Urban Climate 10 (2014) 608–620

Contents lists available at ScienceDirect

Urban Climate

journal homepage: www.elsevier .com/locate/ucl im

Microenvironment particle measurementsin Thessaloniki, Greece

http://dx.doi.org/10.1016/j.uclim.2014.03.0092212-0955/� 2014 Elsevier B.V. All rights reserved.

⇑ Corresponding author. Tel.: +30 2310 996014; fax: +30 2310 996019.E-mail address: zisis@auth.gr (Z. Samaras).

Ilias Vouitsis a, Pekka Taimisto b, Apostolos Kelessis c, Zissis Samaras a,⇑a Laboratory of Applied Thermodynamics, Aristotle University of Thessaloniki, Administration Building, University Campus, PO Box 458,GR-54124 Thessaloniki, Greeceb National Institute for Health and Welfare THL, PO Box 95, FI-70701 Kuopio, Finlandc Municipality of Thessaloniki, Department of Environment, Paparrigopoulou 7, GR-54630 Thessaloniki, Greece

a r t i c l e i n f o

Article history:Received 1 May 2013Revised 8 February 2014Accepted 27 March 2014

Keywords:Commuting modeTraffic conditionsExposureInhaled dose

a b s t r a c t

Monitoring of particulate concentrations in Thessaloniki, Greece,was carried out during April 2011, to assess differences in commut-ers’ exposure to traffic related particulate pollution. Three routeswere monitored in the two directions while bicycling, driving carand travelling by bus. The length of each route was about 8 kmand individual journey times ranged between 18 and 34 min. Cartrips were made with windows closed and with the ventilation sys-tem at moderate setting and with co-driver’s window open. Theresults indicate that mean inhalation doses while bicycling ishigher than those during travelling by bus (15% for PM, 55% forblack carbon and 40% for particle number) and by car (60% openwindow – 70% closed window for PM, 50% open window – 78%closed-window for black carbon and 54% open window – 77%closed window for PN). Individuals who change their travel modefrom car to bicycling and bus commuting in response to policiesaimed at encouraging a modal shift in travel behavior, are thuslikely to experience increased journey-time personal exposuresto traffic-related air pollution. Commuting by car with closed win-dows is the transport mode by which a person experiences theleast exposure to particulate pollution.

� 2014 Elsevier B.V. All rights reserved.

I. Vouitsis et al. / Urban Climate 10 (2014) 608–620 609

1. Introduction

Exposure to air pollution in traffic has been related to both long- and short-term cardiovascular andrespiratory health effects in a number of studies (Katsouyanni et al., 2011; Bernstein, 2012; Hoek et al.,2013; Chiu et al., 2013). Measurements from fixed monitoring stations have been commonly used assurrogates for personal exposure levels to represent community exposure to pollutants and they havebeen the basis of air quality guidelines and policy. However, several studies revealed that these mea-surements significantly underestimate or have little or no association with the exposure of populationsub groups, thus highlighting the need for direct personal exposure measurements (Loth and Ashmore,1994; Vellopoulou and Ashmore, 1998; Adams et al., 2001 and Adams et al., 2009; Gulliver and Briggs,2004; Lai et al., 2004; Peters et al., 2004; Riediker et al., 2004; Adar et al., 2007; McCreanor et al., 2007;Zuurbier et al., 2010). Particulates which are recognized as the main pollutant of concern in terms ofhuman health are of specific interest and short-term exposure (e.g. while commuting) to peak particleconcentrations may be associated with adverse health effects (Katsouyanni et al., 1997; Brook et al.,2011; Lim et al., 2012).

In the last decade, smaller devices for automatic monitoring of particulates at high temporal (i.e.several minutes to 1 h) have become available. The new generation of monitors are light and portablemaking them ideal for personal monitoring. Most of the studies looking at specific microenvironmentshave been done in transport and comparisons have generally been made in levels of concentrationsbetween different transport modes (e.g. walking, in-car, bus etc.) and between transport modes andfixed-site ambient monitoring. Up to now, published studies report high levels of particle concentra-tions inside public means of transport (Chan et al., 2002; Kaur et al., 2005) and that pedestrians andcyclists experience lower exposure concentrations than individuals inside vehicles (Kaur et al., 2007;Int Panis et al., 2010; Knibbs and de Dear, 2010; Geiss et al., 2010; Wang and Gao, 2011; Knibbs et al.,2011; Dons et al., 2012).

The findings indicate that particulate pollution varies largely from street to street and from city tocity. There are numerous variables potentially affecting personal exposure in transit conditions, includ-ing personal/individual factors, mode of transport, traffic characteristics, fuel type, cabin ventilation,meteorology as well as country’s socioeconomic conditions (developing countries are facing most seri-ous air pollution problems from both industrialization and urbanization processes (Cao et al., 2012; Wuet al., 2013a and Wu et al., 2013b), and thus the traffic microenvironment may be more complicated).

Thessaloniki is one of the most polluted cities in Europe (Kassomenos et al., 2011; Vlachokostaset al., 2009; EEA, 2006), in part due to climate and geography, but also because of high traffic density.Previous measurements (Vouitsis et al., 2008) indicate that main street average concentrations weresignificantly higher than those of city’s nearby background which is not affected by traffic: 7.3 � 104

vs. 1.4 � 104 particles cm�3 during working days and 6.1 � 104 vs 0.8 � 104 particles cm�3 duringweekends. For an improved characterization of the situation, the current study was carried out aimingat estimating exposure concentrations during commuting and the corresponding inhalation dosesduring trip for PM1.0, PM2.5, black carbon (BC) and PN. The study was part of the EC funded researchproject TRANSPHORM which aimed to improve the knowledge of transport related airborne particu-late matter. To this aim, we conducted measurements at different locations representing street back-ground. The main aims of the study were (1) to compare average particulate exposure/inhalation dosein different transportation modes, (2) to evaluate the additional exposure/inhalation dose (as com-pared to ambient levels) experienced while commuting, and (3) to provide information for planningand policing applications. It is the first time that kind of data are available for Thessaloniki and willbe very useful for probabilistic exposure estimates based on individual activity data, including thepopulation while in traffic or in other activities besides home and work.

2. Materials and methods

2.1. Study design

A seven-day long monitoring campaign was performed in Thessaloniki from April 05 to April 13,2011. Three commute routes using three modes of travel (bicycle (BL), bus (B) and car (C)) were

Fig. 1. Map of study area, showing commuting routes.

610 I. Vouitsis et al. / Urban Climate 10 (2014) 608–620

monitored. Route choice is an important determinant of exposure (McCreanor et al., 2007; Hertelet al., 2008) and these routes represented examples of typical commuting routes for travel withinThessaloniki (center – suburbs), with a variety of conditions including high and low traffic roadsand street canyons (Fig. 1). Route 1 (followed 3 days) connects the center of Thessaloniki to Kalamariato the south and includes two major roads of the city – Vasilisis Olgas and Tsimiski; route 2 (2 days)leads from the centre to Eleftherio-Kordelio, northwest direction, and includes one major road (Egna-tia) and a road affected by the industrial area (Monastiriou) (ViPaThe, 2013); route 3 (2 days) runsfrom the centre to Thermi, southeast, and includes two major roads (Karamanli and A Papanastasiou)as well as areas with low traffic and not so much affected by pollution. Because our interest was toassess the impact of transportation mode choice on exposures, each monitored route consisted ofthe full trip from origin to destination. Although sampling times were distributed across 4 periods(07:30 – 10:00 am, 10:00:00 – 12:00 am, 13:00 – 15:00 pm, and 15:00 – 17:00 pm) consisting of peaktraffic periods, the number of samples does not allow differentiating in any statistical significant leveland we present our results averaged all over the measurements. Table 1 gives the details of the mea-surement campaign. For the bicycle and bus mode, the equipment was placed in an appropriate bag tobe carried by the commuter. All car trips were made in the same car (Ford Focus Wagon with 1.6 Lgasoline engine, model 2008–2010, representing middle-sized <3 years old cars) in two configuration.The first configuration used in car measurements was with all windows closed (Cwc), with the cabinventilation on the second level and with the cabin air re-circulation turned off. The second studiedconfiguration (Cwo) was with co-driver’s window fully open and the with cabin ventilation turnedoff. The devices used in the study have important advantages in terms of their transportability andshort averaging times. Limitations of the instruments nevertheless need to be recognized. One sourceof error is likely to be biases in the exposure estimates due to incomplete capture of the particle sizerange (Zhu et al., 2005). The P-Trak device, for example, does not sample particles less than 20 nm inaerodynamic diameter. Systematic underestimation of the ultrafine and very fine components istherefore probable.

Table 1Measurement campaign details (Bicycle (BL), Bus (B), Car (C), w/o = window open, w/c = window closed).

Date Mode Route Starttime

Endtime

Total time(min)

Date Mode Route Starttime

Endtime

Total time(min)

05/04/11 BL 1 08:38 09:06 28 06/04/11 C (w/c) 2 07:38 07:59 2105/04/11 BL 1 (back) 09:22 09:52 30 06/04/11 C (w/c) 2 (back) 08:03 08:18 1505/04/11 BL 1 10:07 10:34 27 06/04/11 C (w/o) 2 08:23 08:46 2305/04/11 BL 1 (back) 10:51 11:20 29 06/04/11 C (w/o) 2 (back) 08:50 09:07 1705/04/11 B 1 08:38 09:21 43 06/04/11 C (w/o) 2 13:05 13:29 2405/04/11 B 1 (back) 09:22 10:02 40 06/04/11 C (w/o) 2 (back) 13:37 13:55 1805/04/11 B 1 10:07 10:35 42 06/04/11 C (w/c) 2 14:05 14:29 2405/04/11 B 1 (back) 10:51 11:26 37 06/04/11 C (w/c) 2 (back) 14:30 14:51 2105/04/11 C (w/o) 1 08:42 09:11 29 07/04/11 BL 3 08:15 08:43 2805/04/11 C (w/c) 1 (back) 09:29 09:53 24 07/04/11 BL 3 (back) 08:58 09:27 2905/04/11 C (w/o) 1 09:55 10:23 28 07/04/11 BL 3 09:44 10:15 3005/04/11 C (w/c) 1 (back) 10:34 11:00 26 07/04/11 BL 3 (back) 10:25 15:56 3205/04/11 C (w/o) 1 15:25 15:52 27 07/04/11 B (+) 3 08:15 08:49 3405/04/11 C (w/c) 1 (back) 15:54 16:16 22 07/04/11 B (++) 3 (back) 08:58 09:30 3205/04/11 C (w/o) 1 16:23 16:49 26 07/04/11 B (+) 3 09:44 10:19 3505/04/11 C (w/c) 1 (back) 16:55 17:17 22 07/04/11 B (++) 3 (back) 10:25 10:58 3306/04/11 BL 2 07:41 08:06 25 07/04/11 C (w/c) 3 08:09 8:32 2306/04/11 BL 2 (back) 08:28 08:56 28 07/04/11 C (w/c) 3 (back) 08:35 9:05 3006/04/11 BL 2 09:16 09:43 27 07/04/11 C (w/o) 3 09:15 9:3 2406/04/11 BL 2 (back) 10:03 10:30 27 07/04/11 C (w/o) 3 (back) 9:41 10:07 2606/04/11 B (⁄) 2 07:41 8:19 38 07/04/11 C (w/o) 3 13:28 13:54 2606/04/11 B (⁄⁄) 2 (back) 08:28 9:07 39 07/04/11 C (w/o) 3 (back) 13:57 14:19 2206/04/11 B (⁄⁄) 2 09:16 9:52 36 07/04/11 C (w/c) 3 14:45 15:13 2806/04/11 B (⁄⁄⁄) 2 (back) 10:03 10:43 40 07/04/11 C (w/c) 3 (back) 15:15 15:39 24(⁄) = 10 min intermediate stop, (⁄⁄) = 8 min intermediate stop, (⁄⁄⁄) 6 min intermediate stop, (+) = 6 min intermediate

stop, (++) = 1 min intermediate stop.

Date Mode Route Starttime

Endtime

Total time(min)

Date Mode Route Starttime

Endtime

Total time(min)

08/04/11 BL 1 (back) 09:05 09:32 30 11/04/11 C (w/o) 2 (back) 11:02 11:20 1808/04/11 BL 1 09:37 10:00 25 11/04/11 C (w/o) 2 11:27 11:52 2508/04/11 BL 1 (back) 10:07 10:32 25 11/04/11 C (w/o) 2 (back) 12:06 12:25 1908/04/11 BL 1 10:33 10:57 27 11/04/11 C (w/o) 2 12:27 12:50 2308/04/11 B 1 (back) 09:05 09:34 29 11/04/11 C (w/c) 2 (back) 16:05 16:34 2908/04/11 B 1 09:37 10:07 30 11/04/11 C (w/c) 2 16:35 16:54 2108/04/11 B 1 (back) 10:08 10:33 25 11/04/11 C (w/c) 2 (back) 17:01 17:24 2308/04/11 B 1 10:34 11:02 28 11/04/11 C (w/c) 2 16:27 17:43 1508/04/11 C (w/c) 1 (back) 09:05 9:21 16 12/04/11 BL 3 (back) 12:19 12:51 3208/04/11 C (w/c) 1 09:21 9:49 28 12/04/11 BL 3 12:57 13:27 3008/04/11 C (w/o) 1 (back) 09:50 10:14 24 12/04/11 BL 3 (back) 14:05 14:34 2908/04/11 C (w/o) 1 10:15 10:42 27 12/04/11 BL 3 14:40 12:11 3108/04/11 C (w/o) 1 (back) 14:26 14:49 23 12/04/11 C (w/o) 3 (back) 11:05 11:30 2508/04/11 C (w/o) 1 14:50 15:14 24 12/04/11 C (w/o) 3 11:34 12:00 2608/04/11 C (w/c) 1 (back) 15:40 15:58 18 12/04/11 C (w/o) 3 (back) 12:23 12:45 2208/04/11 C (w/c) 1 15:58 16:20 22 12/04/11 C (w/o) 3 12:48 13:09 2111/04/11 BL 2 (back) 10:58 11:25 27 12/04/11 C (w/c) 3 (back) 15:57 16:16 1911/04/11 BL 2 11:30 11:58 28 12/04/11 C (w/c) 3 16:18 16:37 1911/04/11 BL 2 (back) 12:06 12:29 23 12/04/11 C (w/c) 3 (back) 16:39 16:58 1911/04/11 BL 2 12:30 12:55 25 12/04/11 C (w/c) 3 17:01 17:19 1811/04/11 B (⁄) 2 (back) 10:58 11:35 37 13/04/11 BL 1 10:29 10:53 2411/04/11 B (⁄⁄) 2 11:40 12:15 35 13/04/11 BL 1 (back) 11:07 11:28 2111/04/11 B (⁄⁄⁄) 2 (back) 12:20 12:54 34 13/04/11 BL 1 11:44 12:07 2311/04/11 B (⁄⁄⁄⁄) 2 13:05 13:40 35 13/04/11 BL 1 (back) 12:12 12:35 23(⁄) = 1 min intermediate stop, (⁄⁄) = 7 min intermediate stop, (⁄⁄⁄) 8 min intermediate stop, (⁄⁄⁄⁄) 3 min intermediate

stop.

I. Vouitsis et al. / Urban Climate 10 (2014) 608–620 611

Date Mode Route Start time End time Total time (min)

13/04/11 B 1 10:29 11:06 3713/04/11 B 1 (back) 11:07 11:28 2113/04/11 B 1 11:44 12:09 2513/04/11 B 1 (back) 12:12 12:32 2013/04/11 C (w/o) 1 10:27 10:47 2013/04/11 C (w/o) 1 (back) 10:48 11:04 1613/04/11 C (w/o) 1 11:05 11:30 2513/04/11 C (w/o) 1 (back) 11:56 12:15 1913/04/11 C (w/c) 1 15:14 15:33 1913/04/11 C (w/c) 1 (back) 15:33 15:54 2113/04/11 C (w/c) 1 16:03 16:21 1813/04/11 C (w/c) 1 (back) 16:22 16:41 19

612 I. Vouitsis et al. / Urban Climate 10 (2014) 608–620

2.2. Measurements

The measurements were conducted with a set of PM monitors, capable of measuring particle num-ber concentration (PNC) as a proxy for ultrafine particles, PM1.0, PM2.5, PM10 and BC while bicycling,driving a car and riding a bus, along the 3 different routes leading from the city centre to the suburbsand back. Further, noise was monitored for each commuting mode. Two transportation modes weremeasured simultaneously (bus–car and bike–car) and measurements with open and closed windowswere made in alternating car drives.

Concurrently with the personal measurements, PNC, PM2.5, PM10 and BC were monitored at afixed street station. For this reason, the municipal measurement station located at Egnatia Street, alocation affected by heavy traffic nearly 24 h a day, was used. Table 2 depicts the equipment usedin the campaign. The complete measurement system with an aluminium carrying frame weightedapproximately 10 kg. During biking and bus rides, the equipment was placed in a soft case, whichcould be carried as backpack. During bicycling the measurement bag was installed onto bicycle’s rackand the PNC monitor was carried in a small backpack (to avoid interruptions in measurements due toshaking). The noise monitor was carried in the backpack’s pocket with microphone attached to leftshoulder of the bicyclist. For car measurements a hard plastic case without aluminium frame wasused. The measurement case was installed onto right back seat of the car. The noise monitor was car-ried in the driver’s door pocket with microphone attached to safety belt on the left shoulder of driver.In Cwc option, cabin ventilation was set on level 2 and re-circulation off.

At the fixed site, 24-h filter samples were collected with built-in filter collection systems of pDR(Table 2). Zefluor filters were used for PM sampling. Temperature and humidity fixed site data werecollected from the existing monitoring networks. No wind speed measurements were taken but datafrom the meteorological station of the city indicate gentle to moderate breeze condition during themeasurement campaign (average wind speed ranged from 1.96 to 5.24 m/s). To note that, althoughmeteorology parameters (along the mode of transport and traffic) have been identified as significantfactors influencing exposure concentrations to the different pollutants (Kaur et al., 2007; Knibbs et al.,

Table 2Instrumentation used in the campaign.

Instrument Measured quantity Measurementcondition

Dust Trak DRX 8533 (TSI Inc., Shoreview, MN, USA) – Laser photometer PM2.5 and PM1.0 CommutingMagee Scientific AE 51(Magee Scientific Corporation, Berkeley, CA, USA) –

AethalometerBC Commuting

P-Trak 8525 (TSI Inc., Shoreview, MN, USA) – Ultrafine particle counter PN (0.02-1 lm) CommutingLarson Davis Spark 706 (PCB Piezotronics Inc., Depew, NY, USA) – Noise

dosimeterNoise Commuting

Escort Ilog EI-HSD32L (Elektron Technology plc, Cambridge, UK) Temperature andhumidity

Commuting

PM2.5 cyclone BGI GK2.05 and Buck Libra Plus pump at 4 l/min PM2.5 (Zefluor filters) Fixed siteDustTrak (flow rate 3 l/min) PM10 (Zefluor filters) Fixed site

I. Vouitsis et al. / Urban Climate 10 (2014) 608–620 613

2011), the close examination of this influence is out of the scope of this study. In any case, the readershall be aware that during the warm period of the year (April–September), the occurrence of seabreeze is favoured due to the weak synoptic forcing (Moussiopoulos et al., 2009). The formation ofsea breeze starts in the morning when local sea breeze cells develop, initially above the Thessalonikibay with directions perpendicular to the shore, resulting in an overall inhomogeneous wind pattern. Inthe afternoon, an extended sea breeze originating from the Thermaikos gulf gradually develops, form-ing a rather homogeneous southern flow penetrating to the inland up to the northern boundary of theGTA, resulting in the transport of air masses to the northeast. To note, however, that all measurementspresented here, were made in roughly similar conditions and we’re interested of differences betweenmodes, not actual precise concentrations or even sources of the particles.

Personal monitors were set to log in 1 s–1 min concentrations depending on the downloadingspeed and the memory capacity of the device. Duplicate sampling was conducted for a week to eval-uate the performance of PM monitors; blank filters were also collected to evaluate the effect of filterhandling on PM results.

The average inhaled dose of air pollutants associated with each trip was calculated as a function ofaverage exposure concentration and travel time for each mode applying inhalation rates for bicycle,car, and bus commuting mode derive by de Nazelle et al. (2012). Comparisons between the four travelmodes were made on the basis of descriptive statistics: mean, median and standard deviation (SD),coefficient of variance (CoV) and bicycling:in-bus, bicycling:in-car and in bus:in car ratios for boththe mean and cumulative exposure and inhalation for each route (Briggs et al., 2008). Two ratio mea-sures were computed for PM1.0, PM2.5, PM10 and BC. The ‘‘ratio of the average’’ (RA) and the ‘‘averageratio of the routes’’ (ARR). RA represents the overall ratio of the mean exposure or inhalation dosewhilst commuting by mode j to the mean exposure or inhalation dose whilst commuting by mode k:

FðRAÞj:k ¼PN

i¼1Fj;iPNi¼1Fk;i

ð1a—1eÞ

where Fj,i and Fk,i are the measured exposure concentration (C) or calculated inhalation doses (ID)(j = BL, B; k = B, Cwo, Cwc, i = 1,. . .,N where N is the number of routes). The ARR represents the geomet-ric average of commuting by mode j–mode k ratios across all routes:

FðARRÞj:k ¼YN

i¼1

Fj;i

Fk;i

!1=N

ð2a—2eÞ

Correlation coefficients were also computed between different commuting exposures to eachpollutant.

Despite being a less frequently considered type of environmental pollution, noise has a major neg-ative impact on the quality of life in cities (Klæboe et al., 2000; Can et al., 2011; Ross et al., 2011).Interestingly, the higher the road traffic noise levels people are exposed to, the more likely they areto be annoyed by air pollution (for instance the smell of exhaust fumes) and vice versa (Klæboeet al., 2000), i.e. noise sensitizes them to olfactory insults. Others have assessed the relationshipsbetween a number of factors such as land use, transport, noise, and air pollution at monitoring stations(Murphy and King, 2010; Rahmani et al., 2011).

Leq (equivalent continuous sound level) is the preferred single value figure to describe sound pres-sure levels that vary over time and would produce the same sound energy over the stated period oftime T. A-weighting is applied to instrument-measured sound levels in effort to account for the rela-tive loudness perceived by the human ear. LAeq is defined as equivalent continuous A-weighted soundpressure level. As the measurements were performed with 1 s time resolution, the LAeq-levels of thewhole one way commuting trip were calculated using equation:

LAeq ¼ 10 � log101n

Xn

i¼1

100:1Li

!ð3Þ

where n is the total number of the 1-s measurements and L is the LAeq-value of each 1-s measurement.

614 I. Vouitsis et al. / Urban Climate 10 (2014) 608–620

3. Results and discussion

Fig. 2 and Table 3 summarize the exposure levels during commuting, as well as the fixed stationmeasurements.

Route 1 was the most polluted route (with the exception of bus commuting where the exposure toPM1.0, PM2.5, PM10 and PN was higher on route 2 and of commuting with car with opened windowwhere the exposure to PN and BC was again higher on route 2), while route 3 was the cleanest one. Onroute 1, the higher exposure to PM1.0, PM2.5 and PM10 was during commuting by bus, followed bybicycle and car with opened window. The lowest exposure to all PM components was during commut-ing by car with closed window. Exposure to PN on route 1 was higher during commuting by car withopened window, followed by bus commuting and car with closed window. Exposure to BC was higherduring commuting by bicycle, followed by car with opened window, bus and car with closed windowwhich again showed the lowest exposure levels.

The situation was nearly the same on route 2. For PM components the highest exposure occurredagain during commuting by bus, followed by car with opened window in this case and then by bicycle.Commuting by car with closed window was again the cleanest transport mode. In this route, exposureto PN was slightly higher during commuting by car with opened window than during commuting bybus, whereas again the lowest exposure was associated with car commuting with closed window.Similar was the situation for BC.

In route 3, exposure to PM1.0 and PM2.5 was almost identical for bus and bicycle commuting, com-muting by car with opened window followed while the lower exposure occurred again when commut-ing by car with closed window. Exposure to PM10 was slightly higher for bus commuting, followed bybicycle, car with opened window and car with closed window. Bicycle commuting showed the highestexposure to PN, followed by car with opened window, bus and car with closed window. Surprisingly,commuting by car with closed window showed higher exposure to BC than bicycle commuting andeven higher than bus commuting. Commuting by car with opened window showed the highest expo-sure in this case. To note, however, that the levels of exposure on route 3 were significantly lowerwhen compared to the levels of the other routes.

Measured pollutant concentrations during commuting appear to be different than the concentra-tions measured in fixed street station, depending on the commuting mode and the route travelled.Results are mixed during travelling on routes 1 and 2, while on route 3 pollution levels were found

Fig. 2. Average exposures to PM1.0, PM2.5, PM10, BC and PN in different modes of transport for the three routes.

Table 3PM1.0. PM2.5, PM10, BC and PN concentrations during commuting by bicycle, bus and car (Bicycle (BL), Bus (B), Car (C), w/o = window open, w/c = window closed).

Mode PM1.0 (lg/m3) PM2.5 (lg/m3) PM10 (lg/m3) PNC (#/cm3) � 104 BC (lg/m3)

Mean ± SD Median CoV Mean ± SD Median CoV Mean ± SD Median CoV Mean ± SD Median CoV Mean ± SD Median CoV

Route 1 (3 days)BL 79 ± 34 50 43 81 ± 34 51 42 141 ± 84 70 59 – – – 10.5 ± 6.7 6.8 64B 96 ± 19 91 20 101 ± 22 93 22 140 ± 40 127 29 4.18 ± 10.48 3.6075 25 7.4 ± 3.1 6.8 42Cwo 69 ± 23 65 33 70 ± 23 66 33 84 ± 29 80 34 5.12 ± 3.18 4.2215 62 9.0 ± 4.8 7.0 53Cwc 49 ± 14 45 29 50 ± 14 46 28 60 ± 18 55 30 3.082 ± 1.71 2.3625 55 4.2 ± 2.8 4.0 67

Route 2 (2 days)BL 44 ± 3 35 7 45 ± 3 36 7 57 ± 5 45 9 – – – – – –B 98 ± 23 80 24 105 ± 26 85 25 212 ± 61 171 29 6.43 ± 0.67 6.1625 10 6.2 ± 2.0 3.8 32Cwo 54 ± 7 47 13 55 ± 7 48 13 74 ± 3 63 4 6.67 ± 1.83 5.5725 27 10.9 ± 4.0 8.9 37Cwc 26 ± 5 23 19 26 ± 5 23 19 31 ± 08 27 20 2.25 ± 0.29 1.8825 13 4.2 ± 3.2 4.0 76

Route 3 (2 days)BL 40 ± 12 29 30 41 ± 13 29 27 48 ± 12 35 25 4.39 – – 3.7 – –B 38 ± 3 36 8 41 ± 3 38 7 54 ± 4 48 7 3.43 3.03 3.0 2.8 –Cwo 32 ± 3 29 9 33 ± 3 30 9 43 ± 4 39 9 4.29 ± 1.20 3.17 28 4.8 3.7 –Cwc 14 ± 3 13 21 15 ± 3 13 21 16 ± 3 14 19 3.21 – – 3.8 2.9

Fixed siteRoute 1 83 ± 17 78 20 85 ± 18 79 21 91 ± 21 65 33 4.72 ± 1.36 4.05 29 8.2 ± 1.7 6.7 82Route 2 45 ± 6 41 13 45 ± 6 41 13 47 ± 6 43 13 4.42 ± 0.95 3.76 22 6.0 ± 1.5 5.1 85Route 3 66 ± 16 62 24 67 ± 16 63 24 69 ± 17 63 25 4.29 ± 0.60 3.74 14 3.6 ± 0.6 3.1 86

I.Vouitsis

etal./U

rbanClim

ate10

(2014)608–

620615

616 I. Vouitsis et al. / Urban Climate 10 (2014) 608–620

lower than street concentrations for almost all pollutants and all transport modes. More specifically,for route 1 measured pollutant levels during commuting were generally higher than that of street lev-els for PM10 and BC for bicycle commuting, for PM1.0, PM2.5 and PM10 for bus commuting and for PNand BC for car with open window commuting; for route 2: PM10 for bicycle commuting and all pol-lutants for bus and car with opened window commuting; route 3: PN and BC for bicycle commutingand BC for car (both opened and closed window commuting). Commuting by car with closed windowwas generally found to be associated with lower exposure levels compared with the street back-ground, with the exception of BC. In addition, as indicated by the differences between mean and med-ian values, most exposure peaks are experienced during bicycling in all three routes and for all thepollutants measured and to a lesser extent when commuting by bus and car with both opened andclosed window, on route 2 in particular. However, as shown by CoV values, bicycling measurementswere those with the highest degree of variation and shall be considered very carefully.

Table 4 gives Pearson correlations between average exposures during commuting for each pollu-tant. Strong linear correlations are seen in most cases suggesting significant association.

Considering the inhaled dose of pollutants after accounting for trip duration and inhalation rates,calculating inhaled pollutant dose are shown in Table 5. Patterns are generally similar to those of aver-age exposures, but because of the higher inhalation rate associated with bicycling, the bicycling:in-carratios are larger. Compared to bus riders, cyclists inhaled 35% more PM1.0 and PM2.5, 62% more BCand 56% more particles. Compared to car drivers and depending on the car window option, cyclistsinhaled 50–90% more PM1 and PM2.5, 48–84% more BC and 55–72% more particles. Compared tocar drivers, bus riders inhaled 38–85% more PM1.0 and PM2.5, 37–81% more particles and 20–60%more BC (from car with closed window). Exception was route 2 where bus riders inhaled about 50%more PM1.0 and PM2.5 than cyclists and car drivers with open window inhaled about 4% more BCthan bus riders. In the first case, considering that the measurements conducted simultaneously, theincreased confounding influence of other sources by the bus itself is minimal and self-pollution seemsprobable (tailpipe and/or engine crankcase emissions).

Table 6 shows the ratio of averages (RA) as CRA for exposure and as IDRA for inhalation dose andthe average ratio of routes (AAR) as CARR for exposure and as IDARR for inhalation dose for PM1.0,PM2.5 and PM10. CRA seems to increase with increasing particle size (from an average of 1.54 forPM1.0 to 1.99 for PM10) and in bus:in car with closed windows ratio showed the largest value in mostcases. IDRAs were somewhat higher and showed the same pattern regarding particle size (from anaverage of 2.93 for PM1.0 to 3.76 for PM10). In this case, however, the largest calculated ratio was thatof bicycling: in car with closed windows, obviously due to increased inhalation rate of bicyclists. CARRshowed almost the same pattern for particle size (from an average of 1.58 for PM1.0 to 1.93 for PM10).Again, the largest calculated ratio was that of bicycling: in car with closed windows.

Although comparisons with other studies are difficult due to the wide variety of designs, includ-ing differences in the simultaneity of trips by various modes, the choice of routes by mode, the ven-tilation setting in cars, or the type of monitoring instruments used, as well as the particular settingof the study site, it is clear that our results are in contrast with most of the studies which indicatedgeneral tendencies of higher concentrations in cars than walking or biking (50% higher car concen-trations for PM2.5 or PNC – (Knibbs et al., 2011; Zuurbier et al., 2010; McNabola et al., 2008; Kauret al., 2005, 2007). Two exceptions include studies in the UK, with PM2.5 levels 40–100% higherwalking than in a car (Briggs et al., 2008; Gulliver and Briggs, 2007). Also in Belgium, PM2.5 and

Table 4Pearson correlations between average exposures during bicycling, in-bus and in-car travelling (Bicycle (BL), Bus (B), Car (C),w/o = window open, w/c = window closed).

PM1.0 PM2.5 PNC BC

BL–B 0.557 0.512 – 0.918Cwo 0.901 0.886 0.992 0.838Cwc 0.739 0.711 0.996 0.914Cwo 0.862 0.853 – 0.554Cwc 0.971 0.969 – 0.681

Table 5Calculation of inhalation of pollutants by mode accounting for varying inhalation rate (IR), trip time, and mean concentration bymode (Bicycle (BL), Bus (B), Car (C), w/o = window open, w/c = window closed).

IR (lpm) Trip time Mean concentration Inhaled dosea (trip)

R1 R2 R3 R1 R2 R3 R1 R2 R3

PM1.0 min lg m�3 lgBL 41 26 27 30 79 44 40 83.4 49.2 49.0B 20.1 28 37 34 96 98 38 53.5 72.6 26.2Cwo 19.9 24 22 25 69 54 32 33.0 23.4 15.9Cwc 19.9 21 22 23 49 26 14 20.1 11.3 6.4Fixed site 4.6/4.0 83 45 66PM2.5 lg m�3 lgBL 41 26 27 30 81 45 41 85.5 50.3 50.2B 20.1 28 37 34 101 105 41 56.3 77.8 28.2Cwo 19.9 24 22 25 70 55 33 33.4 23.8 16.4Cwc 19.9 21 22 23 50 26 15 20.5 11.3 6.8Fixed Site 4.6/4.0 85 45 67PM10BL 41 26 27 30 141 57 48 149.3 63.2 58.6B 20.1 28 37 34 14 212 54 78.2 157.2 36.9Cwo 19.9 24 22 25 84 74 43 40.2 32.1 21.2Cwc 19.9 21 22 23 60 31 16 24.5 13.5 7.3Fixed Site 4.6/4.0BC lg m�3 lgBL 41 26 27 30 10.5 3.7 11.1 4.5B 20.1 28 37 34 7.4 6.2 3.0 4.1 4.6 2.1Cwo 19.9 24 22 25 9.0 10.9 4.08 4.3 4.7 2.4Cwc 19.9 21 22 23 4.2 4.2 3.08 1.7 1.8 1.7Fixed Site 4.6/4.0 8.2 6.0 3.6PN # cm�3 #�1010

BL 41 26 27 30 43985 – – 5.4B 20.1 28 37 34 41824 64272 34314 2.4 4.8 2.4Cwo 19.9 24 22 25 51172 66718 42959 2.4 2.9 2.1Cwo 19.9 21 22 23 30732 22487 32139 1.3 0.9 1.5Fixed Site 4.6/4.0 47159 44194 42908

a Inhaled dose (ID) is calculated as: ID = inhalation rate (IR) � concentration (C) � Duration of exposure.

Table 6Ratio of averages for exposure (CRA) and inhalation dose (IDRA) and average ratio of routes for exposure (CARR) and inhalationdose (IDARR).

PM1.0 PM2.5 PM10 BC

CRA CAAR CRA CAAR CRA CAAR CRA CAARBL–B 0.70 0.73 0.68 0.70 0.61 0.62 1.37 1.21BL–Cwc 1.83 1.98 1.84 1.84 2.30 2.34 1.78 1.35BL–Cwo 1.05 1.05 1.06 1.06 1.22 1.13 1.03 0.97B–Cwc 2.61 2.72 2.71 2.71 3.79 3.76 1.36 1.27B–Cwo 1.50 1.44 1.56 1.51 2.02 1.82 0.67 0.66

IDRA IDAAR IDRA IDAAR IDRA IDAAR IDRA IDAAR

BL–B 1.19 1.25 1.15 1.20 1.00 1.07 2.52 1.81BL–Cwc 4.80 5.18 4.82 5.02 5.98 6.10 4.52 2.56BL–Cwo 2.50 2.84 2.53 2.55 2.90 2.72 2.34 1.70B–Cwc 4.04 4.13 4.21 4.28 6.01 5.72 2.05 1.93B–Cwo 2.11 2.02 2.20 2.12 2.91 2.55 0.95 0.93

I. Vouitsis et al. / Urban Climate 10 (2014) 608–620 617

PM10 concentrations measured in cars in two of three cities were significantly lower than concen-tration found while riding bikes in trips immediately following the car trip; differences in PNC werenot significant in these same two cities and the reverse was true for both contaminants in the third

Table 7The median LAeq, LA95 and LA5 levels by commuting mode.

Commuting mode Median LAeq (dB) Median LA95 (dB) Median LA5 (dB)

BL 74.9 61.7 80.3B 73.2 63.7 77.9C-wc 70.7 54.2 74.6C-wo 72.1 55.2 77.5

618 I. Vouitsis et al. / Urban Climate 10 (2014) 608–620

city (Int Panis et al., 2010). Since it has been assumed that this may have occurred because the carshad closed window and used air conditioning which reduces PM concentrations, in our study weconducted measurements with both options. The fact that in most cases car driving was the com-muting mode with the lower exposure/inhalation dose (in all cases when the window was closedand with three exceptions in case of BC when the window was opened), indicates further the heavyparticulate background of the city.

Finally, Table 7 shows noise exposure in terms of median equivalent continuous sound level (LAeq),LA95 (the level exceeded for 95% of the time) and LA5 (the level exceeded for 5% of the time). Car withclosed window was the quietest transportation mode. The median LAeq-levels in a car with open co-driver’s window were 1.4 dB higher than in a car with closed windows. It is notable that even whenthe co-driver’s window was open the car was less noisy than a bus. To note that WHO’s guideline valuefor serious annoyance during daytime and evenings in outdoor living area is 55 dB, which is exceededin all measured transportation modes. Vlachokostas et al. (2012) reported lower average noise levelsin Thessaloniki during walking (70.6 dB), biking (70.9 dB) and driving a car with open front windows(66.3 dB). The routes measured by Vlachokostas et al. (2012) were designated to represent typicalpaths of citizens through Thessaloniki metropolitan centre, and were characterized by high trafficloads, slow speeds and frequent stops during rush hours.

4. Conclusions

This study attempted to assess personal exposure and inhalation dosing to particulate matter(PM1.0, PM2.5, PM10, BC and particle number) during different modes of commuting (bicycle, busand car with all windows closed and with co-driver’s window open) in three different routes in thecity of Thessaloniki. Although with a significant amount of uncertainty, the results presented suggestthat exposure and inhalation doses are likely to be markedly higher for people bicycling in busy streetscompared to car drivers, at least under the driving conditions encountered here. This is in agreementwith Vlachokostas et al. (2014) who estimated fast cycling in the city center as hazardous on the andrecommend precaution. Although in reality bicyclists are usually traveling in cleaner areas, they mustbe informed of the high exposure due to their increased physical activity. Unfortunately, this is not aconsequence of the commuting choice, but a consequence of others’ commuting behavior. Increasedexposure to particulates experienced by bus commuters also. Given that pollutant levels inside busescan be greater than the ambient levels outside the bus as also indicated from several studies (i.e.,Molle et al., 2013) due to either emissions from the bus itself that intrude into the bus cabin (ageand condition of the bus) or due to emissions from other vehicles (opened windows), the use of airfiltration or, better, the use of clean buses, such as electric buses or buses equipped with particulatefilters, will be therefore beneficial not only for outdoor air quality but also for bus passengers. More-over, the longer exposure duration for bus commuters further enlarge the exposure risks to particulatepollution and should be taken into account. Commuting by car with open windows led to significantexposures to black carbon and particle numbers, while commuting by car with closed windows is thetransport mode over which the least exposure to air pollution was monitored. Thus, an occupant of arelatively airtight automobile in which air is recirculated and filtered will likely experience markedlylower exposure concentrations than a cyclist or a bus commuter on a high traffic route. In view of thepresent situation in Thessaloniki and waiting for the implementation of the appropriate measures, carcommuting seems to be the most free of particulates mode as regards personal exposure and inhala-tion dose during commuting. Noise during commuting by all modes of travelling was found to be

I. Vouitsis et al. / Urban Climate 10 (2014) 608–620 619

considerably above recommended value for serious annoyance during daytime and evenings in out-door living area. Although the time of study is not enough to provide guidance for transport policymaking, city authorities should be encouraged to increase the use of clean buses along with othermeasures aiming at the reduction of traffic particle emissions (traffic reduction, bike and bus lanes,different bus ventilation, particle filter for air conditioning).

Acknowledgement

This work was performed under TRANSPHORM project funded by the European Commission (Grantagreement No.: 243406).

References

Adams, H.S., Niewenhuijsen, M.J., Colvile, R.N., McMullen, M.A.S., Khandelwal, P., 2001. Fine particle (PM2.5) personal exposurelevels in transport microenvironments, London, UK. Sci. Total Environ. 279, 29–44.

Adams, H.S., Niewenhuijsen, M.J., Colvile, R.N., Older, M.J., Kendall, M., 2009. Assessment of road users’ elemental carbonpersonal exposure levels, London, UK. Atmos. Environ. 36 (24), 5335–5342.

Adar, S.D., Gold, D.R., Coull, B.A., Schwartz, J., Stone, P.H., Suh, H., 2007. Focused exposures to airborne traffic particles and heartrate variability in the elderly. Epidemiology 18, 95–103.

Bernstein, D.I., 2012. Traffic-related pollutants and wheezing in children. J. Asthma 49, 5–7.Briggs, D.J., de Hoogh, K., Morris, C., Gulliver, J., 2008. Effects of travel mode on exposures to particulate air pollution. Environ.

Int. 34, 12–22.Brook, R.D., Bard, R.L., Burnett, R.T., Shin, H.H., Vette, A., Croghan, C., Phillips, M., Rodes, C., Thornburg, J., Williams, R., 2011.

Differences in blood pressure and vascular responses associated with ambient fine particulate matter exposures measuredat the personal versus community level. Occup. Environ. Med. 68, 224–230.

Can, A., Rademaker, M., Van Renterghem, T., Mishra, V., Van Poppel, M., Touhafi, A., Theunis, J., De Baets, B., Botteldooren, D.,2011. Correlation analysis of noise and ultrafine particle counts in a street canyon. Sci. Total Environ. 409 (3), 564–572.

Cao, J.J., Hang, H., Lee, S.C., Chow, J.C., Zou, C.W., Ho, K.F., Watson, J.G., 2012. Indoor/outdoor relationships for organic andelemental carbon in PM2.5 at residential homes in Guangzhou, China. Aerosol Air Qual. Res. 12, 902–910.

Chan, L.Y., Lau, W.L., Lee, S.C., Chan, C.Y., 2002. Commuter exposure to particulate matter in public transportation modes inHong Kong. Atmos. Environ. 36, 3363–3373.

Chiu, Y.-H.M., Bellinger, D.C., Coull, B.A., Anderson, S., Barber, R., Wright, R.O., Wright, R.J., 2013. Associations between traffic-related black carbon exposure and attention in a prospective birth cohort of urban children. Environ. Health Perspect. 121(7), 859–864.

De Nazelle, A., Scott Fruin, S., Westerdahl, D., Martinez, D., Ripoll, A., Kubesch, N., Nieuwenhuijsen, 2012. A travel modecomparison of commuters’ exposures to air pollutants in Barcelona. Atmos. Environ. 59, 151–159.

Dons, E., Int Panis, L., Van Poppel, M., Theunis, J., Wets, G., 2012. Personal exposure to black carbon in transportmicroenvironments. Atmos. Environ. 55, 392–398.

European Environment Agency (EEA), 2006. Air pollution at street level in European cities. EEA Technical report No. 1/2006.Geiss, O., Barrero-Moreno, J., Tirendi, S., Kotzias, D., 2010. Exposure to particulate matter in vehicle cabins of private cars.

Aerosol Air Qual. Res. 10, 581–588.Gulliver, J., Briggs, D.J., 2004. Personal exposure to particulate air pollution in transport microenvironments. Atmos. Environ. 38,

1–8.Gulliver, J., Briggs, D.J., 2007. Journey-time exposure to particulate air pollution. Atmos. Environ. 41, 7195–7207.Hertel, O., Hvidberg, M., Ketzel, M., Storm, L., Stausgaard, L., 2008. A proper choice of route significantly reduces air pollution

exposure e a study on bicycle and bus trips on urban streets. Sci. Total Environ. 389, 58–70.Hoek, G., Krishnan, R.M., Beelen, R., Peters, A., Ostro, B., Brunekreef, B., Kaufman, J.D., 2013. Long-term air pollution exposure and

cardio-respiratory mortality: a review. Environ. Health 12, 1–15.Int Panis, L., de Geus, B., Vandenbulcke, G., Willems, H., Degraeuwe, B., Bleux, N., Mishra, V., Thomas, I., Meeusen, R., 2010.

Exposure to particulate matter in traffic: a comparison of cyclists and car passengers. Atmos. Environ. 44, 2263–2270.Kassomenos, P.A., Kelessis, A., Paschalidou, A.K., Petrakakis, M., 2011. Identification of sources and processes affecting

particulate pollution in Thessaloniki, Greece. Atmos. Environ. 45 (39), 7293–7300.Katsouyanni, K., Touloumi, G., Spix, C., 1997. Short term effects of ambient sulphur dioxide and particulate matter on mortality

in 12 European cities: results from time series data from the APHEA project. Br. Med. J. 314, 1658–1663.Katsouyanni, K., Gryparis, A., Samoli, E., 2011. Short-Term Effects of Air Pollution on Health. Reference Module in Earth Systems

and Environmental Sciences, from Encyclopedia of Environmental Health, pp. 51–60.Kaur, S., Nieuwenhuijsen, M., Colvile, R., 2005. Personal exposure of street canyon intersection users to PM2.5, ultrafine particle

counts and carbon monoxide in Central London, UK. Atmos. Environ. 39, 3629–3641.Kaur, S., Nieuwenhuijsen, M.J., Colvile, R.N., 2007. Fine particulate matter and carbon monoxide exposure concentrations in

urban street transport microenvironments. Atmos. Environ. 41, 4781–4810.Klæboe, R., Kolbenstvedt, M., Clench-Aas, J., Bartonova, A., 2000. Oslo traffic study-part 1: an integrated approach to assess the

combined effects of noise and air pollution on annoyance. Atmos. Environ. 34 (27), 4727–4736.Knibbs, L.D., de Dear, R.J., 2010. Exposure to ultrafine particles and PM2.5 in four Sydney transport modes. Atmos. Environ. 44,

3224–3227.Knibbs, L.D., Cole-Hunter, T., Morawska, L., 2011. A review of commuter exposure to ultrafine particles and its health effects.

Atmos. Environ. 45, 2611–2622.

620 I. Vouitsis et al. / Urban Climate 10 (2014) 608–620

Lai, H.K., Kendall, M., Ferrier, H., Lindup, I., Alm, S., Hänninen, O., Jantunen, M., Mathys, P., Colvile, R., Ashmore, M.R., Cullinan, P.,Nieuwenhuijsen, M.J., 2004. Personal exposures and microenvironment concentrations of PM2.5, VOC, NO2 and CO inOxford, UK. Atmos. Environ. 38, 6399–6410.

Lim, S., Kim, J., Kim, T., Lee, K., Yang, W., Jun, S., Yu, S., 2012. Personal exposures to PM2.5 and their relationships withmicroenvironmental concentrations. Atmos. Environ. 47, 407–412.

Loth, K.W., Ashmore, M.R., 1994. Personal exposure to air pollution. Clean Air 24, 114–122.McCreanor, J., Cullinan, P., Nieuwenhuijsen, M.J., Stewart-Evans, J., Malliarou, E., Jarup, L., Harrington, R., Svartengren, M., Han,

I.K., Ohman-Strickland, P., Chung, K.F., Zhang, J., 2007. Respiratory effects of exposure to diesel traffic in persons withasthma. N. Engl. J. Med. 357, 2348–2358.

McNabola, A., Broderick, B.M., Gill, L.W., 2008. Relative exposure to fine particulate matter and VOCs between transportmicroenvironments in Dublin: personal exposure and uptake. Atmos. Environ. 42, 6496–6512.

Molle, R., Mazoue, S., Gehin, E., Ionescu, A., 2013. Indoor–outdoor relationships of airborne particles and nitrogen dioxide insideParisian buses. Atmos. Environ. 69, 240–248.

Moussiopoulos, N., Vlachokostas, Ch., Tsilingiridis, G., Douros, I., Hourdakis, E., Naneris, C., Sidiropoulos, C., 2009. Air qualitystatus in greater Thessaloniki area and the emission reductions needed for attaining the EU air quality legislation. Sci. TotalEnviron. 407 (4), 1268–1285.

Murphy, E., King, E.A., 2010. Strategic environmental noise mapping: methodological issues concerning the implementation ofthe EU Environmental Noise Directive and their policy implications. Environ. Int. 36 (3), 290–988.

Peters, A., von Klot, S., Heier, M., Trentinaglia, I., Hormann, A., Wichmann, H.E., Hannelore, L., 2004. Exposure to traffic and theonset of myocardial infarction. N. Engl. J. Med. 351, 1721–1730.

Rahmani, S., Mousavi, S.M., Kamali, M.J., 2011. Modeling of road-traffic noise with the use of genetic algorithm. Appl. SoftComput. 11 (1), 1008–1013.

Riediker, M., Cascio, W.E., Griggs, T.R., Herbst, M.C., Bromberg, P.A., Neas, L., Williams, R.W., Devlin, R.B., 2004. Particulate matterexposure in cars is associated with cardiovascular effects in healthy young men. Am. J. Respir. Crit. Care Med. 169, 934–940.

Ross, Z., Kheirbk, I., Clougherty, J.E., Ito, K., Matte, T., Markowitz, S., Eisl, H., 2011. Noise, air pollutants and traffic: continuousmeasurement and correlation at a high-traffic location in New York City. Environ. Res. 111, 1054–1063.

Vellopoulou, A.V., Ashmore, M.R., 1998. Personal exposures to carbon monoxide in the city of Athens. 1. Commuters’ exposures.Environ. Int. 24, 713–720.

ViPaThe, 2013. Industrial Park of Thessaloniki. Available at <http://www.vipathe.gr/en/parko_thesi.asp>.Vlachokostas, Ch., Achillas, Ch., Moussiopoulos, N., Hourdakis, E., Tsilingiridis, G., Ntziachristos, L., Banias, G., Stavrakakis, N.,

Sidiropoulos, C., 2009. Decision support system for the evaluation of urban air pollution control options: application forparticulate pollution in Thessaloniki, Greece. Sci. Total Environ. 407 (13), 5937–5948.

Vlachokostas, Ch., Achillas, Ch., Michailidou, A.V., Moussiopoulos, N., 2012. Measuring combined exposure to environmentalpressures in urban areas: an air quality and noise pollution assessment approach. Environ. Int. 39, 8–18.

Vlachokostas, C., Banias, G., Athanasiadis, A., Achillas, C., Akylas, V., Moussiopoulos, N., 2014. CENSE: a tool to assess combinedexposure to environmental health stressors in urban area. Environ. Int. 63, 1–10.

Vouitsis, I., Ntziachristos, L., Samaras, Z. Roadside and urban background measurements of ultrafine particles in Thessaloniki,European Aerosol Conference, Thessaloniki, 24–29 August 2008.

Wang, X., Gao, H., 2011. Exposure to fine particle mass and number concentrations in urban transportation environments ofNew York city. Transp. Res D 16, 1361–9209.

Wu, D.L., Lin, M., Chan, C.Y., Li, W.Z., Tao, J., Li, Y.P., Sang, X.F., Bu, C.W., 2013a. Influences of commuting mode, air conditioningmode and meteorological parameters on fine particle (PM2.5) exposure levels in traffic microenvironments. Aerosol AirQual. Res 13, 709–720.

Wu, M., Wu, D., Fan, Q., Wang, B.M., Li, H.W., Fan, S.J., 2013b. Observational studies of the meteorological characteristicsassociated with poor air quality over the Pearl River Delta in China. Atmos. Chem. Phys. 13, 10755–10766.

Zhu, Y.F., Hinds, W.C., Krudysz, M., Kuhn, T., Froines, J., Sioutas, C., 2005. Penetration of freeway ultrafine particles into indoorenvironments. J. Aerosol Sci. 36, 303–322.

Zuurbier, M., Hoek, G., Oldenwening, M., Lenters, V., Meliefste, K., van den Hazel, P., Brunekreef, B., 2010. Commuters’ exposureto particulate matter air pollution is affected by mode of transport, fuel type, and route. Environ. Health Perspect. 118, 783–789.