartimentale Censimento Fonti di Emissione. Febo, A., Perrino, C., Giliberti, C., Allergrini, I., 1996. Use of proper variables to describe some aspects of urban pollution.

In: Allegrini, I., De Santis, F. (Eds.), NATO ASI Series. Springer, Berlin, Haidelberg, pp. 295–315. Field, RA., Goldstone, ME., Laster, JN., Perry, R., 1992. The sources and behaviour of anthropogenic volatile hydrocarbons. Atmospheric Environment 16, 2983–2996. Hallastadius, H., Uneus, L., Walling, S., 1991. A system for evaluation of trace gas concentration in the atmosphere based on the DOAS technique. In: Schiff, I. (Ed.), vol. 1433, Proceedings SPIE, pp. 36–43. Hazuchaa, MJ., Lefohn, AS., 2007. Nonlinearity in human health response to ozone: experimental laboratory considerations. Atmospheric Environment 41, 4559–4570. Karlsson, K., 1990. Environmental control using long path measurements. Proceedings of the 1990 EPA/AWMA— International Symposium Res. Tri. Park, NC, 1–7. IARC Monograph 1987. Supplement 7. Law n. 413 of 04/11/1997 ‘‘Misure urgenti per la prevenzione dell’inquinamento atmosferico da benzene’’ Gazzetta Ufficiale Italiana n. 282 of 03/12/1997. Morales, JA., Treacy, J., Coffey, S., 2004. Urban ozone measurements using differential optical absorption spectroscopy. Analytical and Bioanalytical Chemistry 379, 51–55. Mustafa, MG., 1990. Biochemical basis of ozone toxicity. Free Radical Biology and Medicine 9, 245–265. Partdridge, R.H., Curtis, I.H., Goody, B.A., Woods, P.T., 1995. An evaluation of the performance of an open path atmospheric air-quality monitor manufactured by Opsis, Sweden. NPL Report Qu109, Div Quantum Metrology, National Physical Laboratory, Teddington, Middlesex, U.K. Platt, U., Perner, D., 1980. Direct measurement of atmospheric CH2O, HNO2, O3 and SO2 by differential optical absorption in the near UV. Journal of Geophysical Research 85, 7435–7458. Platt, U., Perner, D., Patz, HW., 1979. Simultaneous measurement of atmospheric CH2O, O3 and NO2 by differential optical absorption. Journal of Geophysical Research 84, 6329–6335. Porstendorfer, J., Buterweck, G., Reineking, A., 1991. Diurnal variation of the concentration of radon and its short-lived daughters in the atmosphere near the ground. Atmospheric Environment 25, 709–713. Pryor, WA., Squadrito, GL., Friedman, M., 1995. The cascade mechanism to explain ozone toxicity: the role of lipid ozonation products. Free Radical Biology and Medicine Dictionary 19, 935–941.