- Email: [email protected]
Variability in exposure to ambient ultrafine particles in urban schools: Comparative assessment between Australia and Spain
Environment International 88 (2016) 142–149
Contents lists available at ScienceDirect
Environment International journal homepage: www.elsevier.com/locate/envint
Variability in exposure to ambient ultrafine particles in urban schools: Comparative assessment between Australia and Spain Mandana Mazaheri a,b, Cristina Reche c, Ioar Rivas c,d, Leigh R. Crilley e, Mar Álvarez-Pedrerol d, Mar Viana c, Aurelio Tobias c, Andrés Alastuey c, Jordi Sunyer d, Xavier Querol c, Lidia Morawska a,b,⁎ a
International Laboratory for Air Quality and Health, Queensland University of Technology, Brisbane, Australia Institute for Health and Biomedical Innovation, Queensland University of Technology, Brisbane, Australia Institute of Environmental Assessment and Water Research, Spanish National Research Council Barcelona, Spain d Centre for Research in Environmental Epidemiology, Barcelona, Spain e School of Geography, Earth and Environmental Sciences, University of Birmingham, Birmingham B15 2TT, UK b c
a r t i c l e
i n f o
Article history: Received 31 August 2015 Received in revised form 23 November 2015 Accepted 20 December 2015 Available online 4 January 2016 Keywords: Ambient particles Exposure New particle formation Urban environments High insolation areas
a b s t r a c t Ambient ultrafine particle number concentrations (PNC) have inhomogeneous spatio-temporal distributions and depend on a number of different urban factors, including background conditions and distant sources. This paper quantitatively compares exposure to ambient ultrafine particles at urban schools in two cities in developed countries, with high insolation climatic conditions, namely Brisbane (Australia) and Barcelona (Spain). The analysis used comprehensive indoor and outdoor air quality measurements at 25 schools in Brisbane and 39 schools in Barcelona. PNC modes were analysed with respect to ambient temperature, land use and urban characteristics, combined with the measured elemental carbon concentrations, NOx (Brisbane) and NO2 (Barcelona). The trends and modes of the quantified weekday average daily cycles of ambient PNC exhibited significant differences between the two cities. PNC increases were observed during traffic rush hours in both cases. However, the midday peak was dominant in Brisbane schools and had the highest contribution to total PNC for both indoors and outdoors. In Barcelona, the contribution from traffic was highest for ambient PNC, while the mid-day peak had a slightly higher contribution for indoor concentrations. Analysis of the relationships between PNC and land use characteristics in Barcelona schools showed a moderate correlation with the percentage of road network area and an anti-correlation with the percentage of green area. No statistically significant correlations were found for Brisbane. Overall, despite many similarities between the two cities, school-based exposure patterns were different. The main source of ambient PNC at schools was shown to be traffic in Barcelona and mid-day new particle formation in Brisbane. The mid-day PNC peak in Brisbane could have been driven by the combined effect of background and meteorological conditions, as well as other local/distant sources. The results have implications for urban development, especially in terms of air quality mitigation and management at schools. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Children's exposure to air pollution, along with characterization of their exposure–response relationship has been of great interest recently. A simple search on the Web of Science using the combined keywords “children”, “air pollution”, “exposure”, “epidemiology” and “schools” results in more than 80 publications in the past 15 years, with more than 30 of them published since 2010 (within the past 5 years). Children are known to be more physiologically susceptible to air pollution and other environmental related health risks than adults, since their organs are still developing and they have higher levels of physical activity and ⁎ Corresponding author at: International Laboratory for Air Quality and Health, Institute of Health and Biomedical Innovation, Queensland University of Technology, GPO Box 2434, Brisbane, QLD 4001, Australia. E-mail address: [email protected] (L. Morawska).
http://dx.doi.org/10.1016/j.envint.2015.12.029 0160-4120/© 2015 Elsevier Ltd. All rights reserved.
breathing rates in comparison to adults (Trasande and Thurston, 2005; Hornung et al., 2009). In 2013, the Health Effects Institute (HEI) identified ultrafine particles (UFPs) as one of the air pollutants of most concern, due to a lack of sufficient studies on their impact on human health (HEI, 2013). However, until 2011, most of the efforts in relation to school-based air quality studies were focused on particulate matter in terms of particle mass and on the PM2.5 and PM10 size ranges (Mejía et al., 2011). Since the Mejía et al. (2011) literature review, some comprehensive studies on UFP characteristics at schools have been conducted around the world, such as in Australia (Ultrafine Particles from Traffic Emissions and Children's Health: UPTECH) (Salimi et al., 2013; Laiman et al., 2014; Mazaheri et al., 2014; Crilley et al., 2016; Ezz et al., 2015), Spain (BRain dEvelopment and Air polluTion ultrafine particles in scHool childrEn: BREATHE) (Reche et al., 2014; Rivas et al., 2014; Viana et al., 2014; Sunyer et al., 2015), Italy (Buonanno et al., 2012, 2013), and the US (Zhang and Zhu, 2012). Indoor concentrations
M. Mazaheri et al. / Environment International 88 (2016) 142–149
of UFP are driven by the infiltration of outdoor particles (especially for naturally ventilated school classrooms), classroom activities during teaching hours, and cleaning before and after school hours (Guo et al., 2008; Tippayawong et al., 2009; Raysoni et al., 2011; Laiman et al., 2014; Reche et al., 2014; Viana et al., 2014). Outdoor sources include a wide range of local and distant sources, such as cooking within the school premises, vehicular traffic (Morawska et al., 1998a,b, Hitchins et al., 2000) or new particle formation mechanisms (Reche et al., 2011a,b). These studies have pointed to the important contribution of environmental factors specific to each study area on UPF characteristics and air pollution in general, in both indoor (teaching classrooms) and outdoor (e.g. playgrounds) school microenvironments. This work uses the large body of air quality data that were collected within the scope of the UPTECH (in Brisbane, Australia) and BREATHE (in Barcelona, Spain) projects and provides a comparative assessment of the differences and similarities in school-based exposure to ambient particle number concentrations (PNC) in high insolation environments with different urban profiles (e.g. urban, fleet and air pollution characteristics). 2. Materials and methods 2.1. Study area The study areas were Brisbane, Australia and Barcelona, Spain, which share a range of similar features, particularly climatic conditions, but with different traffic and urban development characteristics (Table 1). The statistics provided in the table are average values for Brisbane and Barcelona, which were current at the time of the UPTECH (2010–2012) and BREATHE (2012–2013) projects. 2.2. Data source The data used for this study were available from the UPTECH and BREATHE project databases: • The UPTECH data were collected at 3 outdoor and 2 indoor sites (naturally ventilated teaching classrooms) in 25 schools within the Brisbane Metropolitan Area, from October 2010 to August 2012. The UPTECH indoor data used in this study were averaged over the two classrooms, as no major differences were observed between
Table 1 General urban and air quality characteristics in Brisbane and Barcelona. General urban and air quality statistics
Brisbane, Australia
Barcelona, Spain
Population1,2 Total fleet density (km−2)1,2,3 Total passenger vehicles density (km−2)1,2,3 % of passenger vehicles to total registered vehicles1,2,3 % of motor vehicles to total registered vehicles — petrol1,3 % of motor vehicles to total registered vehicles — diesel1,3 % of motor vehicles to total registered vehicles — LPG/dual/other1,3 Maximum temperature (°C)4,5 Minimum temperature (°C)4,5 Relative humidity (%)4,5 Daily mean ambient PNC during weekdays (cm−3)6 Daily mean minimum and maximum traffic counts6 Total % of green area7,2 Total % of residential area7,2 Total % of roads7,2
241,264 3757 2810 75 77 21 2
1,602,386 9050 5800 62 64 35 1
21.9–30.2 10.1–21.5 65 8985 34–1063 50 17 5
19.2–32.3 3–18.5 65 20,548 2,630–54,522 28 25 23
1 2 3 4 5 6 7
(Australian Bureau of Statistics). (Barcelona City Council). https://sedeapl.dgt.gob.es/IEST2. (Australian Bureau of Meteorology). Faculty of Physics from Barcelona University. Averaged values during the monitoring campaigns. QLD Government (https://data.qld.gov.au/).
143
classrooms at each school (Laiman et al., 2014) (https://www.qut. edu.au/research/research-projects/uptech). • The BREATHE data were collected at one indoor (naturally ventilated teaching classroom) and one outdoor location in 36 schools in Barcelona, as well as 3 in Sant Cugat del Vallès (an adjacent municipality located 5 km away from Barcelona), through two separate campaigns conducted between 27 January and 22 June 2012, and 14 September 2012 and 22 February 2013 (http://www.creal.cat/projectebreathe). The data from both campaigns were used in this paper.
Considering the study designs for both projects, vehicular traffic was expected to be the main source of air pollution in the vicinity of the schools. General characteristics of the studied schools in the two cities are summarised in Table 2. For the UPTECH, air quality parameters were measured at one school at a time for 2 consecutive weeks at each school, and for the BREATHE, they were measured at two schools at a time for one week. Parameters that were measured during weekdays (school days) and were common between the two projects or relevant to support the data analysis were used for this paper. These were: concurrently measured 24-h time-series of indoor and outdoor PNC at each school; offline measurements of outdoor elemental carbon (EC) concentrations in PM2.5 during school hours; passive measurements of outdoor NO2 concentrations during school hours for the BREATHE campaign; and time-series of outdoor NOx concentrations for the UPTECH (time-series of NO2 concentrations were not available for the whole UPTECH campaign). Meteorological conditions (temperature, humidity and solar radiation) for the UPTECH were measured at a central outdoor location at each school and at a reference urban location in Barcelona for the BREATHE. Realtime data were collected in 30 s sampling intervals for the UPTECH and 10 min sampling intervals for BREATHE. Local traffic counts in proximity of each school were available with 5 min resolutions for the UPTECH schools, while traffic counts for the BREATHE schools were carried out over 15 min, for two of the sampling days during morning traffic rush hours (09:15 CET approx.). Detailed information on the general aspects, design, instrumentation and data quality controls for each project are available in the previously published papers for the UPTECH (Salimi et al., 2013; Laiman et al., 2014; Crilley et al., 2016) and BREATHE (Reche et al., 2014; Rivas et al., 2014). This paper uses the hourly averaged data that were measured during weekdays for both of the projects. Some assumptions and compromises were made when analysing the data, mainly due to differences between the UPTECH and BREATHE study designs. The BREATHE used DiScmini at stationary sites for indoor and outdoor particle monitoring, measuring concentrations and average sizes of particles in the 10–700 nm range; whereas the UPTECH employed TSI condensation particle counters (CPCs) (models 3787 and 3781), measuring total concentrations of particles larger than 5 and 6 nm, respectively. Comparisons of the PNC characteristics for both projects were made based on the DiScmini Table 2 General school characteristics at the time of measurement campaigns.
School age (years) Floor type Schooling hours Classroom cleaning schedule Building material Traffic rush hours Window frames Teaching equipment
Classroom furniture
Brisbane, Australia
Barcelona, Spain
40–146 Carpet 09:00–15:00 05:00–09:00 and 15:00–18:00
7–9 Ceramic tiles 09:00–17:00 5:00–9:00 and 17:00–20:00
Bricks, concrete, wood 07:00–08:00 and 17:00–18:00 Wood and/or aluminium Whiteboard, desktop computer/s, overhead projector Standard school tables and chairs, bookshelves
Bricks, concrete, wood 8:00–9:00 and 20:00–21:00 Wood/PVC/aluminium Blackboard, overhead projector, desktop computer/s Standard school tables and chairs, bookshelves
144
M. Mazaheri et al. / Environment International 88 (2016) 142–149
and CPC measurements; even though they had somewhat different specifications for detecting and counting particles. This method was considered practical, given that both study areas were within urban settings in developed countries, with similar air pollution sources and PNCs in the UFP size range (b0.1 μm) (Morawska et al., 1998a,b; Reche et al., 2014; Salimi et al., 2014). In addition, this paper is not focused on a direct comparison of PNC absolute values between the two study areas; rather it compares the trends and influencing factors. The ambient air quality parameters measured at one outdoor location at each school were used for the analysis. In the case of the UPTECH, the outdoor data were taken from a centrally located outdoor location at each school. Outdoor data was only available for one location at each school for the BREATHE. This approach does not take into account the potential impact of PNC spatial variation within the school grounds. However, it was considered redundant, since PNC spatial variation was not observed for the majority of the UPTECH schools studied (Salimi et al., 2013). For simplicity, BNE is used wherever referring to the UPTECH schools in Brisbane, Australia and BCN, wherever referring to the BREATHE schools in Barcelona, Spain. 3. Data analysis 3.1. Daily cycles of the air quality parameters The mean daily cycles of indoor and outdoor PNCs (BNE and BCN), local traffic counts and outdoor NOx (BNE) were quantified for each school using the measured time-series, in order to analyse general PNC trends throughout the day (including traffic rush hours and during mid-day). Diurnal cycles were calculated in R using the openair package (Carslaw and Ropkins, 2012). In addition, the contribution of each mode to total PNC was estimated by calculating the sum of PNC for each mode as a percentage of the sum of total PNC during the day.
selected as the proxy for general climatic conditions. Schools were grouped according to their temperature ranges across the year, rather than based on the conventional definition of seasons. This is of most importance in Brisbane (BNE), where there is no conventional change of season and seasonal variation mainly consists of warmer (November to April) and cooler (May–October) months. In the case of Barcelona (BCN), December, January and February are the coldest months, while the warm period lasts about six months, from May to October, with July and August being the warmest months. PNC daily cycles were analysed with respect to the school campaigns conducted in the warmer and cooler months, in order to evaluate the impact of weather conditions throughout the year. 3.2.3. Urban characteristics Trends in mean ambient PNC time-series between schools in the two cities were compared against each city's urban characteristics; namely total built-up area, total number of registered vehicles and total number of vehicles using petrol and diesel fuels (Table 1). Furthermore, the influence of general urban characteristics, in terms of total road network and green areas, on mean ambient PNC was estimated according to local urban characteristics for the BNE and BCN schools. Open access datasets provided by the Queensland Government (https://data.qld. gov.au/) and the Barcelona City Council (http://www.bcn.cat/ estadistica/catala/dades/barris/index.htm) were employed for land use and locality analysis. For Brisbane, green areas comprised of the sum of data subsets for “recreation and culture”, “residual native cover”, “grazing native vegetation”, “other conserved area” and “other forest plantation”. In Barcelona, green areas were defined as the urban and forest parks. 4. Results and discussions 4.1. Temporal variations of the air quality parameters
3.2. Classification of the schools The impact of traffic on air pollution, as well as ambient climatic factors, on the diurnal trends of indoor and outdoor PNC, was determined using the relevant measured environmental parameters as signatures. The schools were grouped accordingly, as described in the following subsections: 3.2.1. Traffic conditions Instead of the local traffic counts, school hour EC, NO2 and NOx that were measured at the same time and location as outdoor PNC were used as proxies to determine the schools that were most likely affected by traffic related air pollution (Wang et al., 2013). This approach was used since traffic counts alone cannot explain the air quality levels, and the dispersion of traffic related emissions heavily depend on the local meteorological conditions (e.g. wind direction and speed) and site characteristics (e.g. topography, distant from the road, barriers, green space, orientation of the building). Classification analysis was performed based on the 50th percentile of the average concentrations of NO2 (for BCN) and NOx (for BNE) during school hours across all the schools. Schools with average concentrations higher than the 50th percentile were grouped as high traffic schools and the remainder was grouped as low traffic schools. Corresponding school hour mean EC concentrations for the high and low traffic schools and their trends were compared with mean PNC, since EC is the main component of traffic related particles. 3.2.2. Ambient climatic conditions The studied classrooms were all naturally ventilated. Thermal comfort in the classrooms was maintained by opening or closing the windows, as a result of variations in outdoor temperature and the thermal comfort preference of the occupants. Therefore, temperature was
Daily cycles of hourly averaged indoor and outdoor PNC for all BNE and BCN schools are available in the supplementary information (SI) document (Figs. S1 and S2). Two peaks, coinciding with the morning and evening traffic rush hours, were observed for all average PNC time-series in the two cities. At each individual BNE school, all average indoor PNC time-series and almost all outdoor PNC time-series exhibited an additional peak during mid-day. Indoor mean hourly PNCs were higher than for outdoors for more than half of the BNE schools. Trends in school-based average PNC daily cycles for the BCN were similar to the BNE; however the mid-day occurrence was less frequent and indoor PNCs were generally lower than outdoors. Higher indoor mid-day PNC was also observed for BCN schools, although with a much lower frequency than BNE. Fig. 1 presents mean daily cycles of the hourly averaged indoor and outdoor PNCs over all the schools for each project. The solid lines are the PNC and the shaded areas show the 95% confidence intervals. The corresponding traffic counts and NOx daily cycles, along with their 95% confidence intervals are shown in the SI document for BNE (Fig. S3). BNE NOx and traffic counts at the closest busy road to the school showed similar trends in terms of morning and evening traffic rush hours: 07:00–08:00 and 17:00–18:00. Traffic rush hours in Barcelona have been reported as 8:00–9:00 and 20:00–21:00 (Ajuntament de Barcelona, 2015). Three PNC peaks were observed for school-based indoor and outdoor time-series in both cities, corresponding to the morning and evening traffic rush hours and a mid-day peak (Fig. 1). The mid-day PNC mode for Brisbane (BNE) schools was around 11:00 and 14:00, which coincided with a small and rather insignificant peak in corresponding traffic counts and NOx, in comparison to the rush hour peaks (SI file, Fig. S3). This mid-day peak for Barcelona (BCN) schools was much smaller than the ones attributed to traffic rush hours.
M. Mazaheri et al. / Environment International 88 (2016) 142–149
145
Fig. 1. Daily cycles of the hourly averaged indoor and outdoor PNCs (cm−3) over all the BNE and BCN schools. Shaded areas represent 95% confidence intervals.
The results demonstrate clear differences between indoor and outdoor PNC daily cycles in the two cities. The major differences between PNC daily cycles for the two cities were: (a) in the magnitude and proportion of the mid-day peak in respect to the traffic rush hours for each city, which is further analysed in the following sub-section (Section 4.2); and (b) the mid-day PNC peak was more pronounced for indoors than outdoors in the BNE schools in comparison to BCN. Previous studies on ambient PNC increases during mid-day in Brisbane and Barcelona have attributed this peak to nucleation processes and new particle formation (Cheung et al., 2011, 2012; Reche et al., 2011a,b, 2014; Brines et al., 2015). In order to further investigate indoor PNC peaks in respect to ambient climatic conditions, the schools were grouped according to whether the campaign was conducted during the warmer or cooler months of the year (as described in Section 3.2) (Fig. 2).
4.1.1. UPTECH (BNE) Daily cycles of the hourly averaged indoor and outdoor PNCs for all BNE schools showed a small increase in indoor PNC before the start of teaching hours (around 07:00), which were lower than outdoor concentrations and were associated with early morning cleaning activities (05:00–09:00). Indoor PNCs were higher than outdoors between 08:00 and 18:00, corresponding with morning and evening cleaning activities (05:00–09:00 and 15:00–18:00) and teaching activities (09:00– 15:00). Comparing the daily PNC cycles for warmer and cooler months, BNE indoor PNCs generally exhibited a similar pattern to outdoors during the warmer months, when classrooms tended to have the windows open at all times during school hours and AERs were higher (Fig. 2). Previous analysis of indoor air quality for the BNE classrooms showed that the average air exchange rate during the warmer months was slightly higher than for the cooler months (0.6 vs. 0.5 h−1) (Laiman et al.,
Fig. 2. Mean daily cycles of indoor and outdoor PNCs (cm−3) during the cooler and warmer months for the BNE and BCN schools. Shaded areas represent 95% confidence intervals.
146
M. Mazaheri et al. / Environment International 88 (2016) 142–149
Fig. 3. Overall mean daily outdoor and indoor PNC (cm−3) for the identified high traffic (HT) and low traffic (LT) schools in BNE and BCN.
2014). During the warmer months, indoor PNCs were lower than outdoors before the start of each school day (09:00), gradually peaked above outdoor levels around 13:00 and then decreased to below outdoor levels at 17:00. In the cooler months, indoor PNC patterns differed during mid-day compared to the warmer months and on average, were 25% higher than outdoor levels. Indoor and outdoor PNC time-series measured during the weekend were used to further investigate the higher mid-day PNC peak indoors for BNE schools. During the weekends, classrooms were not in use, all of the doors and windows were shut, and no cleaning activities took place. Indoor PNC during mid-day was lower than outdoors during weekends for all of the BNE schools, as well as those grouped according to warmer and cooler months (SI file: Figs. S4 and S5). The results indicate that the prominent indoor mid-day peak was due to the combined effect of lower air exchange rates and potential indoor sources favouring new particle formation. It should be noted that the Brisbane schools did not have a dining hall, but a small stall-like shop (canteen) that served a variety of cold and warm foods. Furthermore, the canteens were not in close proximity to the studied classrooms and their immediate impact on indoor PNC during mid-day was implausible. Therefore, cooking related emissions were not considered as a driver for the observed mid-day peak in the average daily cycles.
hours. These peaks were more evident during the warmer period, especially for outdoor concentrations. Given that indoor PNCs were always lower than outdoors for hours outside of school activities in both cities (i.e. teaching and maintenance/cleaning hours), the higher indoor PNC indicated that indoor activities, cleaning methods and schedules were the most influential factors, which varied from one study area to another (Table 2). This was despite the fact that the classrooms were naturally ventilated and traffic density was generally higher for the BCN schools (Table 1). One of the significant findings of this research is the observed higher indoor versus outdoor PNC for UPTECH schools. Investigations in relation to identifying the specific source/s of the mid-day indoor PNC peaks require further data on the composition and size distribution of the indoor PNC and are the topic of our future research. Cluster analysis of the ambient/outdoor particle number size distributions (PNSD) in the BNE schools that were measured during the UPTECH project identified five clusters associated with three main particle sources, namely regional background, photochemically induced particles and traffic related particles (Salimi et al., 2014). For BCN schools, mean hourly indoor and outdoor particle sizes measured by the DiScmini were smaller during mid-day (SI file, Fig. S6). These outcomes support this work's findings, even though indoor PNSDs were not available for BNE schools.
4.1.2. BREATHE (BCN) Indoor PNC patterns were very similar to those found for outdoors, indicating that outdoor PNC could be responsible for indoor PNC variations. Similar to BNE, this was due to natural ventilation in the classrooms, as a result of opening windows in order to attain comfortable room temperatures. However, it is important to note that two main differences, namely: (i) increases in indoor PNC during morning rush hours usually occurred with a delay, coinciding with the beginning of school hours; and (ii) indoor PNC increases in the late afternoon/ evening were observed at around 17:00–19:00, not coinciding with outdoor peaks. This was associated with the cleaning of classrooms, which occurred once a day, mostly in the late afternoon/early evening (around 17:00–20:00) (Reche et al., 2014). In contrast to BNE schools, average indoor PNC daily cycles for the BCN showed that indoor concentrations were always lower than the outdoors by at least 10%, when averaged over all schools, as well as warmer or cooler months (Figs. 1 and 2). BCN schools were much newer than BNE schools (Table 2). Although building structure and age could be a defining factor when discussing the PNC I/O ratios, this could not be the case for the BCN indoor PNC being lower than outdoors. This is because all the classrooms were naturally ventilated and had windows opened, and therefore the differences in the effect of particle infiltration due to the gaps/leaks in the structure would have been negligible in this case. Small mid-day PNC peaks were observed for both indoors and outdoors at all schools, but they were generally lower than those observed during traffic rush
4.2. Influence of urban characteristics on ambient air quality Using the 50th percentile of NOx and NO2 concentrations as proxies for traffic related emissions, 13 out of the 25 BNE schools and 20 out of the 39 BCN schools were identified as high traffic schools (i.e. more likely to be affected by traffic related air pollution) and the remainder as low traffic schools. Mean school hour NOx for the high and low traffic BNE schools were 24 ppb and 10 ppb, respectively. In case of the BCN, mean NO2 was 28 ppb for the high traffic schools and 17 ppb for low traffic schools.
Fig. 4. Box plot of the overall mean outdoor EC (μg m−3) during school hours for the identified high traffic (HT) and low traffic (LT) schools in BNE and BCN.
M. Mazaheri et al. / Environment International 88 (2016) 142–149
147
Fig. 5. Fractions of the identified PNC peaks to the total mean daily PNC in BNE and BCN schools; H-O = High Traffic-Outdoors, H-I = High Traffic-Indoors, L-O = Low Traffic-Outdoors, L-I = Low Traffic-Indoors, A.M. Traffic = morning traffic rush hour, P.M. Traffic = evening traffic rush hour.
The school-based mean daily cycles of PNC for the BNE schools showed similar trends, where PNC peaked during traffic rush hours (Fig. S1). Similar trends were observed for the BCN PNC time-series (Fig. S2). Box plots of the mean indoor and outdoor daily PNC and outdoor school hour EC for the high and low traffic schools are presented in Figs. 3 and 4, respectively. Overall, significantly higher PNC and EC levels were observed in the BCN schools, and mean indoor and outdoor PNCs were found to be higher for high traffic schools. This was more evident for BCN, where the difference in outdoor PNC between the high and low traffic schools was 54% and the indoor difference was 38%. Significant differences (N10%) in the mean PNCs were not observed for the high and low traffic BNE schools. Comparison of the mean daily PNC between the high and low traffic schools in each city showed higher concentrations in BCN schools than in BNE. For high traffic schools, the differences between BCN and BNE were found as 89% for outdoors and 47% for indoors. For the low traffic BCN and BNE schools, the differences between mean daily PNC were 47% for outdoors and 11% for indoors. These results in terms of mean PNC daily cycles during weekdays for indoor and outdoor PNCs (cm−3) in the high and low traffic schools in the two cities are available in the SI file (Fig. S7). The corresponding weekend trends for the BNE schools are also available in the SI file (Fig. S8, weekend data are not available for BCN). The traffic classified weekend PNC trends are consistent with the weekend results for cold and warm months, where indoor concentrations were lower than outdoors. Similar trends were found for mean outdoor EC concentration between the high and low traffic schools. The differences in mean ambient
EC between the high and low traffic schools for the BCN were 47%, whereas the difference for the BNE was 30%. Differences between the mean ambient EC during school hours in the two cities were found to be as 86% for high traffic and 71% for low traffic schools. The observed PNC peaks between the two cities were compared by calculating the fraction of each PNC peak in relation to the total daily PNC for high and low traffic schools (Fig. 5). Mean concentrations of each identified peak are available in Table S1. As expected, based on PNC time-series (Figs. 1 and 2), the contribution of the mid-day peak was significantly higher for the BNE schools in comparison to the BCN in all cases, especially for indoors. Similar to the BNE results, but less pronounced, the mid-day peak made a greater contribution to indoor PNC than total traffic in the BCN schools, pointing to the significance of indoor sources. In terms of traffic, the morning rush hour (A.M. Traffic) made a larger contribution to total ambient and indoor PNC in both cities. The exception was for low traffic BCN schools, where evening rush hour (P.M. Traffic) made an equal or higher contribution to total PNC than the morning rush hour. In contrast to the BNE, the contribution of total daily traffic rush hours was generally higher than the mid-day peak for outdoor PNC in the BCN schools. Lower PNC levels, in general, together with much lower traffic counts and a smaller diesel fleet in Brisbane, favour new particle formation, which was responsible for the dominant peak during mid-day (Hameri et al., 1996; Ronkko et al., 2006; Hamed et al., 2007; Pey et al., 2008). In contrast to Brisbane, the dominant peaks in PNC timeseries in Barcelona were associated with traffic rush hours, given that
Fig. 6. Correlations between the ambient PNC and percentages of road network and green areas for each district/suburb in BNE (Brisbane, Australia) and BCN (Barcelona, Spain).
148
M. Mazaheri et al. / Environment International 88 (2016) 142–149
most of the BCN schools were located within busy urban areas or in the city centre of Barcelona. Higher PNC, higher traffic and a larger fleet density in proximity to the BCN schools facilitated the condensation sink and could have resulted in fewer mid-day new particle formation events. Furthermore, differences in the lowest particle size detection limit in PNC measurements between BNE and BCN should also be noted, as the mid-day peak particles are in the smaller size range (please see Section 2.2 Data source: 5 and 6 nm for Brisbane vs. 10 nm for Barcelona). These findings are supported by the recently published analysis of new particle formation characteristics in high insolation and subtropical areas, including Brisbane and Barcelona (Brines et al., 2015). District classification of the BNE and BCN schools found that the schools were located in 24 different districts/suburbs in both cities. Percentages of road network area and green area were calculated as a fraction of the total area for each identified district/suburb and their correlations with the corresponding mean daily PNC were determined (Fig. 6). For BCN schools, moderate correlations were found between mean daily PNC and road network percentages, while moderate anticorrelations were found with percentages of green areas (R2 = 0.3). No statistically significant correlations were found for BNE school districts/suburbs, where mean daily PNC variations across the suburbs were lower than that of BCN (stdev: 2.48 × 103 vs. 8.01 × 103 cm−3). The density of passenger cars in Barcelona was twice that of Brisbane during the study periods (Table 1), while the fraction of the total fleet using petrol was 77% in Brisbane and 64% in Barcelona and fractions for the diesel-fuel fleet were 21% for Brisbane and 35% for Barcelona. The percentage ratio of the total number of registered vehicles and road network area was 15% for Barcelona in comparison to 0.2% for Brisbane. These outcomes can be explained by the differences in urban and traffic characteristics between the two cities, as well as the inhomogeneous nature of PNC spatio-temporal distribution, which was highly affected by background conditions, as well as the major non-traffic sources of PNC (i.e. airport and port operations) (Mazaheri et al., 2011; Keuken et al., 2015). Overall, these results imply that the main driver of ambient PNC for schools in Barcelona was traffic, and for Brisbane schools, it was new particle formation during mid-day. The findings of this paper indicate that the high mid-day PNC peak in BNE schools could have been driven by the combined effect of background and meteorological conditions, as well as other local/distant sources, which subdued the PNC levels during traffic rush hours. These results have important implications for urban development, especially in terms of air quality mitigation and management at schools. This includes emission mitigation recommendations in regards to traffic emissions in general and especially in Barcelona; e.g. amending school's start and end times to avoid morning or evening traffic rush hours, ensuring that children's outdoor activities are outside traffic rush hours and using barrier or distance from the busy roads to disturb the direct and fast transport of emissions. Despite the costs, if all other mitigation strategies fail or are ineffective in addressing the issue, implementing heating, ventilation and air-conditioning (HVAC) and air filtering systems are also recommended. The variation in exposure to PNC levels may have different implications in respect to the associated health effects for children in the two cities. Part of the air quality data in Barcelona schools (BREATHE project) have already been published, reporting associations between exposure to higher traffic related air pollutants and smaller growth in cognitive development in school children (Sunyer et al., 2015). The PNC exposure–response relationships for the BNE children are being currently investigated as part of UPTECH project. Acknowledgements Mandana Mazaheri was supported by the 2015 Endeavour Fellowship funded by the Australian Government — Department of Education. UPTECH was supported by the Australian Research Council (ARC Linkage Grant LP0990134) and other funding organisations; i.e. QLD Department of Transport and Main Roads and QLD Department of
Education, Training and Employment; and BREATHE by the European Community Seventh Framework Program (ERC-Advanced Grant: 268479) and national projects IMPACT (CGL2011-26574), VAMOS (CLG2010-19464-CLI), CECAT (CTM2011-14730-E) and Generalitat de Catalunya 2015 SGR33. Authors thank Samuel Clifford, Md Mahmudur Rahman and Jayandana Karunasinghe from QUT and all the UPTECH and BREATHE team members.
Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.envint.2015.12.029.
References Ajuntament de Barcelona, 2015. Dades bàsiques de mobilitat. from http://prod-mobilitat. s3.amazonaws.com/DB_2012_compr.pdf. Australian Bureau of Meteorology, 2015. Climate statistics. from http://www.bom.gov. au/climate/averages/tables/cw_040913.shtml. Australian Bureau of Statistics, 2015. Brisbane Inner City statistics. from http://stat.abs.gov. au/itt/r.jsp?RegionSummary®ion=305&dataset=ABS_NRP9_ASGS&geoconcept= REGION&measure=MEASURE&datasetASGS=ABS_NRP9_ASGS&datasetLGA=ABS_ NRP9_LGA®ionLGA=REGION®ionASGS=REGION. Barcelona City Council, 2015. Figures by neighborhoods. from http://www.bcn.cat/ estadistica/catala/dades/barris/index.htm. Brines, M., Dall'Osto, M., Beddows, D.C.S., Harrison, R.M., Gómez-Moreno, F., Núñez, L., Artíñano, B., Costabile, F., Gobbi, G.P., Salimi, F., Morawska, L., Sioutas, C., Querol, X., 2015. Traffic and nucleation events as main sources of ultrafine particles in highinsolation developed world cities. Atmos. Chem. Phys. 15 (10), 5929–5945. Buonanno, G., Marini, S., Morawska, L., Fuoco, F.C., 2012. Individual dose and exposure of Italian children to ultrafine particles. Sci. Total Environ. 438, 271–277. Buonanno, G., Fuoco, F.C., Morawska, L., Stabile, L., 2013. Airborne particle concentrations at schools measured at different spatial scales. Atmos. Environ. 67, 38–45. Carslaw, D.C., Ropkins, K., 2012. Openair — an R package for air quality data analysis. Environ. Model. Softw. 27-28, 52–61. Cheung, H.C., Morawska, L., Ristovski, Z.D., 2011. Observation of new particle formation in subtropical urban environment. Atmos. Chem. Phys. 11 (8), 3823–3833. Cheung, H.C., Morawska, L., Ristovski, Z.D., Wainwright, D., 2012. Influence of medium range transport of particles from nucleation burst on particle number concentration within the urban airshed. Atmos. Chem. Phys. 12 (11), 4951–4962. Crilley, L.R., Ayoko, G.A., Mazaheri, M., Morawska, L., 2016. Factors influencing the outdoor concentration of carbonaceous aerosols at urban schools in Brisbane, Australia: implications for children's exposure. Environ. Pollut. 208, Part A, 249–255. Ezz, W., Mazaheri, M., Robinson, P., Johnson, G., Clifford, S., He, C., Morawska, L., Marks, G., 2015. Ultrafine Particles from Traffic Emissions and Children's Health (UPTECH) in Brisbane, Queensland (Australia): study design and implementation. Int. J. Environ. Res. Public Health 12 (2), 1687–1702. Guo, H., Morawska, L., He, C., Gilbert, D., 2008. Impact of ventilation scenario on air exchange rates and on indoor particle number concentrations in an air-conditioned classroom. Atmos. Environ. 42 (4), 757–768. Hamed, A., Joutsensaari, J., Mikkonen, S., Sogacheva, L., Dal Maso, M., Kulmala, M., Cavalli, F., Fuzzi, S., Facchini, M.C., Decesari, S., Mircea, M., Lehtinen, K.E.J., Laaksonen, A., 2007. Nucleation and growth of new particles in Po Valley, Italy. Atmos. Chem. Phys. 7, 355–376. Hameri, K., Kulmala, M., Aalto, P., Leszczynski, K., Visuri, R., Hamekoski, K., 1996. The investigations of aerosol particle formation in urban background area of Helsinki. Atmos. Res. 41 (3–4), 281–298. HEI, 2013. Understanding the Health Effects of Ambient Ultrafine Particles. HEI Perspectives 3 — HEI Review Panel on Ultrafine Particles. Health Effects Institute, Boston, Massachusetts. Hitchins, J., Morawska, L., Wolff, R., Gilbert, D., 2000. Concentrations of submicrometre particles from vehicle emissions near a major road. Atmos. Environ. 34 (1), 51–59. Hornung, R.W., Lanphear, B.P., Dietrich, K.N., 2009. Age of greatest susceptibility to childhood lead exposure: a new statistical approach. Environ. Health Perspect. 117 (8), 1309–1312. Keuken, M.P., Moerman, M., Zandveld, P., Henzing, J.S., Hoek, G., 2015. Total and sizeresolved particle number and black carbon concentrations in urban areas near Schiphol airport (The Netherlands). Atmos. Environ. 104, 132–142. Laiman, R., He, C., Mazaheri, M., Clifford, S., Salimi, F., Crilley, L.R., Megat Mokhtar, M.A., Morawska, L., 2014. Characteristics of ultrafine particle sources and deposition rates in primary school classrooms. Atmos. Environ. 94, 28–35. Mazaheri, M., Johnson, G.R., Morawska, L., 2011. An inventory of particle and gaseous emissions from large aircraft thrust engine operations at an airport. Atmos. Environ. 45 (20), 3500–3507. Mazaheri, M., Clifford, S., Jayaratne, R., Megat Mokhtar, M.A., Fuoco, F., Buonanno, G., Morawska, L., 2014. School children's personal exposure to ultrafine particles in the urban environment. Environ. Sci. Technol. 48 (1), 113–120. Mejía, J.F., Choy, S.L., Mengersen, K., Morawska, L., 2011. Methodology for assessing exposure and impacts of air pollutants in school children: data collection, analysis and health effects — a literature review. Atmos. Environ. 45 (4), 813–823.
M. Mazaheri et al. / Environment International 88 (2016) 142–149 Morawska, L., Bofinger, N.D., Kocis, L., Nwankwoala, A., 1998a. Submicrometer and supermicrometer particles from diesel vehicle emissions. Environ. Sci. Technol. 32 (14), 2033–2042. Morawska, L., Thomas, S., Bofinger, N., Wainwright, D.D., Neale, D., 1998b. Comprehensive characterization of aerosols in a subtropical urban atmosphere: particle size distribution and correlation with gaseous pollutants. Atmos. Environ. 32, 2467–2478. Pey, J., Rodriguez, S., Querol, X., Alastuey, A., Moreno, T., Putaud, J.P., Van Dingenen, R., 2008. Variations of urban aerosols in the western Mediterranean. Atmos. Environ. 42 (40), 9052–9062. Raysoni, A.U., Sarnat, J.A., Sarnat, S.E., Garcia, J.H., Holguin, F., Luèvano, S.F., Li, W.-W., 2011. Binational school-based monitoring of traffic-related air pollutants in El Paso, Texas (USA) and Ciudad Juárez, Chihuahua (México). Environ. Pollut. 159 (10), 2476–2486. Reche, C., Querol, X., Alastuey, A., Viana, M., Pey, J., Moreno, T., Rodríguez, S., González, Y., Fernández-Camacho, R., De La Campa, A.M.S., De La Rosa, J., Dall'Osto, M., Prévt, A.S.H., Hueglin, C., Harrison, R.M., Quincey, P., 2011a. New considerations for PM, black carbon and particle number concentration for air quality monitoring across different European cities. Atmos. Chem. Phys. 11 (13), 6207–6227. Reche, C., Viana, M., Moreno, T., Querol, X., Alastuey, A., Pey, J., Pandolfi, M., Prévôt, A., Mohr, C., Richard, A., Artiñano, B., Gomez-Moreno, F.J., Cots, N., 2011b. Peculiarities in atmospheric particle number and size-resolved speciation in an urban area in the western Mediterranean: results from the DAURE campaign. Atmos. Environ. 45 (30), 5282–5293. Reche, C., Viana, M., Rivas, I., Bouso, L., Alvarez-Pedrerol, M., Alastuey, A., Sunyer, J., Querol, X., 2014. Outdoor and indoor UFP in primary schools across Barcelona. Sci. Total Environ. 493, 943–953. Rivas, I., Viana, M., Moreno, T., Pandolfi, M., Amato, F., Reche, C., Bouso, L., AlvarezPedrerol, M., Alastuey, A., Sunyer, J., Querol, X., 2014. Child exposure to indoor and outdoor air pollutants in schools in Barcelona, Spain. Environ. Int. 69, 200–212. Ronkko, T., Virtanen, A., Vaaraslahti, K., Keskinen, J., Pirjola, L., Lappi, M., 2006. Effect of dilution conditions and driving parameters on nucleation mode particles in diesel exhaust: laboratory and on-road study. Atmos. Environ. 40 (16), 2893–2901.
149
Salimi, F., Mazaheri, M., Clifford, S., Crilley, L.R., Laiman, R., Morawska, L., 2013. Spatial variation of particle number concentration in school microscale environments and its impact on exposure assessment. Environ. Sci. Technol. 47 (10), 5251–5258. Salimi, F., Ristovski, Z., Mazaheri, M., Laiman, R., Crilley, L.R., He, C., Clifford, S., Morawska, L., 2014. Assessment and application of clustering techniques to atmospheric particle number size distribution for the purpose of source apportionment. Atmos. Chem. Phys. 14 (21), 11883–11892. Sunyer, J., Esnaola, M., Alvarez-Pedrerol, M., Forns, J., Rivas, I., Lopez-Vicente, M., SuadesGonzalez, E., Foraster, M., Garcia-Esteban, R., Basagana, X., Viana, M., Cirach, M., Moreno, T., Alastuey, A., Sebastian-Galles, N., Nieuwenhuijsen, M., Querol, X., 2015. Association between traffic-related air pollution in schools and cognitive development in primary school children: a prospective cohort study. PLoS Med. 12 (3), e1001792. Tippayawong, N., Khuntong, P., Nitatwichit, C., Khunatorn, Y., Tantakitti, C., 2009. Indoor/ outdoor relationships of size-resolved particle concentrations in naturally ventilated school environments. Build. Environ. 44 (1), 188–197. Trasande, L., Thurston, G.D., 2005. The role of air pollution in asthma and other pediatric morbidities. J. Allergy Clin. Immunol. 115 (4), 689–699. Viana, M., Rivas, I., Querol, X., Alastuey, A., Sunyer, J., Alvarez-Pedrerol, M., Bouso, L., Sioutas, C., 2014. Indoor/outdoor relationships and mass closure of quasi-ultrafine, accumulation and coarse particles in Barcelona schools. Atmos. Chem. Phys. 14 (9), 4459–4472. Wang, M., Beelen, R., Basagana, X., Becker, T., Cesaroni, G., de Hoogh, K., Dedele, A., Declercq, C., Dimakopoulou, K., Eeftens, M., Forastiere, F., Galassi, C., Grazuleviciene, R., Hoffmann, B., Heinrich, J., Iakovides, M., Kunzli, N., Korek, M., Lindley, S., Molter, A., Mosler, G., Madsen, C., Nieuwenhuijsen, M., Phuleria, H., Pedeli, X., RaaschouNielsen, O., Ranzi, A., Stephanou, E., Sugiri, D., Stempfelet, M., Tsai, M.Y., Lanki, T., Udvardy, O., Varro, M.J., Wolf, K., Weinmayr, G., Yli-Tuomi, T., Hoek, G., Brunekreef, B., 2013. Evaluation of land use regression models for NO2 and particulate matter in 20 European study areas: the ESCAPE project. Environ. Sci. Technol. 47 (9), 4357–4364. Zhang, Q., Zhu, Y., 2012. Characterizing ultrafine particles and other air pollutants at five schools in south Texas. Indoor Air 22 (1), 33–42.
Contents lists available at ScienceDirect
Environment International journal homepage: www.elsevier.com/locate/envint
Variability in exposure to ambient ultrafine particles in urban schools: Comparative assessment between Australia and Spain Mandana Mazaheri a,b, Cristina Reche c, Ioar Rivas c,d, Leigh R. Crilley e, Mar Álvarez-Pedrerol d, Mar Viana c, Aurelio Tobias c, Andrés Alastuey c, Jordi Sunyer d, Xavier Querol c, Lidia Morawska a,b,⁎ a
International Laboratory for Air Quality and Health, Queensland University of Technology, Brisbane, Australia Institute for Health and Biomedical Innovation, Queensland University of Technology, Brisbane, Australia Institute of Environmental Assessment and Water Research, Spanish National Research Council Barcelona, Spain d Centre for Research in Environmental Epidemiology, Barcelona, Spain e School of Geography, Earth and Environmental Sciences, University of Birmingham, Birmingham B15 2TT, UK b c
a r t i c l e
i n f o
Article history: Received 31 August 2015 Received in revised form 23 November 2015 Accepted 20 December 2015 Available online 4 January 2016 Keywords: Ambient particles Exposure New particle formation Urban environments High insolation areas
a b s t r a c t Ambient ultrafine particle number concentrations (PNC) have inhomogeneous spatio-temporal distributions and depend on a number of different urban factors, including background conditions and distant sources. This paper quantitatively compares exposure to ambient ultrafine particles at urban schools in two cities in developed countries, with high insolation climatic conditions, namely Brisbane (Australia) and Barcelona (Spain). The analysis used comprehensive indoor and outdoor air quality measurements at 25 schools in Brisbane and 39 schools in Barcelona. PNC modes were analysed with respect to ambient temperature, land use and urban characteristics, combined with the measured elemental carbon concentrations, NOx (Brisbane) and NO2 (Barcelona). The trends and modes of the quantified weekday average daily cycles of ambient PNC exhibited significant differences between the two cities. PNC increases were observed during traffic rush hours in both cases. However, the midday peak was dominant in Brisbane schools and had the highest contribution to total PNC for both indoors and outdoors. In Barcelona, the contribution from traffic was highest for ambient PNC, while the mid-day peak had a slightly higher contribution for indoor concentrations. Analysis of the relationships between PNC and land use characteristics in Barcelona schools showed a moderate correlation with the percentage of road network area and an anti-correlation with the percentage of green area. No statistically significant correlations were found for Brisbane. Overall, despite many similarities between the two cities, school-based exposure patterns were different. The main source of ambient PNC at schools was shown to be traffic in Barcelona and mid-day new particle formation in Brisbane. The mid-day PNC peak in Brisbane could have been driven by the combined effect of background and meteorological conditions, as well as other local/distant sources. The results have implications for urban development, especially in terms of air quality mitigation and management at schools. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Children's exposure to air pollution, along with characterization of their exposure–response relationship has been of great interest recently. A simple search on the Web of Science using the combined keywords “children”, “air pollution”, “exposure”, “epidemiology” and “schools” results in more than 80 publications in the past 15 years, with more than 30 of them published since 2010 (within the past 5 years). Children are known to be more physiologically susceptible to air pollution and other environmental related health risks than adults, since their organs are still developing and they have higher levels of physical activity and ⁎ Corresponding author at: International Laboratory for Air Quality and Health, Institute of Health and Biomedical Innovation, Queensland University of Technology, GPO Box 2434, Brisbane, QLD 4001, Australia. E-mail address: [email protected] (L. Morawska).
http://dx.doi.org/10.1016/j.envint.2015.12.029 0160-4120/© 2015 Elsevier Ltd. All rights reserved.
breathing rates in comparison to adults (Trasande and Thurston, 2005; Hornung et al., 2009). In 2013, the Health Effects Institute (HEI) identified ultrafine particles (UFPs) as one of the air pollutants of most concern, due to a lack of sufficient studies on their impact on human health (HEI, 2013). However, until 2011, most of the efforts in relation to school-based air quality studies were focused on particulate matter in terms of particle mass and on the PM2.5 and PM10 size ranges (Mejía et al., 2011). Since the Mejía et al. (2011) literature review, some comprehensive studies on UFP characteristics at schools have been conducted around the world, such as in Australia (Ultrafine Particles from Traffic Emissions and Children's Health: UPTECH) (Salimi et al., 2013; Laiman et al., 2014; Mazaheri et al., 2014; Crilley et al., 2016; Ezz et al., 2015), Spain (BRain dEvelopment and Air polluTion ultrafine particles in scHool childrEn: BREATHE) (Reche et al., 2014; Rivas et al., 2014; Viana et al., 2014; Sunyer et al., 2015), Italy (Buonanno et al., 2012, 2013), and the US (Zhang and Zhu, 2012). Indoor concentrations
M. Mazaheri et al. / Environment International 88 (2016) 142–149
of UFP are driven by the infiltration of outdoor particles (especially for naturally ventilated school classrooms), classroom activities during teaching hours, and cleaning before and after school hours (Guo et al., 2008; Tippayawong et al., 2009; Raysoni et al., 2011; Laiman et al., 2014; Reche et al., 2014; Viana et al., 2014). Outdoor sources include a wide range of local and distant sources, such as cooking within the school premises, vehicular traffic (Morawska et al., 1998a,b, Hitchins et al., 2000) or new particle formation mechanisms (Reche et al., 2011a,b). These studies have pointed to the important contribution of environmental factors specific to each study area on UPF characteristics and air pollution in general, in both indoor (teaching classrooms) and outdoor (e.g. playgrounds) school microenvironments. This work uses the large body of air quality data that were collected within the scope of the UPTECH (in Brisbane, Australia) and BREATHE (in Barcelona, Spain) projects and provides a comparative assessment of the differences and similarities in school-based exposure to ambient particle number concentrations (PNC) in high insolation environments with different urban profiles (e.g. urban, fleet and air pollution characteristics). 2. Materials and methods 2.1. Study area The study areas were Brisbane, Australia and Barcelona, Spain, which share a range of similar features, particularly climatic conditions, but with different traffic and urban development characteristics (Table 1). The statistics provided in the table are average values for Brisbane and Barcelona, which were current at the time of the UPTECH (2010–2012) and BREATHE (2012–2013) projects. 2.2. Data source The data used for this study were available from the UPTECH and BREATHE project databases: • The UPTECH data were collected at 3 outdoor and 2 indoor sites (naturally ventilated teaching classrooms) in 25 schools within the Brisbane Metropolitan Area, from October 2010 to August 2012. The UPTECH indoor data used in this study were averaged over the two classrooms, as no major differences were observed between
Table 1 General urban and air quality characteristics in Brisbane and Barcelona. General urban and air quality statistics
Brisbane, Australia
Barcelona, Spain
Population1,2 Total fleet density (km−2)1,2,3 Total passenger vehicles density (km−2)1,2,3 % of passenger vehicles to total registered vehicles1,2,3 % of motor vehicles to total registered vehicles — petrol1,3 % of motor vehicles to total registered vehicles — diesel1,3 % of motor vehicles to total registered vehicles — LPG/dual/other1,3 Maximum temperature (°C)4,5 Minimum temperature (°C)4,5 Relative humidity (%)4,5 Daily mean ambient PNC during weekdays (cm−3)6 Daily mean minimum and maximum traffic counts6 Total % of green area7,2 Total % of residential area7,2 Total % of roads7,2
241,264 3757 2810 75 77 21 2
1,602,386 9050 5800 62 64 35 1
21.9–30.2 10.1–21.5 65 8985 34–1063 50 17 5
19.2–32.3 3–18.5 65 20,548 2,630–54,522 28 25 23
1 2 3 4 5 6 7
(Australian Bureau of Statistics). (Barcelona City Council). https://sedeapl.dgt.gob.es/IEST2. (Australian Bureau of Meteorology). Faculty of Physics from Barcelona University. Averaged values during the monitoring campaigns. QLD Government (https://data.qld.gov.au/).
143
classrooms at each school (Laiman et al., 2014) (https://www.qut. edu.au/research/research-projects/uptech). • The BREATHE data were collected at one indoor (naturally ventilated teaching classroom) and one outdoor location in 36 schools in Barcelona, as well as 3 in Sant Cugat del Vallès (an adjacent municipality located 5 km away from Barcelona), through two separate campaigns conducted between 27 January and 22 June 2012, and 14 September 2012 and 22 February 2013 (http://www.creal.cat/projectebreathe). The data from both campaigns were used in this paper.
Considering the study designs for both projects, vehicular traffic was expected to be the main source of air pollution in the vicinity of the schools. General characteristics of the studied schools in the two cities are summarised in Table 2. For the UPTECH, air quality parameters were measured at one school at a time for 2 consecutive weeks at each school, and for the BREATHE, they were measured at two schools at a time for one week. Parameters that were measured during weekdays (school days) and were common between the two projects or relevant to support the data analysis were used for this paper. These were: concurrently measured 24-h time-series of indoor and outdoor PNC at each school; offline measurements of outdoor elemental carbon (EC) concentrations in PM2.5 during school hours; passive measurements of outdoor NO2 concentrations during school hours for the BREATHE campaign; and time-series of outdoor NOx concentrations for the UPTECH (time-series of NO2 concentrations were not available for the whole UPTECH campaign). Meteorological conditions (temperature, humidity and solar radiation) for the UPTECH were measured at a central outdoor location at each school and at a reference urban location in Barcelona for the BREATHE. Realtime data were collected in 30 s sampling intervals for the UPTECH and 10 min sampling intervals for BREATHE. Local traffic counts in proximity of each school were available with 5 min resolutions for the UPTECH schools, while traffic counts for the BREATHE schools were carried out over 15 min, for two of the sampling days during morning traffic rush hours (09:15 CET approx.). Detailed information on the general aspects, design, instrumentation and data quality controls for each project are available in the previously published papers for the UPTECH (Salimi et al., 2013; Laiman et al., 2014; Crilley et al., 2016) and BREATHE (Reche et al., 2014; Rivas et al., 2014). This paper uses the hourly averaged data that were measured during weekdays for both of the projects. Some assumptions and compromises were made when analysing the data, mainly due to differences between the UPTECH and BREATHE study designs. The BREATHE used DiScmini at stationary sites for indoor and outdoor particle monitoring, measuring concentrations and average sizes of particles in the 10–700 nm range; whereas the UPTECH employed TSI condensation particle counters (CPCs) (models 3787 and 3781), measuring total concentrations of particles larger than 5 and 6 nm, respectively. Comparisons of the PNC characteristics for both projects were made based on the DiScmini Table 2 General school characteristics at the time of measurement campaigns.
School age (years) Floor type Schooling hours Classroom cleaning schedule Building material Traffic rush hours Window frames Teaching equipment
Classroom furniture
Brisbane, Australia
Barcelona, Spain
40–146 Carpet 09:00–15:00 05:00–09:00 and 15:00–18:00
7–9 Ceramic tiles 09:00–17:00 5:00–9:00 and 17:00–20:00
Bricks, concrete, wood 07:00–08:00 and 17:00–18:00 Wood and/or aluminium Whiteboard, desktop computer/s, overhead projector Standard school tables and chairs, bookshelves
Bricks, concrete, wood 8:00–9:00 and 20:00–21:00 Wood/PVC/aluminium Blackboard, overhead projector, desktop computer/s Standard school tables and chairs, bookshelves
144
M. Mazaheri et al. / Environment International 88 (2016) 142–149
and CPC measurements; even though they had somewhat different specifications for detecting and counting particles. This method was considered practical, given that both study areas were within urban settings in developed countries, with similar air pollution sources and PNCs in the UFP size range (b0.1 μm) (Morawska et al., 1998a,b; Reche et al., 2014; Salimi et al., 2014). In addition, this paper is not focused on a direct comparison of PNC absolute values between the two study areas; rather it compares the trends and influencing factors. The ambient air quality parameters measured at one outdoor location at each school were used for the analysis. In the case of the UPTECH, the outdoor data were taken from a centrally located outdoor location at each school. Outdoor data was only available for one location at each school for the BREATHE. This approach does not take into account the potential impact of PNC spatial variation within the school grounds. However, it was considered redundant, since PNC spatial variation was not observed for the majority of the UPTECH schools studied (Salimi et al., 2013). For simplicity, BNE is used wherever referring to the UPTECH schools in Brisbane, Australia and BCN, wherever referring to the BREATHE schools in Barcelona, Spain. 3. Data analysis 3.1. Daily cycles of the air quality parameters The mean daily cycles of indoor and outdoor PNCs (BNE and BCN), local traffic counts and outdoor NOx (BNE) were quantified for each school using the measured time-series, in order to analyse general PNC trends throughout the day (including traffic rush hours and during mid-day). Diurnal cycles were calculated in R using the openair package (Carslaw and Ropkins, 2012). In addition, the contribution of each mode to total PNC was estimated by calculating the sum of PNC for each mode as a percentage of the sum of total PNC during the day.
selected as the proxy for general climatic conditions. Schools were grouped according to their temperature ranges across the year, rather than based on the conventional definition of seasons. This is of most importance in Brisbane (BNE), where there is no conventional change of season and seasonal variation mainly consists of warmer (November to April) and cooler (May–October) months. In the case of Barcelona (BCN), December, January and February are the coldest months, while the warm period lasts about six months, from May to October, with July and August being the warmest months. PNC daily cycles were analysed with respect to the school campaigns conducted in the warmer and cooler months, in order to evaluate the impact of weather conditions throughout the year. 3.2.3. Urban characteristics Trends in mean ambient PNC time-series between schools in the two cities were compared against each city's urban characteristics; namely total built-up area, total number of registered vehicles and total number of vehicles using petrol and diesel fuels (Table 1). Furthermore, the influence of general urban characteristics, in terms of total road network and green areas, on mean ambient PNC was estimated according to local urban characteristics for the BNE and BCN schools. Open access datasets provided by the Queensland Government (https://data.qld. gov.au/) and the Barcelona City Council (http://www.bcn.cat/ estadistica/catala/dades/barris/index.htm) were employed for land use and locality analysis. For Brisbane, green areas comprised of the sum of data subsets for “recreation and culture”, “residual native cover”, “grazing native vegetation”, “other conserved area” and “other forest plantation”. In Barcelona, green areas were defined as the urban and forest parks. 4. Results and discussions 4.1. Temporal variations of the air quality parameters
3.2. Classification of the schools The impact of traffic on air pollution, as well as ambient climatic factors, on the diurnal trends of indoor and outdoor PNC, was determined using the relevant measured environmental parameters as signatures. The schools were grouped accordingly, as described in the following subsections: 3.2.1. Traffic conditions Instead of the local traffic counts, school hour EC, NO2 and NOx that were measured at the same time and location as outdoor PNC were used as proxies to determine the schools that were most likely affected by traffic related air pollution (Wang et al., 2013). This approach was used since traffic counts alone cannot explain the air quality levels, and the dispersion of traffic related emissions heavily depend on the local meteorological conditions (e.g. wind direction and speed) and site characteristics (e.g. topography, distant from the road, barriers, green space, orientation of the building). Classification analysis was performed based on the 50th percentile of the average concentrations of NO2 (for BCN) and NOx (for BNE) during school hours across all the schools. Schools with average concentrations higher than the 50th percentile were grouped as high traffic schools and the remainder was grouped as low traffic schools. Corresponding school hour mean EC concentrations for the high and low traffic schools and their trends were compared with mean PNC, since EC is the main component of traffic related particles. 3.2.2. Ambient climatic conditions The studied classrooms were all naturally ventilated. Thermal comfort in the classrooms was maintained by opening or closing the windows, as a result of variations in outdoor temperature and the thermal comfort preference of the occupants. Therefore, temperature was
Daily cycles of hourly averaged indoor and outdoor PNC for all BNE and BCN schools are available in the supplementary information (SI) document (Figs. S1 and S2). Two peaks, coinciding with the morning and evening traffic rush hours, were observed for all average PNC time-series in the two cities. At each individual BNE school, all average indoor PNC time-series and almost all outdoor PNC time-series exhibited an additional peak during mid-day. Indoor mean hourly PNCs were higher than for outdoors for more than half of the BNE schools. Trends in school-based average PNC daily cycles for the BCN were similar to the BNE; however the mid-day occurrence was less frequent and indoor PNCs were generally lower than outdoors. Higher indoor mid-day PNC was also observed for BCN schools, although with a much lower frequency than BNE. Fig. 1 presents mean daily cycles of the hourly averaged indoor and outdoor PNCs over all the schools for each project. The solid lines are the PNC and the shaded areas show the 95% confidence intervals. The corresponding traffic counts and NOx daily cycles, along with their 95% confidence intervals are shown in the SI document for BNE (Fig. S3). BNE NOx and traffic counts at the closest busy road to the school showed similar trends in terms of morning and evening traffic rush hours: 07:00–08:00 and 17:00–18:00. Traffic rush hours in Barcelona have been reported as 8:00–9:00 and 20:00–21:00 (Ajuntament de Barcelona, 2015). Three PNC peaks were observed for school-based indoor and outdoor time-series in both cities, corresponding to the morning and evening traffic rush hours and a mid-day peak (Fig. 1). The mid-day PNC mode for Brisbane (BNE) schools was around 11:00 and 14:00, which coincided with a small and rather insignificant peak in corresponding traffic counts and NOx, in comparison to the rush hour peaks (SI file, Fig. S3). This mid-day peak for Barcelona (BCN) schools was much smaller than the ones attributed to traffic rush hours.
M. Mazaheri et al. / Environment International 88 (2016) 142–149
145
Fig. 1. Daily cycles of the hourly averaged indoor and outdoor PNCs (cm−3) over all the BNE and BCN schools. Shaded areas represent 95% confidence intervals.
The results demonstrate clear differences between indoor and outdoor PNC daily cycles in the two cities. The major differences between PNC daily cycles for the two cities were: (a) in the magnitude and proportion of the mid-day peak in respect to the traffic rush hours for each city, which is further analysed in the following sub-section (Section 4.2); and (b) the mid-day PNC peak was more pronounced for indoors than outdoors in the BNE schools in comparison to BCN. Previous studies on ambient PNC increases during mid-day in Brisbane and Barcelona have attributed this peak to nucleation processes and new particle formation (Cheung et al., 2011, 2012; Reche et al., 2011a,b, 2014; Brines et al., 2015). In order to further investigate indoor PNC peaks in respect to ambient climatic conditions, the schools were grouped according to whether the campaign was conducted during the warmer or cooler months of the year (as described in Section 3.2) (Fig. 2).
4.1.1. UPTECH (BNE) Daily cycles of the hourly averaged indoor and outdoor PNCs for all BNE schools showed a small increase in indoor PNC before the start of teaching hours (around 07:00), which were lower than outdoor concentrations and were associated with early morning cleaning activities (05:00–09:00). Indoor PNCs were higher than outdoors between 08:00 and 18:00, corresponding with morning and evening cleaning activities (05:00–09:00 and 15:00–18:00) and teaching activities (09:00– 15:00). Comparing the daily PNC cycles for warmer and cooler months, BNE indoor PNCs generally exhibited a similar pattern to outdoors during the warmer months, when classrooms tended to have the windows open at all times during school hours and AERs were higher (Fig. 2). Previous analysis of indoor air quality for the BNE classrooms showed that the average air exchange rate during the warmer months was slightly higher than for the cooler months (0.6 vs. 0.5 h−1) (Laiman et al.,
Fig. 2. Mean daily cycles of indoor and outdoor PNCs (cm−3) during the cooler and warmer months for the BNE and BCN schools. Shaded areas represent 95% confidence intervals.
146
M. Mazaheri et al. / Environment International 88 (2016) 142–149
Fig. 3. Overall mean daily outdoor and indoor PNC (cm−3) for the identified high traffic (HT) and low traffic (LT) schools in BNE and BCN.
2014). During the warmer months, indoor PNCs were lower than outdoors before the start of each school day (09:00), gradually peaked above outdoor levels around 13:00 and then decreased to below outdoor levels at 17:00. In the cooler months, indoor PNC patterns differed during mid-day compared to the warmer months and on average, were 25% higher than outdoor levels. Indoor and outdoor PNC time-series measured during the weekend were used to further investigate the higher mid-day PNC peak indoors for BNE schools. During the weekends, classrooms were not in use, all of the doors and windows were shut, and no cleaning activities took place. Indoor PNC during mid-day was lower than outdoors during weekends for all of the BNE schools, as well as those grouped according to warmer and cooler months (SI file: Figs. S4 and S5). The results indicate that the prominent indoor mid-day peak was due to the combined effect of lower air exchange rates and potential indoor sources favouring new particle formation. It should be noted that the Brisbane schools did not have a dining hall, but a small stall-like shop (canteen) that served a variety of cold and warm foods. Furthermore, the canteens were not in close proximity to the studied classrooms and their immediate impact on indoor PNC during mid-day was implausible. Therefore, cooking related emissions were not considered as a driver for the observed mid-day peak in the average daily cycles.
hours. These peaks were more evident during the warmer period, especially for outdoor concentrations. Given that indoor PNCs were always lower than outdoors for hours outside of school activities in both cities (i.e. teaching and maintenance/cleaning hours), the higher indoor PNC indicated that indoor activities, cleaning methods and schedules were the most influential factors, which varied from one study area to another (Table 2). This was despite the fact that the classrooms were naturally ventilated and traffic density was generally higher for the BCN schools (Table 1). One of the significant findings of this research is the observed higher indoor versus outdoor PNC for UPTECH schools. Investigations in relation to identifying the specific source/s of the mid-day indoor PNC peaks require further data on the composition and size distribution of the indoor PNC and are the topic of our future research. Cluster analysis of the ambient/outdoor particle number size distributions (PNSD) in the BNE schools that were measured during the UPTECH project identified five clusters associated with three main particle sources, namely regional background, photochemically induced particles and traffic related particles (Salimi et al., 2014). For BCN schools, mean hourly indoor and outdoor particle sizes measured by the DiScmini were smaller during mid-day (SI file, Fig. S6). These outcomes support this work's findings, even though indoor PNSDs were not available for BNE schools.
4.1.2. BREATHE (BCN) Indoor PNC patterns were very similar to those found for outdoors, indicating that outdoor PNC could be responsible for indoor PNC variations. Similar to BNE, this was due to natural ventilation in the classrooms, as a result of opening windows in order to attain comfortable room temperatures. However, it is important to note that two main differences, namely: (i) increases in indoor PNC during morning rush hours usually occurred with a delay, coinciding with the beginning of school hours; and (ii) indoor PNC increases in the late afternoon/ evening were observed at around 17:00–19:00, not coinciding with outdoor peaks. This was associated with the cleaning of classrooms, which occurred once a day, mostly in the late afternoon/early evening (around 17:00–20:00) (Reche et al., 2014). In contrast to BNE schools, average indoor PNC daily cycles for the BCN showed that indoor concentrations were always lower than the outdoors by at least 10%, when averaged over all schools, as well as warmer or cooler months (Figs. 1 and 2). BCN schools were much newer than BNE schools (Table 2). Although building structure and age could be a defining factor when discussing the PNC I/O ratios, this could not be the case for the BCN indoor PNC being lower than outdoors. This is because all the classrooms were naturally ventilated and had windows opened, and therefore the differences in the effect of particle infiltration due to the gaps/leaks in the structure would have been negligible in this case. Small mid-day PNC peaks were observed for both indoors and outdoors at all schools, but they were generally lower than those observed during traffic rush
4.2. Influence of urban characteristics on ambient air quality Using the 50th percentile of NOx and NO2 concentrations as proxies for traffic related emissions, 13 out of the 25 BNE schools and 20 out of the 39 BCN schools were identified as high traffic schools (i.e. more likely to be affected by traffic related air pollution) and the remainder as low traffic schools. Mean school hour NOx for the high and low traffic BNE schools were 24 ppb and 10 ppb, respectively. In case of the BCN, mean NO2 was 28 ppb for the high traffic schools and 17 ppb for low traffic schools.
Fig. 4. Box plot of the overall mean outdoor EC (μg m−3) during school hours for the identified high traffic (HT) and low traffic (LT) schools in BNE and BCN.
M. Mazaheri et al. / Environment International 88 (2016) 142–149
147
Fig. 5. Fractions of the identified PNC peaks to the total mean daily PNC in BNE and BCN schools; H-O = High Traffic-Outdoors, H-I = High Traffic-Indoors, L-O = Low Traffic-Outdoors, L-I = Low Traffic-Indoors, A.M. Traffic = morning traffic rush hour, P.M. Traffic = evening traffic rush hour.
The school-based mean daily cycles of PNC for the BNE schools showed similar trends, where PNC peaked during traffic rush hours (Fig. S1). Similar trends were observed for the BCN PNC time-series (Fig. S2). Box plots of the mean indoor and outdoor daily PNC and outdoor school hour EC for the high and low traffic schools are presented in Figs. 3 and 4, respectively. Overall, significantly higher PNC and EC levels were observed in the BCN schools, and mean indoor and outdoor PNCs were found to be higher for high traffic schools. This was more evident for BCN, where the difference in outdoor PNC between the high and low traffic schools was 54% and the indoor difference was 38%. Significant differences (N10%) in the mean PNCs were not observed for the high and low traffic BNE schools. Comparison of the mean daily PNC between the high and low traffic schools in each city showed higher concentrations in BCN schools than in BNE. For high traffic schools, the differences between BCN and BNE were found as 89% for outdoors and 47% for indoors. For the low traffic BCN and BNE schools, the differences between mean daily PNC were 47% for outdoors and 11% for indoors. These results in terms of mean PNC daily cycles during weekdays for indoor and outdoor PNCs (cm−3) in the high and low traffic schools in the two cities are available in the SI file (Fig. S7). The corresponding weekend trends for the BNE schools are also available in the SI file (Fig. S8, weekend data are not available for BCN). The traffic classified weekend PNC trends are consistent with the weekend results for cold and warm months, where indoor concentrations were lower than outdoors. Similar trends were found for mean outdoor EC concentration between the high and low traffic schools. The differences in mean ambient
EC between the high and low traffic schools for the BCN were 47%, whereas the difference for the BNE was 30%. Differences between the mean ambient EC during school hours in the two cities were found to be as 86% for high traffic and 71% for low traffic schools. The observed PNC peaks between the two cities were compared by calculating the fraction of each PNC peak in relation to the total daily PNC for high and low traffic schools (Fig. 5). Mean concentrations of each identified peak are available in Table S1. As expected, based on PNC time-series (Figs. 1 and 2), the contribution of the mid-day peak was significantly higher for the BNE schools in comparison to the BCN in all cases, especially for indoors. Similar to the BNE results, but less pronounced, the mid-day peak made a greater contribution to indoor PNC than total traffic in the BCN schools, pointing to the significance of indoor sources. In terms of traffic, the morning rush hour (A.M. Traffic) made a larger contribution to total ambient and indoor PNC in both cities. The exception was for low traffic BCN schools, where evening rush hour (P.M. Traffic) made an equal or higher contribution to total PNC than the morning rush hour. In contrast to the BNE, the contribution of total daily traffic rush hours was generally higher than the mid-day peak for outdoor PNC in the BCN schools. Lower PNC levels, in general, together with much lower traffic counts and a smaller diesel fleet in Brisbane, favour new particle formation, which was responsible for the dominant peak during mid-day (Hameri et al., 1996; Ronkko et al., 2006; Hamed et al., 2007; Pey et al., 2008). In contrast to Brisbane, the dominant peaks in PNC timeseries in Barcelona were associated with traffic rush hours, given that
Fig. 6. Correlations between the ambient PNC and percentages of road network and green areas for each district/suburb in BNE (Brisbane, Australia) and BCN (Barcelona, Spain).
148
M. Mazaheri et al. / Environment International 88 (2016) 142–149
most of the BCN schools were located within busy urban areas or in the city centre of Barcelona. Higher PNC, higher traffic and a larger fleet density in proximity to the BCN schools facilitated the condensation sink and could have resulted in fewer mid-day new particle formation events. Furthermore, differences in the lowest particle size detection limit in PNC measurements between BNE and BCN should also be noted, as the mid-day peak particles are in the smaller size range (please see Section 2.2 Data source: 5 and 6 nm for Brisbane vs. 10 nm for Barcelona). These findings are supported by the recently published analysis of new particle formation characteristics in high insolation and subtropical areas, including Brisbane and Barcelona (Brines et al., 2015). District classification of the BNE and BCN schools found that the schools were located in 24 different districts/suburbs in both cities. Percentages of road network area and green area were calculated as a fraction of the total area for each identified district/suburb and their correlations with the corresponding mean daily PNC were determined (Fig. 6). For BCN schools, moderate correlations were found between mean daily PNC and road network percentages, while moderate anticorrelations were found with percentages of green areas (R2 = 0.3). No statistically significant correlations were found for BNE school districts/suburbs, where mean daily PNC variations across the suburbs were lower than that of BCN (stdev: 2.48 × 103 vs. 8.01 × 103 cm−3). The density of passenger cars in Barcelona was twice that of Brisbane during the study periods (Table 1), while the fraction of the total fleet using petrol was 77% in Brisbane and 64% in Barcelona and fractions for the diesel-fuel fleet were 21% for Brisbane and 35% for Barcelona. The percentage ratio of the total number of registered vehicles and road network area was 15% for Barcelona in comparison to 0.2% for Brisbane. These outcomes can be explained by the differences in urban and traffic characteristics between the two cities, as well as the inhomogeneous nature of PNC spatio-temporal distribution, which was highly affected by background conditions, as well as the major non-traffic sources of PNC (i.e. airport and port operations) (Mazaheri et al., 2011; Keuken et al., 2015). Overall, these results imply that the main driver of ambient PNC for schools in Barcelona was traffic, and for Brisbane schools, it was new particle formation during mid-day. The findings of this paper indicate that the high mid-day PNC peak in BNE schools could have been driven by the combined effect of background and meteorological conditions, as well as other local/distant sources, which subdued the PNC levels during traffic rush hours. These results have important implications for urban development, especially in terms of air quality mitigation and management at schools. This includes emission mitigation recommendations in regards to traffic emissions in general and especially in Barcelona; e.g. amending school's start and end times to avoid morning or evening traffic rush hours, ensuring that children's outdoor activities are outside traffic rush hours and using barrier or distance from the busy roads to disturb the direct and fast transport of emissions. Despite the costs, if all other mitigation strategies fail or are ineffective in addressing the issue, implementing heating, ventilation and air-conditioning (HVAC) and air filtering systems are also recommended. The variation in exposure to PNC levels may have different implications in respect to the associated health effects for children in the two cities. Part of the air quality data in Barcelona schools (BREATHE project) have already been published, reporting associations between exposure to higher traffic related air pollutants and smaller growth in cognitive development in school children (Sunyer et al., 2015). The PNC exposure–response relationships for the BNE children are being currently investigated as part of UPTECH project. Acknowledgements Mandana Mazaheri was supported by the 2015 Endeavour Fellowship funded by the Australian Government — Department of Education. UPTECH was supported by the Australian Research Council (ARC Linkage Grant LP0990134) and other funding organisations; i.e. QLD Department of Transport and Main Roads and QLD Department of
Education, Training and Employment; and BREATHE by the European Community Seventh Framework Program (ERC-Advanced Grant: 268479) and national projects IMPACT (CGL2011-26574), VAMOS (CLG2010-19464-CLI), CECAT (CTM2011-14730-E) and Generalitat de Catalunya 2015 SGR33. Authors thank Samuel Clifford, Md Mahmudur Rahman and Jayandana Karunasinghe from QUT and all the UPTECH and BREATHE team members.
Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.envint.2015.12.029.
References Ajuntament de Barcelona, 2015. Dades bàsiques de mobilitat. from http://prod-mobilitat. s3.amazonaws.com/DB_2012_compr.pdf. Australian Bureau of Meteorology, 2015. Climate statistics. from http://www.bom.gov. au/climate/averages/tables/cw_040913.shtml. Australian Bureau of Statistics, 2015. Brisbane Inner City statistics. from http://stat.abs.gov. au/itt/r.jsp?RegionSummary®ion=305&dataset=ABS_NRP9_ASGS&geoconcept= REGION&measure=MEASURE&datasetASGS=ABS_NRP9_ASGS&datasetLGA=ABS_ NRP9_LGA®ionLGA=REGION®ionASGS=REGION. Barcelona City Council, 2015. Figures by neighborhoods. from http://www.bcn.cat/ estadistica/catala/dades/barris/index.htm. Brines, M., Dall'Osto, M., Beddows, D.C.S., Harrison, R.M., Gómez-Moreno, F., Núñez, L., Artíñano, B., Costabile, F., Gobbi, G.P., Salimi, F., Morawska, L., Sioutas, C., Querol, X., 2015. Traffic and nucleation events as main sources of ultrafine particles in highinsolation developed world cities. Atmos. Chem. Phys. 15 (10), 5929–5945. Buonanno, G., Marini, S., Morawska, L., Fuoco, F.C., 2012. Individual dose and exposure of Italian children to ultrafine particles. Sci. Total Environ. 438, 271–277. Buonanno, G., Fuoco, F.C., Morawska, L., Stabile, L., 2013. Airborne particle concentrations at schools measured at different spatial scales. Atmos. Environ. 67, 38–45. Carslaw, D.C., Ropkins, K., 2012. Openair — an R package for air quality data analysis. Environ. Model. Softw. 27-28, 52–61. Cheung, H.C., Morawska, L., Ristovski, Z.D., 2011. Observation of new particle formation in subtropical urban environment. Atmos. Chem. Phys. 11 (8), 3823–3833. Cheung, H.C., Morawska, L., Ristovski, Z.D., Wainwright, D., 2012. Influence of medium range transport of particles from nucleation burst on particle number concentration within the urban airshed. Atmos. Chem. Phys. 12 (11), 4951–4962. Crilley, L.R., Ayoko, G.A., Mazaheri, M., Morawska, L., 2016. Factors influencing the outdoor concentration of carbonaceous aerosols at urban schools in Brisbane, Australia: implications for children's exposure. Environ. Pollut. 208, Part A, 249–255. Ezz, W., Mazaheri, M., Robinson, P., Johnson, G., Clifford, S., He, C., Morawska, L., Marks, G., 2015. Ultrafine Particles from Traffic Emissions and Children's Health (UPTECH) in Brisbane, Queensland (Australia): study design and implementation. Int. J. Environ. Res. Public Health 12 (2), 1687–1702. Guo, H., Morawska, L., He, C., Gilbert, D., 2008. Impact of ventilation scenario on air exchange rates and on indoor particle number concentrations in an air-conditioned classroom. Atmos. Environ. 42 (4), 757–768. Hamed, A., Joutsensaari, J., Mikkonen, S., Sogacheva, L., Dal Maso, M., Kulmala, M., Cavalli, F., Fuzzi, S., Facchini, M.C., Decesari, S., Mircea, M., Lehtinen, K.E.J., Laaksonen, A., 2007. Nucleation and growth of new particles in Po Valley, Italy. Atmos. Chem. Phys. 7, 355–376. Hameri, K., Kulmala, M., Aalto, P., Leszczynski, K., Visuri, R., Hamekoski, K., 1996. The investigations of aerosol particle formation in urban background area of Helsinki. Atmos. Res. 41 (3–4), 281–298. HEI, 2013. Understanding the Health Effects of Ambient Ultrafine Particles. HEI Perspectives 3 — HEI Review Panel on Ultrafine Particles. Health Effects Institute, Boston, Massachusetts. Hitchins, J., Morawska, L., Wolff, R., Gilbert, D., 2000. Concentrations of submicrometre particles from vehicle emissions near a major road. Atmos. Environ. 34 (1), 51–59. Hornung, R.W., Lanphear, B.P., Dietrich, K.N., 2009. Age of greatest susceptibility to childhood lead exposure: a new statistical approach. Environ. Health Perspect. 117 (8), 1309–1312. Keuken, M.P., Moerman, M., Zandveld, P., Henzing, J.S., Hoek, G., 2015. Total and sizeresolved particle number and black carbon concentrations in urban areas near Schiphol airport (The Netherlands). Atmos. Environ. 104, 132–142. Laiman, R., He, C., Mazaheri, M., Clifford, S., Salimi, F., Crilley, L.R., Megat Mokhtar, M.A., Morawska, L., 2014. Characteristics of ultrafine particle sources and deposition rates in primary school classrooms. Atmos. Environ. 94, 28–35. Mazaheri, M., Johnson, G.R., Morawska, L., 2011. An inventory of particle and gaseous emissions from large aircraft thrust engine operations at an airport. Atmos. Environ. 45 (20), 3500–3507. Mazaheri, M., Clifford, S., Jayaratne, R., Megat Mokhtar, M.A., Fuoco, F., Buonanno, G., Morawska, L., 2014. School children's personal exposure to ultrafine particles in the urban environment. Environ. Sci. Technol. 48 (1), 113–120. Mejía, J.F., Choy, S.L., Mengersen, K., Morawska, L., 2011. Methodology for assessing exposure and impacts of air pollutants in school children: data collection, analysis and health effects — a literature review. Atmos. Environ. 45 (4), 813–823.
M. Mazaheri et al. / Environment International 88 (2016) 142–149 Morawska, L., Bofinger, N.D., Kocis, L., Nwankwoala, A., 1998a. Submicrometer and supermicrometer particles from diesel vehicle emissions. Environ. Sci. Technol. 32 (14), 2033–2042. Morawska, L., Thomas, S., Bofinger, N., Wainwright, D.D., Neale, D., 1998b. Comprehensive characterization of aerosols in a subtropical urban atmosphere: particle size distribution and correlation with gaseous pollutants. Atmos. Environ. 32, 2467–2478. Pey, J., Rodriguez, S., Querol, X., Alastuey, A., Moreno, T., Putaud, J.P., Van Dingenen, R., 2008. Variations of urban aerosols in the western Mediterranean. Atmos. Environ. 42 (40), 9052–9062. Raysoni, A.U., Sarnat, J.A., Sarnat, S.E., Garcia, J.H., Holguin, F., Luèvano, S.F., Li, W.-W., 2011. Binational school-based monitoring of traffic-related air pollutants in El Paso, Texas (USA) and Ciudad Juárez, Chihuahua (México). Environ. Pollut. 159 (10), 2476–2486. Reche, C., Querol, X., Alastuey, A., Viana, M., Pey, J., Moreno, T., Rodríguez, S., González, Y., Fernández-Camacho, R., De La Campa, A.M.S., De La Rosa, J., Dall'Osto, M., Prévt, A.S.H., Hueglin, C., Harrison, R.M., Quincey, P., 2011a. New considerations for PM, black carbon and particle number concentration for air quality monitoring across different European cities. Atmos. Chem. Phys. 11 (13), 6207–6227. Reche, C., Viana, M., Moreno, T., Querol, X., Alastuey, A., Pey, J., Pandolfi, M., Prévôt, A., Mohr, C., Richard, A., Artiñano, B., Gomez-Moreno, F.J., Cots, N., 2011b. Peculiarities in atmospheric particle number and size-resolved speciation in an urban area in the western Mediterranean: results from the DAURE campaign. Atmos. Environ. 45 (30), 5282–5293. Reche, C., Viana, M., Rivas, I., Bouso, L., Alvarez-Pedrerol, M., Alastuey, A., Sunyer, J., Querol, X., 2014. Outdoor and indoor UFP in primary schools across Barcelona. Sci. Total Environ. 493, 943–953. Rivas, I., Viana, M., Moreno, T., Pandolfi, M., Amato, F., Reche, C., Bouso, L., AlvarezPedrerol, M., Alastuey, A., Sunyer, J., Querol, X., 2014. Child exposure to indoor and outdoor air pollutants in schools in Barcelona, Spain. Environ. Int. 69, 200–212. Ronkko, T., Virtanen, A., Vaaraslahti, K., Keskinen, J., Pirjola, L., Lappi, M., 2006. Effect of dilution conditions and driving parameters on nucleation mode particles in diesel exhaust: laboratory and on-road study. Atmos. Environ. 40 (16), 2893–2901.
149
Salimi, F., Mazaheri, M., Clifford, S., Crilley, L.R., Laiman, R., Morawska, L., 2013. Spatial variation of particle number concentration in school microscale environments and its impact on exposure assessment. Environ. Sci. Technol. 47 (10), 5251–5258. Salimi, F., Ristovski, Z., Mazaheri, M., Laiman, R., Crilley, L.R., He, C., Clifford, S., Morawska, L., 2014. Assessment and application of clustering techniques to atmospheric particle number size distribution for the purpose of source apportionment. Atmos. Chem. Phys. 14 (21), 11883–11892. Sunyer, J., Esnaola, M., Alvarez-Pedrerol, M., Forns, J., Rivas, I., Lopez-Vicente, M., SuadesGonzalez, E., Foraster, M., Garcia-Esteban, R., Basagana, X., Viana, M., Cirach, M., Moreno, T., Alastuey, A., Sebastian-Galles, N., Nieuwenhuijsen, M., Querol, X., 2015. Association between traffic-related air pollution in schools and cognitive development in primary school children: a prospective cohort study. PLoS Med. 12 (3), e1001792. Tippayawong, N., Khuntong, P., Nitatwichit, C., Khunatorn, Y., Tantakitti, C., 2009. Indoor/ outdoor relationships of size-resolved particle concentrations in naturally ventilated school environments. Build. Environ. 44 (1), 188–197. Trasande, L., Thurston, G.D., 2005. The role of air pollution in asthma and other pediatric morbidities. J. Allergy Clin. Immunol. 115 (4), 689–699. Viana, M., Rivas, I., Querol, X., Alastuey, A., Sunyer, J., Alvarez-Pedrerol, M., Bouso, L., Sioutas, C., 2014. Indoor/outdoor relationships and mass closure of quasi-ultrafine, accumulation and coarse particles in Barcelona schools. Atmos. Chem. Phys. 14 (9), 4459–4472. Wang, M., Beelen, R., Basagana, X., Becker, T., Cesaroni, G., de Hoogh, K., Dedele, A., Declercq, C., Dimakopoulou, K., Eeftens, M., Forastiere, F., Galassi, C., Grazuleviciene, R., Hoffmann, B., Heinrich, J., Iakovides, M., Kunzli, N., Korek, M., Lindley, S., Molter, A., Mosler, G., Madsen, C., Nieuwenhuijsen, M., Phuleria, H., Pedeli, X., RaaschouNielsen, O., Ranzi, A., Stephanou, E., Sugiri, D., Stempfelet, M., Tsai, M.Y., Lanki, T., Udvardy, O., Varro, M.J., Wolf, K., Weinmayr, G., Yli-Tuomi, T., Hoek, G., Brunekreef, B., 2013. Evaluation of land use regression models for NO2 and particulate matter in 20 European study areas: the ESCAPE project. Environ. Sci. Technol. 47 (9), 4357–4364. Zhang, Q., Zhu, Y., 2012. Characterizing ultrafine particles and other air pollutants at five schools in south Texas. Indoor Air 22 (1), 33–42.