)p<°@° *x9J9*<°
=C<>°
=)p=x !6Eg°
#Jh=°A°$E>K°
*>°@°)¢):°
i=+>°L6d°¨ $FM°E=N° q)¤°3=° ' N 7xj=9° Oy° F=° L3,¥° ¢PE°E?°*E° ©=rQMx %E ?
8R<9°¢LE°G?°-E°?sRSz° T9=9°M° U°?h¤°+7dx¢k=9C<: M ° E8V?9° W{° £El?°¢XH°E?°-H°¢X=x°?rYZz° %"N N @°¤°.x¢??:°=N
°*I\¦>°%E=° &|L ?]¥°@°E?°"E° /6L@a6°° ° :7>°)° r^6@bhu° °rL8@c6E?°6¤°@°?= +R}° = %"N N =0_x° /°8¤°Mx°=m?8~L8°Bs+°@° -8HL¡-h° +:°??.Lx°?° %"N v-e=°EN«°- *Rn*5h>° %#HN N =NG@N !J9FG:A N :K9N D95FB !JCx<9° 2>° §° ®° ¯° Ex<° 3`h=° w?° (f° ASSESSMENT OF FINE PARTICULATE MATTER IN HEAVY TRAFFIC AREAS IN FIJI by Shavneet Ambar Mani A thesis submitted in partial fulfillment of the requirement for the degree of Master in Science on Chemistry Copyright © 2020 by Shavneet Ambar Mani School of Biological and Chemical Sciences Faculty of Science, Technology and Environment The University of the South Pacific March, 2020 Dedication To my mum; if only words could describe the blessing that you are……. i Acknowledgment In retrospect, this section seemed like the easiest to write. It has quickly become the hardest as I try to turn emotions into words. A thousand faces immediately flash up, all of whom have played a beat to tune this thesis. I begin with name of almighty, Lord Ganesha, the source of my will and breath, both of which he has provided in the most difficult of times. Second to god, I thank my mum, Yeshmin Lata Narayan, who has believed in me more than I ever could. Thank you for reminding me that I was capable, especially when my nerves frayed. I must sincerely extend my deepest gratitude to my supervisor, Dr Francis Mani. I admire your knowledge and patience, especially since I have led you to age a decade in the past 2 years. Thanks for keeping me in check! Special thanks also goes to Mrs Monika Lemestre and Dr Peggy Gunkel- Grillon of the Institute of Exact and Applied Sciences at the University of New Caledonia. Sincere appreciation and gratitude also goes to Professor Richard Peltier of the University of Massachusetts for guiding me throughout the research and providing professional insights. Many thanks to my academic advisors and technical staff including Mr Steven Sutcliffe, Dr Ajal Kumar, Joslin Lal, Thomas Tunidau, Shelvin Prasad and everyone at the “chem prep room”. I thank the Research Office and the Faculty of Science, Technology and Environment for providing all the necessary funding and support. I would certainly be made to disappear if I did not mention the members of my “tea gang”, Shivam Hanushri Goundar, Kunal Singh, Kavnil Lal, Zahra Azid and Janice Mani. Thank you so much Shamal and Sweta or should I say shamuweta? I would also like to acknowledge Aysheal Chand, Vinal Prasad and Krishneel Prasad for their guest appearances. ii My hearty appreciation goes to Mr and Mrs Ronal Deo and the kids, for showing me what it means to be humble. Special thanks also goes to Mr Davendra Nand and Mrs Roselyn Karan. You all were the team working in the background and have provided immense moral support. Finally, I would like to thank members of my family. Special thanks to my dad, Mr Shailendra Mani. Thanks to my namesake, Mr Shavneet Mani (Monu), my grandmother, Sheeran Kumari, Shaya Malini, Mohammed Salim, Kapul Prasad, Rahul Chand (Papa Joe), Avishal Singh and Sagita Devi. Before I finish, thank you so much Ashley (my niece) for forcefully teaching me what you have learnt at pre-school. I hope I don’t have to sit for an exam. Thanks to everyone that I may have missed. You know who you are so smile. iii Abstract Traffic emissions are a primary source of PM2.5-related air pollution in many urban environments. However, air quality data on heavy traffic areas in urban centres across the Pacific Island Countries (PICs) including Fiji, remain largely absent. To bridge the data gap, we monitored roadside PM2.5 concentrations in two of the largest cities in the south pacific island countries, Suva and Lautoka City in Fiji. A high volume air sampler was used for sampling in densely populated and heavy trafficked areas in Suva from September 2018 to January 2019 and in Lautoka from January to March 2019. Daily mean PM2.5 concentrations in Suva and Lautoka cities were reported to be 21.6 ± 13.3 μg/m3 and 67.2 ± 35.2 μg/m3, respectively. In comparison, PM2.5 concentrations in Lautoka City were more than twice the World Health Organisation 24 hour mean guideline concentration of 25 μg/m3. Correlations between PM2.5 and meteorological parameters including wind, rainfall, humidity and temperature were also investigated. Real time concentrations of PM2.5 and Black Carbon (BC) were also monitored to assess the diurnal patterns. Real time concentrations were monitored at four different sites, namely, Samabula, Suva City Bus Station, Reservoir Road community and Lautoka City. The BC concentrations were 3.9 ± 2.9 μg /m3 and 2.6 ± 2.7 μg /m3, 2.4 ± 2.3 μg /m3 and at 4.02 ± 4.7 μg /m3 at Samabula, Suva City Bus Station, Reservoir Road and Lautoka City, respectively. Concentrations of Al, As, Ba, Ca, Cd, Co Cr, Cu, Fe, K, Mg, Mn, Mo, Na, Ni, P, Pb, S, Si, Sr, V, and Zn were investigated along with their Enrichment Factors and these are presented in Chapter 4. Elements such as Al, Ba, Ca, K, Na and Zn were observed to have concentrations greater than 1 μg/m3 in both the cities. This study demonstrates that PM2.5 related air quality is already worse in Lautoka City in comparison to Suva, with mean concentrations greater than twice the WHO daily mean guidelines. Due to the absence of air quality legislation and significantly increasing number of cars, air quality could deteriorate to levels with significant public health impacts in future unless intervention is made. Finally, policy recommendations to reduce PM2.5 and BC emissions from land transport sector is also provided. iv List of Acronyms AGL Above Ground Level Al Aluminum As Arsenic Ba Barium BC Black Carbon Ca Calcium CBD Central Business District Cd Cadmium Co Cobalt CO2 Carbon Dioxide Cr Chromium Cu Copper EC Elemental Carbon EF Enrichment Factor EPA Victoria Environmental Protection Authority Victoria Fe Iron FJD Fiji Dollar GHG Green House Gasses HSU Hatridge Smoke Unit HVS High Volume Sampler HYSPLIT Hybrid Single Particle Lagrangian Integrated Trajectory K Potassium Mg Magnesium Mn Manganese Mo Molybdenum LTA Land Transport Authority LVS Low Volume Sampler Na Sodium v Ni Nickle NGO’s Non-Governmental Organizations NOx Nitrogen Oxides NVOC Non Volatile Organic Compound OC Organic Carbon OECD Organization for Economic Cooperation and Development P Phosphorus PAH’s Poly Aromatic Hydrocarbons Pb Lead PICs Pacific Island Countries PM Particulate Matter PM10 Particles having aerodynamic diameter equal less than 10 μm PM2.5 Particles having aerodynamic diameter equal or less than 2.5 μm RH Relative humidity S Sulfur Si Silicon SSI Size Selective Inlet SO2 Sulfur Dioxide SOA Secondary Organic Aerosols Sr Strontium TC Total Carbon TIN’s Traffic Infringement Notices USD United States Dollar US EPA / EPA United States Environmental Protection Agency V Vanadium VOC Volatile Organic Compound WHO World Health Organization WSOC Water Soluble Organic Carbon Zn Zinc vi Table of Contents Dedication ...... i Acknowledgment ...... ii Abstract ...... iv List of Acronyms ...... v Table of Contents ...... vii List of Tables...... ix List of Figures ...... x 1.0 Introduction ...... 1 1.1 Background and Significance ...... 1 1.2 Aim ...... 7 1.3 Objectives ...... 7 2.0 Literature Review ...... 8 2.1 Definition of particulate and composition ...... 8 2.2 Health Impacts and Exposure Limits of PM2.5 ...... 11 2.3 Sources of PM2.5 ...... 18 2.4 PM2.5 in a PIC: Fiji ...... 20 2.5 Dispersion Factors ...... 23 2.5.1 Wind speed and direction ...... 23 2.5.2 Rainfall...... 24 2.5.3 Relative Humidity ...... 24 2.5.4 Temperature ...... 25 3.0 Methodology ...... 26 3.1 Sampling Site ...... 26 3.1.1 Site 1 – Suva ...... 26 3.1.2 Road Networking and Traffic around Suva Site ...... 29 3.1.3 Site 2 – Lautoka ...... 34 3.1.4 Road Networking and Traffic around Lautoka Site...... 34 3.2 PM2.5 Concentrations ...... 39 3.2.1 Gravimetric Analysis ...... 39 3.2.2 Real Time PM2.5 Concentrations ...... 43 3.2 Real Time Black Carbon Concentrations ...... 44 3.2.1 MicroAeth AE51...... 44 vii 3.3 Heavy Metal Characterization ...... 45 3.3.1 Calculation of Metal Concentration ...... 46 3.4 Statistical Analysis ...... 47 4.0 Results and Discussion ...... 48 4.1 PM2.5 Concentration ...... 48 4.2 Geography and Metrological Correlation ...... 55 4.2.1 Rain ...... 57 4.2.2 Wind...... 59 4.2.3 Relative Humidity ...... 64 4.2.4 Temperature ...... 64 4.3 Real Time Trends in PM2.5 and BC ...... 65 4.4 High Volume Sampler vs Real Time Data ...... 71 4.5 Elemental Composition of PM2.5 ...... 72 4.5.1 Element Concentrations ...... 72 4.5.2 Element Correlations ...... 80 4.5.3 Enrichment Factor...... 85 4.6 Implications and Policy Comparisons ...... 87 5.0 Conclusion and Recommendations ...... 91 Bibliography ...... 94 Appendix ...... 126 Appendix A ...... 126 viii List of Tables Chapter 1 Literature Review Table 1 Short Term and Long Term Health Effects of PM2.5 Chapter 4 Results and Discussion Table 2 Fine particulate matter concentrations categorized by months in the two cities Table 3 Number of buses operated and trips made by companies at the Lautoka Bus Station Table 4 Common signature elements and sources Table 5 Elemental concentrations in current study compared to roadside concentrations in some global cities. Table 6 Correlation between elements in Suva Table 7 Correlations between elements in Lautoka ix List of Figures Chapter 1 Introduction Figure 1 Projected Traffic Volume in Suva Nausori Corridor in 2014 Figure 2 Projected Traffic Volume in Suva Nausori Corridor by 2030 Chapter 2 Literature Review Figure 3 Size Distribution and Dominant Chemical Species Figure 4 PM2.5 concentration and health effect Figure 5 PM2.5 and Overall Air Quality Figure 6 Most Populous Areas in Fiji Chapter 3 Methodology Figure 7 Street view of Sampling Site 1 Figure 8 Satellite view of sampling Site 1 Figure 9 Typical Traffic Movement Rates as on Tuesdays (AM), 07 55 hrs. Figure 10 Typical Traffic Movement Rates as on Tuesdays (PM) , 17 25 hrs. Figure 11 Typical Traffic Movement Rates as on Sundays (AM) , 07 55 hrs. Figure 12 Typical Traffic Movement Rates as on Sundays (PM) , 17 25 hrs. Figure 13 Street view of sampling Site 2 Figure 14 Satellite view of sampling Site 2 Figure 15 Typical Traffic movement rates on Tuesdays at 08 30 hrs (AM peak hour). Figure 16 Typical traffic movement rates on Tuesdays, 18 05 hrs (PM Peak hour) . Figure 17 Cross- sectional view of the Size Selective Inlet (SSI) Chapter 4 Results and Discussion Figure 18 PM2.5 concentrations categorized by day at the sampling sites Figure 19 Roadside PM2.5 concentrations in some global cities x Figure 20 Air quality during sampling days in Suva City (A) and Lautoka City (B); Health effects on population as per the sampling days in Suva City (C) and Lautoka City (D) Figure 21 Correlation Plot of PM2.5 and Meteorological variables in Suva (A) and Lautoka City (B) Figure 22 Altitude Map of Viti Levu Figure 23 Difference in cloud buildup between western and eastern sides of Viti Levu Figure 24 Wind Rose of Sampling in Suva (A) and Lautoka City (B) Figure 25 (A) Typical Backward wind trajectories arriving at Suva City and (B) movement across the Suva Peninsula Figure 26 (A) Typical Backward wind trajectories arriving in Lautoka City and (B) movement across Lautoka Figure 27 Diurnal Variation of PM2.5 and BC concentrations in Samabula (A), Suva City (B), Reservoir Road (C) and Lautoka City (D) Figure 28 Visible smoke emissions from buses in Lautoka City Figure 29 Location of Lautoka bus station and surrounding buildings Figure 30 Backward wind trajectories arriving from the SW and NW quadrants Figure 31 Element to element ratios highlights elements that have greater concentration in Lautoka in comparison to Suva and vice-versa Figure 32 Enrichment Factor (EF) of elements in Suva and Lautoka cities xi 1.0 Introduction Overview This chapter introduces air pollution as an environmental and health risk factor in the Pacific Island Countries (PICs), particularly, Fiji. It highlights vehicle emissions as a leading source of fine particulate matter (PM2.5) in urban areas. This chapter also argues on the strong rise in vehicle numbers, urban development and unavailability of PM2.5 related data in heavy traffic areas as a motivation for this research. Finally, it presents the aim and objectives of this thesis. 1.1 Background and Significance Air pollution has become the leading environmental risk factor, affecting the environment, climate and public health. Today, one third of deaths due to stroke, heart disease, and lung cancer are attributed to air pollution (Karagulian et al., 2015; World Health Organisation, 2018). Urban areas in particular are experiencing rapid deterioration of air quality and facing greater health repercussions. In urban environments, the dominant air pollutant is usually fine particulate matter or PM2.5 (Mainka & Zajusz-Zubek, 2019). These are microscopic particles having an aerodynamic diameter of 2.5 micrometers or less (Marcazzan et al., 2001). The effects of PM2.5 on the human health and the range of sources has made PM2.5 the primary marker of air quality (Ott et al., 2008). Road traffic is a major source of fine particulate matter in urban areas in both developed and developing countries (Janssen et al., 2011; Thorpe & Harrison, 2008). Motor vehicles release PM2.5 primarily through exhaust and non-exhaust emissions. Exhaust emissions include direct and tailpipe emissions. Non-exhaust emissions are emissions which arise from mechanical abrasion including road, brake and tire wear as well as road dust re-suspension (Amato et al., 2014). When combined, these processes are one of the major sources of PM2.5 in many cities globally. Hence, the continuous monitoring of PM2.5 is a requirement in many countries and pollution 1 mitigation strategies are designed using the data collected (Baumann et al., 2006). This is vital for improving the longevity and quality of life of the general population. However, the lack of data as a result of unavailable expertise and lack of instruments makes it difficult to deduce the levels of PM2.5. This also impedes the design of proper and tailored mitigation strategies in many developing countries. Moreover, weak pollution control policies and unregulated development possibly further aggravates the effects of traffic related PM2.5 in such countries. The Pacific Island Countries (PICs) are such examples where data about PM2.5 and air quality in general are almost non-existent. Moreover, a huge influx in second hand, re- conditioned and used vehicles are further adding to underestimated air quality issues in the PICs. The increase in the number of cheap second hand vehicles is causing greater traffic jams and increased air pollution as well as damaging existing road infrastructure in the PICs (Pacifc Community, 2017). In emphasis, it is now argued that the PICs such as Samoa and Fiji have become dumping grounds as it imports high numbers of vehicles that developed countries do not want (Sanerivi, 2017; Sevura, 2005). Fiji is one of the fastest developing countries out of all the PICs, serving as the hub of the Pacific. The two cities in Fiji, namely, Suva and Lautoka are arguably more economically active than most towns and cities in the South Pacific Islands. A rise in the number of second hand and new vehicles have been noted in Suva and Lautoka, however, the environmental impact of this growth has only marginally been examined. To date, only one study to evaluate ambient PM2.5 in Suva City has taken place while no data exists for Lautoka City or for anywhere else in Fiji. The prior study done in Suva City determined the concentration of ambient PM2.5 to be 7.4 μg/m3 (Isley et al., 2017). However, some attributes of this study provides limitations for it to be used to consider the concentrations that the general population is exposed to. Isley et al. (2017) reported low concentrations, likely attributed to the higher sampling height of this study (18 m above ground, on a building located at Suva wharf). This work (Isley et al., 2017; Isley et al., 2018) noted low concentrations and a large influence of marine aerosols in the composition of fine particulate matter. The study showed that marine aerosols were the largest contributor to PM2.5 mass with high amounts of Na, Ca and K ions. The current study distinguishes itself from the previous 2 study by focusing on ground level roadside and traffic related emissions in areas of high population density and activity. Information about fine particulate matter concentrations at ground level, especially in dense traffic areas is therefore yet to be fully examined in Fiji. The absence of such air quality data and the rapid unchecked development sets the foundation of this project. Presently, there has been significant influx in the number of vehicles being imported. The intensity and duration of traffic has also significantly increased, implying a possible rise in PM2.5 emissions and consequently, greater apprehension on people’s well-being. Traffic congested areas are avenues of high exposure and unavailability of ground level PM2.5 data in these two cities has delayed air quality legislations and policies, leaving the people’s health and environment at risk in Suva and Lautoka City. The need for an assessment of PM2.5 is further substantiated by the rising risk and cases of public health issues in Fiji. In recent years, Fiji has noted greater health concerns such as higher asthma and “wheezing” rates (Isley et al., 2017). Fiji has the highest rate of asthma related mortality in the world when considering age normalized asthma related deaths (The Global Asthma Network, 2018). There were more than 2200 asthma related hospitalization at the major hospitals in Fiji between 2010 to 2014, 16% of which resulted in mortality (Ministry of Health and Medical Services, 2017). Many studies have associated elevated PM2.5 concentrations with increased asthma related hospitalizations (Tecer et al., 2008a; Xing et al., 2016). Non-communicable diseases such as asthma comprise most of the top ten causes of mortality and disability in Fiji (Asian Development Bank, 2017). Fiji also has one of the highest rates of diabetes related amputations and cardiovascular disease cases in the world (Chand, 2018; Morgan, 2015). Both diabetes and cardiovascular diseases have been positively associated with long term air pollutant exposures including PM2.5 (De Marco et al., 2018; Fiordelisi et al., 2017; Pearson et al., 2010). While quantitative associations of air pollution and health effects are beyond the scope of this research, air quality should be acknowledged as possible contributing sources to the prevalence of these diseases. 3 There is a need to establish baseline data in Suva and Lautoka City as road traffic emissions continue to increase. Reports published by the Fiji Roads Authority (2014) highlight that the traffic build-up in Suva is quite heavy and is only going to get worse with projections from 2014 to 2030, as depicted by Figures 1 and 2 respectively. The report, based on in-person (manual) traffic counts and modelling, shows the traffic volume in traffic prone areas of Central Business District (CBD) and the Greater Suva Area (GSA). Moreover, the report further highlights the strong growth in the number of vehicles involved in peak hour traffic from 36, 700 vehicles per hour (vph) in 2014 to 43, 5000 vph by 2030. This growth in traffic would prolong the journey through these roads by 25% (Fiji Roads Authority, 2014). This would mean more time spent in traffic, and greater exposure to emission related particulate pollution. Hence, it is essential that PM2.5 monitoring programmes be undertaken immediately to establish baseline data which would provide a comparative platform to determine the impacts, sustainability and feasibility of current and future developments. 4 Figure 1 Projected Traffic Volume in Suva Nausori Corridor in 2014 Source: (Fiji Roads Authority, 2014) 5 Figure 2 Projected Traffic Volume in Suva Nausori Corridor by 2030 Source: Fiji Roads Authority, 2014 6 1.2 Aim The central goal of this project is to determine the concentration and heavy metal composition of fine particulate matter (PM2.5) in dense traffic areas in Suva and Lautoka cities, in order to assess the air quality and contributions from vehicle traffic to the air pollution in Fiji 1.3 Objectives I. To develop EPA standard local capacity to measure fine particulate matter (PM2.5) at the University of the South Pacific. II. To assess daily averages of PM2.5 and metal compositions of PM2.5 in traffic laden areas in Suva and Lautoka City. III. To investigate diurnal patterns of PM2.5 and Black Carbon (BC) using real time PM2.5 and BC sensors and correlating the diurnal trends to traffic patterns. 7 2.0 Literature Review Overview This chapter provides literature on key concepts and arguments that are presented in this research. It presents the definition and effects of PM2.5 as well as information on common sources and dispersion factors. This chapter also highlights available information about air quality in Fiji and evidences of developments that may conversely deteriorate it. 2.1 Definition of particulate and composition Particulate matter has been more broadly defined as a combination of small solid particles and liquid droplets that are dispersed in the atmosphere. These particles may be a mixture of visible particulates, such as dust, soot, smoke and dirt or the combination of compounds and elements, invisible to the human eye (US Environmental Protection Agency, 2004). More specifically, particulate matter is a composition of different chemical compounds fused together, largely, at the source of emission or by undergoing reactions in the atmosphere. Total Suspended Solids (TSP) is a general term used to describe solid particles and liquid droplets in the air with varying sizes (Sezer Turalıoğlu et al., 2005). However, specific terms have been given to particulate matter of specific sizes. Particulate matter is usually divided into two main categories, PM10 and PM2.5, on the basis of its size. PM10 is defined as particles having an aerodynamic diameter of less than 10 μm while PM2.5 or fine particulate matter are particulates with an aerodynamic diameter less than 2.5 μm (Caiazzo et al., 2013; Marcazzan et al., 2001; Martuzevicius et al., 2008; Wang et al., 2013). While these two sizes remain the primary subjects, particulate matter ranging in between 10 μm to 2.5 μm, known as coarse particulate matter, is also increasingly being included in many research (Adar et al., 2014; Ljubimova et al., 2018; Ott et al., 2008). 8 While both PM10 and PM2.5 sized particles are inhalable, PM2.5 has been more strongly associated with ill health impacts than coarse fractions, hence remains one of the primary markers of air quality (Ott et al., 2008). In addition, PM2.5 has been noted to have the most varied composition, being a good marker for anthropogenic activities, as demonstrated by Figure 3. Figure 3 Size Distribution and Dominant Chemical Species Source: (Watson et al., 1998) The composition of PM2.5 is largely dependent on the source of emission. It is also used to identify and quantify the contribution of each source to the overall contributions of particulate matter. This is vital in understanding the chemical processes that these particles undergo in the atmosphere (Jeong et al., 2016). In general, PM2.5 comprises largely of carbonaceous compounds such as organic carbon (OC), elemental carbon (EC), Black Carbon (BC) and Water Soluble Organic Carbon (WSOC) which typically amounts to more than 50% of the PM mass (Chow et al., 2002; Harrison et al., 2004; HEI Panel on the Health Effects of Traffic Related Air Pollution, 2010; Tao et al., 2013). The total carbon content (TC) in particulates is usually categorized into organic carbon and elemental carbon. Elemental carbon (EC) is usually less than one fifth of the TC, 9 having light absorbing nature. Black Carbon or BC is also generally known to contribute strongly to the light absorbing characteristic of aerosols (Rattigan et al., 2010). This leads to some overlap between EC and BC. While not certain, general consensus has been made to distinguish EC and BC depending on the “operational measurement methods” (Briggs & Long, 2016). Briggs and Long (2016) also suggest that BC can be defined as carbonaceous particles with strong light absorptive properties, measured using light attenuation methods while EC is “operationally defined” based on thermal properties. However, these definitions fail to provide clear distinction between EC and BC. Petzold et al. (2013) argue that no precise and comprehensive terminology exits within the scientific community to exactly quantify carbonaceous matter in atmospheric aerosols. Current terminologies are limited or bound by either the specific property being examined or by the method of examination of the carbon in the aerosol. Black Carbon is usually of significant interest due to its toxicity and subsequent health impacts (Janssen et al., 2011). Reports by the World Health Organisation (Janssen, 2012) highlight that BC acts as a carrier of varying combustion driven chemical compounds which may be of varying toxicity to sensitive targets in the human body such as the lungs and major defence cells. Black Carbon is formed by incomplete combustion of carbon based compounds. It is a strong indicator of the negative effects of PM2.5 and is largely monitored together with PM2.5 itself (Targino et al., 2016). BC is also the paramount light absorbing species in the atmosphere, and largely affects the earth’s temperature and climate through aerosol radiative forcing. This phenomenon alters the radiative properties of earth and has been acknowledged as the second most important component in global warming, second only to CO2 when considering direct forcing (Jacobson, 2001). The unique surface properties of BC also makes it a good adsorption site for many semi-volatile organic compounds such as Poly Aromatic Hydrocarbons (PAH’s), thereby enhancing its overall carcinogenic and toxic properties (Viidanoja et al., 2002). This has important influence on the overall PM2.5 properties in comparison to other constituents. PM2.5 further constitutes of water soluble as well as inorganic and organic ions, such 2- - + - 3- 2- 2+ + + 2+ - - - as SO4 , NO3 , NH4 , NO2 , PO4 , C2O4 , Ca , Na , K , Mg , Cl , F , HCOO , - CH3COO and signature elements including heavy and trace heavy metals (Wang et 10 al., 2006). These signature elements are used to determine the contributions of the different sources of PM2.5. For example, high concentrations of Cl, Na, Mg, K and Ca may indicate a presence of sea aerosols. The presence of S, K, H, Zn, Ba, Fe, Pb, Cu, Zn, Al and Mn largely indicate combustion and traffic associated activities and emissions (Gillies et al., 2001; Heal et al., 2005; Isley et al., 2017; Jeong et al., 2016; Shakya et al., 2017; Zhang et al., 2017). However, while the elemental and metal constituent is essential in identifying the sources of emission, it is these components that exacerbate the health impacts of PM2.5. 2.2 Health Impacts and Exposure Limits of PM2.5 Air pollution has become the leading environmental cause of mortality globally. According to the World Health Organisation (2018), fine particulate matter has the greatest impact on human health in comparison to other air pollutants. In emphasis, 1 in 8 deaths worldwide were attributed to air pollution in 2016, with 91% of the world’s population breathing unclean air in the same year (World Health Organisation, 2018). The effects are more pronounced in children and young adults due to their weak immune system and developing bodily function, therefore, potentially compromising future generations (Kurt et al., 2016). Consequently, countries continue to research on the pathways in which fine particulate matter affects health and intervention measures that would curb the health deterioration mediated by PM2.5. The health impacts of PM2.5 highlighted by studies in the past three years is summarised by Isley et al. (2018). To demonstrate further, Table 1 outlines some key findings about PM2.5 and its association with disease and mortality targeting specific bodily functions, published in the last two years. 11 Table 1 Short Term and Long Term Health Effects of PM2.5 Subject of Exposure Study Affected System Findings Interest range Short term exposure to increased PM2.5 concentrations increased the odds Increased of receiving treatment for Acute Lower Respiratory Infections (ALRI). (Horne et al., PM2.5 Short term Respiratory Primarily, bronchiolitis was seen to be dominating the very young 2018) exposure affected population (0-2 years) while influenza dominated the old. The primary etiologic agent determined was Respiratory Syncytial Virus (RSV). PM2.5 (Pirozzi et al., Short term exposure to PM2.5 (above 12 μg, for 4-5 days) was positively Ozone 2018) Short Term Respiratory associated with increased mortality, emergency department visits and Nitrogen hospitalisation due to pneumonia and severe pneumonia for older adults. Oxide Exposure to PM2.5 significantly associated with lower percent predicted (Bergstra et PM2.5 Long Term Respiratory Peak Expiratory Flow (PEF) of 2.80%. Also, significantly associated with al., 2018) NOx coughing and wheezing. Approximately 1 in 7 lung cancer associated mortality can be attributed to (Guo et al., PM2.5 Long Term Respiratory PM2.5. Increased exposure to PM2.5 is associated with lung cancer and 2017) death, especially that containing Poly Aromatic Hydrocarbons (PAH’s). 12 3 Exposure to increased ambient PM2.5 concentrations by 10 μg/m was (Gharibvand PM2.5 Long Term Respiratory positively associated with a 31% increased incident lung et al., 2017) Adenocarcinoma (AD)- a lung cancer subtype Exposure to PM2.5 gives rise to reactive oxygen species (ROS), which (Deng et al., PM2.5 Long Term Respiratory through cellular processes promotes the malicious growth of lung cancer 2017b) cells and tumour progression. PM2.5 resulted in autonomic imbalance. The constituents of fine (Chen et al., particulate matter further triggered thrombogenecity. PM2.5 that is PM2.5 Short Term Cardiovascular 2017) deposited in the pulmonary alveoli alters Heart Rate Variability (HRV) immediately. Exposure to elevated levels of PM2.5 led to significant increase in the (Zhang et al., PM2.5 heart rates of patients with cardiovascular disease. The lengthening of PR, Short Term Cardiovascular 2018c) Ozone QRS and QT intervals was also observed. Short term exposure to PM2.5 affects acute cardiovascular disease. Increases risk of myocardial infarction (MI) and cardiac arrhythmia. Long (Fiordelisi et PM2.5 term exposure increases risk of coronary events and accelerates the Long Term Cardiovascular al., 2017) PM10 development of cardiometabolic disorders. Long term exposure is also associated with heart failure related mortality. 13 0.6% of all cardiovascular disease and 1.5% of all respiratory disease, (De Marco et Cardiovascular / 18.1% of all Ischaemic Heart Disorder and 9.2% of Chronic Obstructive PM2.5 Long Term al., 2018) Respiratory disease Pulmonary Disease in 2015-2016 in Metropolis of Rome was attributed to PM2.5 exposure. 3 (Gutiérrez- An increase of 10 μg/m of PM2.5 was associated with an increased Cardiovascular / Avila et al., PM2.5 Short Term cardiovascular and cerebrovascular mortality by 1.22% and 3.43% Cerebrovascular 2018) respectively. Exposure to PM2.5 significantly increased moderate-to-severe depressive symptoms. Short term exposure was also significantly associated with (Pun et al., Cognitive function/ PM2.5 Short Term incident anxiety symptoms. PM2.5 may affect mental health by increased 2017) Cerebrovascular neuroinflammation, oxidative stress, cerebrovascular damage and neurodegeneration. (Lu et al., PM2.5 Cognitive function Short term exposure to PM2.5 was significantly associated with delirium Short Term 2017) SO2 and Mental Health in surgical adults Exposure to PM2.5 causes decline in cognitive function, particularly, (Tallon et al., PM2.5 Cognitive Functions through mediated by depression in older adults. PM2.5 exposure was Long Term 2017) NO2 and Mental Health associated with 0.22 point decrease in CCFM scores, indicating equivalent aging of 1.6 years. Long term PM2.5 exposure results in neurotoxic effects which increases (Younan et Cognitive functions PM2.5 Long Term delinquent behaviour of urban dwelling adolescents undergoing al., 2018) and Mental Health psychological adversities. 14 Long term PM2.5 exposure of women in gestational periods were found to Neurodevelopment cause olfactory dysfunction, serving as an early marker to greater effects (Winckelmans Long Term PM2.5 and central nervous of the central nervous system in the born infants. Short term exposure (1 et al., 2017) Short Term system month) of women in gestational periods altered expressions of neuroactive ligand receptor interaction gene members. Exposure to PM2.5 after Hepatocellular Carcinoma (HCC) diagnosis (Deng et al., resulted in shortened survival. Higher exposures resulted in stronger PM2.5 Long Term Liver Function 2017a) decrease in survival period. PM2.5 may create oxidative stress, liver inflammation, steatosis and genotoxicity, mediating liver cancer. (Xu et al., Exposure to PM2.5 is associated with increased rapid decline in renal PM2.5 Long Term Renal/ Urinary 2018) function and increased risk of membranous nephropathy PM2.5 exposure results in a decrease in sperm quality and quantity. Exposure is also associated with increased infertility. In pregnant women, (Zhang et al., Long Term PM2.5 Reproductive system it has been associated with intra uterine growth retardation, low birth 2018a) Short Term weight of foetuses. PM2.5 may lead to reproductive toxicity in both and females. (Mazidi & Exposure to PM2.5 was significantly associated with prevalence of type 2 Speakman, PM2.5 Long term Overall diabetes. 2017) 15 The health ramifications induced by high exposure to PM2.5 has also caused economic stress in many countries, if not globally. The impact of ambient air pollution, largely led by PM2.5 was estimated to cost the OECD countries around USD 1.7 trillion in 2010. In the same year, the cost of health impacts due to air pollution in China and India was estimated to be USD 1.4 trillion and USD 0.5 trillion respectively (OECD, 2016). By 2015, these costs had substantially increased to almost USD 1.6 trillion for China and USD 0.8 trillion for India, equating to 8.4% and 11.6% of its annual GDP respectively (Roy & Braathen, 2017). Approximately 90% of these costs were attributed to fine particulate matter. The increase in morbidity and mortality due to exposure to PM2.5, further lowers the productivity of labor, thereby causing stress on all sectors of the economy. According to the World Bank (2016), lowered productivity due to lost labor cost the global economy USD 225 billion in 2013. These figures continue to rise as countries struggle to limit the concentration and exposure levels of PM2.5 and to meet the WHO exposure guidelines. The World Health Organization, in 2005, revised and laid out stringent guidelines in an effort to limit high exposure to PM2.5. These guidelines set the limit of 24 hour daily 3 average concentration of PM2.5 to be less than 25 μg/m and the annual mean to be equal or less than 10 μg/m3 (World Health Organisation, 2006). The United States Environmental Protection Agency (EPA), after tremendous revisions, followed the WHO, albeit setting out slightly lower standards, at 35 μg/m3 as the daily 24 hour mean and the annual mean at 12 μg/m3 (Esworthy, 2015). The Environment Protection Authority Victoria or EPA Victoria, the environmental regulatory and watchdog institution of the Australian State of Victoria, has set out much more rigorous standards, limiting the daily averages to 25 μg/m3 but the annual mean to be at 8 μg/m3. The Australian air quality standards also divide the concentrations in groups relative to both the health effects and air quality in general as shown by Figures 4 and 5. 16 Figure 4 24 hour mean PM2.5 concentrations and exposure effect on selected population group Source:(Environment Protection Authority Victoria, 2018) Figure 5 PM2.5 and Overall Air Quality Source:(Environment Protection Authority Victoria, 2018) Despite these thoroughly reviewed guidelines, a large number of countries struggle to achieve these set targets. In emphasis, Kelly and Fussell (2015) argue that in spite of the guidelines and monumental research, urban areas worldwide are experiencing deteriorating air quality and mitigation efforts have largely been stalled. A plausible explanation would be that the WHO guidelines are largely meant for policy formulation and the fact that it is not a legally binding legislation may contribute to delayed action in some countries (Krzyzanowski & Cohen, 2008). More importantly, the lack of technologies, especially in developing countries and the expansive range 17 of sources makes it practically much more difficult to curb fine particulate matter (Bachmann, 2007). 2.3 Sources of PM2.5 Fine Particulate matter, or PM2.5 is produced through an array of sources including industrial processes, vehicular and residential combustion, biomass burning, road dust, mechanical resistance and granulation, agricultural sectors, coal combustion, electricity production, secondary nitrate and sulfate formation, secondary organic aerosols (SOA) formation and natural sources such as weathering, wind erosion and sea sprays (Chuersuwan et al., 2008; Harrison et al., 1997; Heo et al., 2009; Lee et al., 2003; Masri et al., 2015; Qin et al., 2006; Tsapakis et al., 2002; Tucker, 2000; Zheng et al., 2002). The European Union suggests five key economic sectors, (Commercial and Residential, Transport, Industry, Agriculture, Waste and Energy ) to be responsible for almost all the PM2.5 related air pollution in many countries (European Environment Agency, 2019). Developing countries such as India and the South African nations have a larger contribution from the residential sector, suggesting greater biomass burning. On the other hand, the transport sector is seen to be the key contributor to the overall PM2.5 in the European Union countries and Japan (Karagulian et al., 2016). According to Kurt et al. (2016), approximately 20% of air pollution related mortality can be attributed to traffic related PM in Germany, United Kingdom and United States. Other studies demonstrate that as much as 25% of urban ambient air pollution due to PM2.5 is produced by traffic (Karagulian et al., 2015). Traffic related emissions from the transport sector is the key source in many urban environments globally (Ferm & Sjöberg, 2015; Harrison et al., 1997; Targino et al., 2016; Zhang & Batterman, 2013; Zhang et al., 2017). Traffic related PM2.5 emissions are a huge contributor to fine particulate matter concentrations. Tail pipe and direct exhaust emissions from vehicles are the largest primary contributor to PM2.5 from the transport sector, followed by emissions from wear of brakes and clutches (Ketzel et al., 2007). Apart from primary emissions, other gaseous pollutants from the transport sector also contribute to the overall ambient 18 PM2.5 concentrations through secondary formation pathways. These include NOx precursors which in time form particulate nitrates and Volatile Organic Compounds (VOC) precursors, which ultimately form particulate organic carbon (Hodan & Barnard, 2004). VOC’s emitted through tail pipes, such as formaldehyde, acetaldehyde, benzene and 1, 3 butadiene also increase the toxicity of PM2.5 and cause adverse impacts on health (Davis et al., 2007; Hime et al., 2018). Other compounds such as black carbon and heavy metals also form large proportion of traffic emitted PM2.5 and are known to have detrimental health consequences (Heal et al., 2005; Shakya et al., 2017). Furthermore, more recognition must be given to traffic as a dominant and nocuous source as high numbers of people spend considerable amount of time on or near roads on a daily basis (HEI Panel on the Health Effects of Traffic Related Air Pollution, 2010). This directly exposes them to fine particulate matter generated by vehicle exhaust, which is more toxic when in comparison to other non-exhaust sources (Hime et al., 2018). More importantly, in traffic microenvironments, the most vulnerable groups, including children, old aged adults and pregnant women are exposed, resulting in more pronounced health repercussions such as asthma, respiratory illness, cancer and cardiovascular diseases (Guo et al., 2017; Horne et al., 2018; Kelly & Fussell, 2015; Liu et al., 2017; Pirozzi et al., 2018; Tallon et al., 2017; Winckelmans et al., 2017; Zhang et al., 2018c). As a consequence, the attribution of traffic towards ambient air pollution and particulate matter concentrations makes it a key research and monitoring problem. While global data for larger countries are available, understanding traffic related air pollution in developing countries is still a challenge. In addition, many studies also argue that the impact of particulate matter and air pollution in general are much higher in developing countries (Cohen et al., 2005; Massey et al., 2009; World Health Organisation, 2018). Focus is now being given to developing states like India, China and Africa, however the developing Pacific Island Countries (PICs) are yet to be fully explored (Wu et al., 2017). The scarce available literature on ambient PM2.5 concentration in PICs such as Fiji has the potential to offset the sustainable development goals as these countries gain momentum in developing. 19 2.4 PM2.5 in a PIC: Fiji PM2.5 data in Fiji remains largely unavailable, except for data collected by Isley et al. (2017) who provided ambient PM2.5 concentrations, measured at Suva wharf during a field campaign between 2013 to 2014. Apart from the current study, no previous studies have been conducted to investigate air quality in heavily trafficked areas in the pacific island nation of Fiji. Fiji is the most diverse and fastest growing economy in the PICs (The World Bank, 2018). It boasts more development and stronger participation in global trade, making it an exception in comparison to other PICs (Juswanto, 2016). The rapid economic and population growth in Fiji is inevitably pushing towards greater urbanization. The 2017 census data states that 55.9% of the 884 887 Fijians now reside in urban areas. The urban population has increased by almost 10% in the last two decades. In emphasis, Figure 6 outlines that a total of 243 795 people (27.55% of the population) reside in densely populated areas of Suva City, Nasinu and Nausori alone (Fiji Bureau of Statistics, 2018a). The urban population boom and strong economic growth have now become key drivers in increasing car ownership per household in Fiji. The number of vehicle registrations, in the past 5 years has increased by more than 40%, standing at 117 561 vehicles at the end of 2017 (Fiji Bureau of Statistics, 2018b). The increase in motor vehicles in Fiji can further be attributed to the 53.13% decrease on motor vehicle import duty conceded by the Government of Fiji (The Fijian Government, 2010). It further reduced the fiscal and import duty from 15% to 5% in 2016-2017 and to zero fiscal and import excise duties in 2017-2018 on new passenger hybrid and electric vehicles brought into Fiji, allowing for more vehicle imports (Fiji Revenue and Customs Authority, 2016, 2017). These factors have led to the substantial influx in the number of vehicles in Fiji. 20 Key Urban Centres Population Count 2017 Census Suva – 93 874 Lautoka --71 573 Nasinu – 92 046 Nadi - 70 533 Nausori – 57 886 Figure 6 Most Populous Areas in Fiji The main island of Viti Levu is home to a large proportion of the population. The population density is the highest between the Suva-Nausori corridor and Nadi-Lautoka Corridor. When combined, these areas are home to 385 892 people or 43% of the total population. 21 The significant rise in the number of vehicles has huge impacts on traffic congestion and travel time (Fiji Roads Authority, 2014). Steady prolongation of both traffic range and intensity as well as travel time have been observed during peak hours in the two cities in Fiji, namely, Lautoka and Suva city (Deo, 2018; Kumar, 2017). Apart from the loss of productivity due to delays, the underlying threat imposed on the affected population by these traffic congestions is potentially much greater. Traffic congestion results in higher tail pipe emissions including fine particulate matter and greenhouse gases (GHG’s) (Grote et al., 2016). The emissions are further increased by the use of very old buses as the main source of transport. Reportedly, buses in Fiji have very high emissions as 97% use fuels below Euro IV standard. Furthermore, more than 50% of these buses are more than 20 years old and 25% are more than 30 years old, producing visible smoke emissions (Haworth & Starkey, 2017). Laws and regulations on air pollution have largely failed to stagnate or to reduce emissions, irrespective of the source of emission in Fiji. While air pollution has been noted as an important environmental issue by the Department of Environment (DoE), Fiji is yet to implement a National Air Quality Policy, inclusive of fine particulate matter standards (Isley & Taylor, 2018; United Nations Environment Programme, 2015). Existing policies such as the Environmental Management Regulations (2007) have provision for emission license for industries, while plans such as the National Air Pollution Control Strategy (endorsed under the Environment Management Act (2005)) layout long term strategic framework for air pollution control (The Government of Fiji, 2012; United Nations Environment Programme, 2015). However, there are no watchdog organizations with the capacity to provide continuous monitoring and assessment of the emissions from different sectors and therefore the implementation process is almost non-existent. For instance, Isley and Taylor (2018) argue that despite laws which place a FJD $10 000 penalty on burning household garbage without a license, surveys show that more than half the population living in Suva burn household and green waste. This undoubtedly demonstrates the lapses between policy formulation and enforcement and is common throughout the Pacific Islands and other developing countries. In terms of vehicle emissions, the transport regulatory body of Fiji, the Land Transport Authority (LTA), (under the Land Transport Act of 1998 and Land Transport (Vehicle 22 Registration and Construction) Regulations (2000)), has the authority to issue Traffic Infringement Notices (TIN’s) and or Vehicle Defect Notices for vehicles that emit excessive smoke for greater than 10 seconds (High Court of Fiji, 2004; Parliment of the Republic of Fiji, 1998). However, these measures have largely failed to address the issue of vehicle emissions. Numerous reports demonstrate that “smokey” vehicles are quite widespread in Fiji (Sevura, 2005). More recently, the LTA revised the acceptable smoke opacity level from 70% to 50% in an effort to reduce “visible” vehicular smoke emissions, however, this factor is only considered during vehicle inspection for the permit renewal process (Pratibha, 2017). Multiple assessments through various stakeholders have highlighted that the laws surrounding vehicle emissions standards are inadequate and poorly enforced and regulated (Isley et al., 2016; Secretariat of the Pacific Regional Environment Programme, 2014; Sevura, 2005). The lax enforcement measures and the increase in vehicle count may be critical contributing factors in the increase in fine particulate matter concentrations in Fiji. Moreover, the different climate conditions in the two cities under study may favor reduced dispersion and particulate accumulation, further contributing to increase in air pollution. 2.5 Dispersion Factors 2.5.1 Wind speed and direction Wind speed and direction have extensive influence on the dispersion and migration of fine particulate matter. The wind speed has a negative effect on PM2.5 concentrations (Harrison et al., 1997). Greater wind speed causes a faster migration and dispersion of pollutants, thus diluting it (Zhang et al., 2015). Similarly, wind direction also strongly influences both the temporal and spatial distribution of PM2.5 (Guerra et al., 2006). Studies focusing on traffic emissions suggest that depending on the geographical location, upwind location may show lower concentrations of PM2.5 in comparison to downwind locations from the source. Notably, it also takes greater time for the pollutants such as fine particulate matter to be dispersed and drop to background levels at downwind sites in comparison to upwind sites (Hitchins et al., 2000; Hu et al., 2009). 23 2.5.2 Rainfall Rain has a negative effect on the concentrations of PM2.5. Rainfall has been noted to “washout” and “scavenge” (wet deposition) fine particulate matter, thereby decreasing the concentrations. (Tecer et al., 2008b; Wang & Ogawa, 2015). This process, although leading to lower concentrations of PM2.5 in the atmosphere, contributes heavily to rain water pollution (Ouyang et al., 2015). This may have further implications as heavy metals from primary emissions of PM2.5 and sulphur as well as nitrogen related compounds formed from secondary PM2.5 formation are removed due to their solubility in rain water. The polluted rain water further increases concentrations of heavy metals in urban soils and also contaminates water sources, thereby exacerbating public exposure and health risk (Ouyang et al., 2015) . 2.5.3 Relative Humidity Humidity has been noted to have a curvilinear (inverted U) relationship with PM2.5 concentrations (Lou et al., 2017). At lower humidity levels, typically lower than 70% relative humidity (RH), particles exhibit hygroscopic behaviour and tend to accumulate, giving a higher concentration of fine particulate matter. At higher levels of humidity (70-100% RH), greater hygroscopic growth causes particles to become too heavy to retain aerodynamic movement and thus dry deposition occurs. This leads to a decrease in PM2.5 concentrations (Lou et al., 2017; Wang & Ogawa, 2015). Hygroscopic components of aerosols, such as sulphates, nitrates and water soluble carbonaceous species, which also account for large proportions of PM2.5, can have increased water uptake in high humidity conditions (greater than 70% RH). This leads to an increase in the light scattering ability of PM2.5, which is commonly associated with haze formation resulting in impaired atmospheric visibility (Chan et al., 1999; Chen, 2014). 24 2.5.4 Temperature Generally, temperature has been positively correlated with concentrations of PM2.5 (Tai et al., 2010; Wang & Ogawa, 2015). Temperature, unlike other meteorological factors which mainly affect the dispersion, plays a critical role in secondary PM2.5 formation by strongly influencing tropospheric gas phase reactions (Aw & Kleeman, 2003). Components such as Organic Carbon (OC) and Elemental Carbon (EC) increase with rising temperatures due to increased gas phase partitioning and faster gas-to-particle conversion. Increase in temperature also promotes oxidation of SO2, supplemented by temperature dependent rate constants and greater concentrations of oxidants (Dawson et al., 2007; Tai et al., 2010). These factors lead to greater organic and sulfate particulates, which in turn increase the overall PM2.5 concentrations. However, the contrary is seen with respect to nitrates at elevated temperatures. Ammonium nitrate shows increased volatility with rise in temperature, favoring nitric acid formation, thereby decreasing particulate nitrate concentrations (Aw & Kleeman, 2003). Notably, these temperature related contradictory responses of sulfates and nitrates equipoises any substantial change to the ambient fine particulate matter concentrations. 25 3.0 Methodology The following section presents information about the sampling areas and the sampling techniques. It outlines the methods and selection used to carry out the research, including sampling methods, instrumentation and data analysis. 3.1 Sampling Site Concentrations of fine particulate matter were monitored in heavy traffic areas in Lautoka and Suva City. These are the only two cities in Fiji and have a quite high population density as well as strong economic activity. 3.1.1 Site 1 – Suva The field campaign in Suva City was undertaken at the Fiji National University’s (FNU) Derrick Campus in Samabula (latitude: -18.1262, longitude: 178.4402). The University sits at the junction of a total of 14 lane crossroads, and is one of the busiest intersections in Suva. It is located at the intersection of major roadways, particularly, Princess Roads (4 lane), Ratu Mara Road (4 lanes), Edinburgh Drive (3 lanes), Waimanu Road (3 lanes), at an elevation of 60 m above sea level. Figures 7 and 8 provide street and satellite view of the site respectively. It is an area of high population density and surrounded by many public agencies and buildings. The sampling site is located 1.15 km away from Fiji’s largest hospital, the Colonial War Memorial Hospital. A number of institutions such as the Gospel High School, Suva Primary School, Samabula Primary School, banks, Australian High Commission, Residence of Indian High Commissioner, Five Princess hotel, Calvary Temples, Hindu Temples, Police Stations, bakeries, restaurants and numerous businesses and sales shops are located in a 500 m radius. This makes it one of the busiest areas and with the heavy traffic. Furthermore, the site is located 2 kilometers away from the capital city of Suva. 26 Sampling Site Figure 7 Street view of Sampling Site 1 27 Sampling Site s Figure 8 Satellite view of sampling Site 1 28 3.1.2 Road Networking and Traffic around Suva Site The Fiji National University campus in Samabula is positioned in the middle of key road networks and used largely to commute to and out of main Suva Central Business District (CBD) by residents of Nausori and Nasinu. Traffic mainly flows through the Kings Road from Nasinu and Nausori, through Ratu Mara Road, passing in front of the site, and using Edinburgh drive to the main city. A nearby Land Transport Authority traffic sensor indicates vehicle counts of around 27 000, approximately 89% of which are passenger vehicles. The number of vehicles passing the site, however, is much greater considering contributions to traffic from Princess Road which is also a another roadway between Suva and Nausori. Princess Road serves as an alternate route to Nausori and is the major link to many residential area networks of Greater Suva Area (GSA). Increase in traffic intensity also causes traffic to move much slower and increases the total journey time. Currently, traffic movement is the slowest along Princess and Kings Road during Peak traffic. Figures 9, 10, 11 and 12 generated by Google Maps highlight the rate of traffic movement during the morning and evening peak hours of Tuesday (weekday) and Sunday (Weekend) respectively. These maps correlate to the observed traffic patterns on the ground. 29 Figure 9 Typical Traffic Movement Rates as on Tuesdays (AM), 07 55 hrs. 30 bn Figure 10 Typical Traffic Movement Rates as on Tuesdays (PM) , 17 25 hrs. 31 Figure 11 Typical Traffic Movement Rates as on Sundays (AM) , 07 55 hrs. 32 Figure 12 Typical Traffic Movement Rates as on Sundays (PM) , 17 25 hrs. 33 3.1.3 Site 2 – Lautoka Lautoka is a rapidly developing city in Fiji. It is the second out of the two cities and like Suva, has quite high population density. It serves as the main trade and commercial activity point for residents of Lautoka, Nadi and Ba. Unlike Suva City, Lautoka City is characterized by more local industries, such as sugar mills, food processing factories and wood chip mills, which may influence traffic emissions. The site was located in the city center, adjacent to the Lautoka Bus Station and market (latitude: -17.6038˚S, longitude: 177.4538˚E). Samplers were positioned near the exit way of buses that enter the main bus station in Lautoka. Figures 13 and 14 show the street and satellite view of the site respectively. The site is located on Vakabale Street, which connects to the main Vitogo Parade Road and Tukani Street. Various institutions such as the Natabua High School, Drasa Avenue School, Lautoka Central Primary School, Post office, botanical garden and play parks, banks, shopping malls, University of the South Pacific Lautoka Campus, supermarkets, restaurants, cinema houses, churches and temples as well as hotels are located within 1 km radius of the site. The Lautoka hospital and sugar mill are located 1.2 km and 1.4 km away respectively. 3.1.4 Road Networking and Traffic around Lautoka Site The road networks in Lautoka City are not as complex as Suva. Kings Road and Vitogo Parade are major service roads, linked with many 2 lane (one way) arterial roads leading to majority of the city points. Moreover, unlike Suva City, Lautoka City is yet to develop a long term transport strategy which would detail traffic intensity and flow. Land transport Authority’s traffic sensors located 3 km away from the sampling site show an average of 11 381 vehicles per day of which around 84.5% vehicles are passenger vehicles. While the average vehicle number is less than half of that in Suva, large vehicles make 15.5% of average vehicles in Lautoka in comparison to 11 % in Suva. Moreover, traffic sensors in Lautoka do not include vehicles arriving in Lautoka City from Suva and Nadi via Queens’s road. Traffic sensor located 12 km away from the site at Queens’s road shows a greater number of vehicles with an average of 15672, approximately 10% of which are large vehicles. Data are not available on manual 34 vehicle counts, however, traffic map layering using Google Maps provides fairly accurate description of traffic movement rates in surrounding areas of the sampling site. This is demonstrated by Figures 15 and 16. In addition, traffic is almost non- existent on Sundays, hence traffic movement rate maps are not provided. 35 Sampling Site 2 Figure 13 Street view of sampling Site 2 36 Sampling Site 2 Figure 14 Satellite view of sampling Site 2 37 Figure 15 Typical Traffic movement rates surrounding sampling site on Tuesdays at 08 30 hrs (AM peak hour). Figure 16 Typical traffic movement rates on Tuesdays, 1805 hrs (PM Peak hour) 38 3.2 PM2.5 Concentrations 3.2.1 Gravimetric Analysis The US EPA outlines the use of gravimetric analysis as the best method for PM2.5 analysis, and this method is used by many air monitoring networks such as Speciation State and Local Air Monitoring Stations (SLAMS) and Speciation Trends Network (STN) (US Environmental Protection Agency, 2017a). Sampling and gravimetric analysis was done with reference to US 40 CFR 50, Appendix L to Part 50 - Reference Method for Determination of Fine Particulate Matter as PM2.5 in the Atmosphere (US Environmental Protection Agency, 2017b). This method was further supplemented by Quality Assurance Guidance – Document 2.12, to ensure high quality data can be obtained (US Environmental Protection Agency, 2016). Air sampling was done using the Ecotech HiVol 3000 sampler. 3.1.2.1 Instrumentation Detail for Gravimetric Analysis 3.2.1.1.1 Ecotech 3000 High Volume Sampler The Ecotech Hivol 3000 is a CE (Conformité Européene) compliant high volume sampler, which conforms to the European Health, Safety and Environmental legislations. It also conforms to IEC 61010-01, which outlines the safety requirements for electrical equipment’s for measurement, control and laboratory use. The HiVol 3000 has also been approved for US EPA manual reference method: RFPS-0706-162 (Ecotech, 2017). The HiVol 3000 has true volumetric flow control combined with user programmable data logging capacity. The HiVol 3000 was fitted with a PM2.5 size selective inlet (SSI) to enable sampling for PM2.5. The PM2.5 SSI isolates all particles having an Equivalent Aerodynamic Diameter (EAD) of 2.5 microns or less. All sampling was done at a flow rate of 67.8 m3/hr (1.13 m3/ min) for 24- hour periods. 3.2.1.1.2 Size Selective Inlet (SSI) Operational Mechanism The SSI is of impactor well type, which draws air at a rate of 67.8 m3/ hr. As air enters the SSI, particles are accelerated through multiple accelerator nozzles, gaining momentum. Particles larger than 2.5 μm, then, gain enough momentum and collide 39 (impaction) on a greased collection shim. It is in this process that particles greater than 2.5 microns are trapped onto the greased shim and only particles with diameter of 2.5 μm or less pass through (Ecotech, 2013). Figure 17 provides a cross-sectional view of a SSI, including the movement of air through the acceleration plates and greased collection shim. Figure 17 Cross- sectional view of the Size Selective Inlet (SSI) Source: Ecotech (2013) 3.2.1.2 Sampling PM2.5 sampling was done at the aforementioned sites in Suva and Lautoka City. Sampling was done according to the US EPA PM2.5 1 in 3 day sampling schedule, using Ecotech 3000 HiVol sampler at a flow rate of 67.8 m3/hr (Ecotech, 2013; Vojtisek-Lom et al., 2015; Zhu et al., 2018). At the end of each sampling, the logged data on the HiVol was transferred to laptops using the Airodis software and stored. Manual notes of total volume, total corrected volume, average flow rate and total run time was also noted through the status display. Any flags such as grass cutting, sudden 40 or extreme metrological or physical change around the sites were also noted. The hivol sampler was calibrated using an orifice and a digital manometer through pressure drop every four weeks to maintain optimum operation and in accordance with the guidelines (US Environmental Protection Agency, 2016). A 24 hour sample was collected beginning 00 00 hrs (midnight) to the next midnight. A total of 42 samples were collected in Suva City from 2nd September to 3rd January 2019 and 23 samples collected in Lautoka City from 13th January to 8th March 2019. Pre-conditioned and pre-weighed Polytetrafluoroethylene (PTFE) coated filters were used during each sampling. These filters were post-conditioned and post weighed to attain the mass difference and to determine the PM2.5 concentrations. 3.2.1.3 Filter Conditioning and Weighing For gravimetric analysis, 8 inch x 10 inch borosilicate filters bonded with Polytetrafluoroethylene (PTFE) (Pallflex EMFAB TX40H120-WW, Pall Cooperation, USA) were used to collect the samples. These filters are constructed using pure borosilicate glass microfibers and reinforced with woven glass cloth before being bonded with PTFE. Every filter paper is flushed with distilled water to remove any water soluble residue. The borosilicate and PTFE enable the filters to have low air resistance and enhance purity, non-hygroscopic and non-reactive properties (Pall Corporation, 2018). The Emfab filters have been known to show extremely low variability in mass, demonstrating remarkable stability in comparison to quartz and non-bonded glass filters (Brown et al., 2006). These filters are also able to withstand folding during weighing and transportation, making it the best choice for ambient air monitoring studies. PTFE membranes are also preferred for elemental and ion analysis due to lower elemental concentration in blanks and are recommended by agencies such as US EPA and Canadian Air and Precipitation Monitoring Network (CAPMoN) (Perrino et al., 2013) . The filters were conditioned in a controlled environment before and after sampling. The conditioning was done in batches for 72 hours as recommended by US Environmental Protection Agency (2016), at 21.1± 1.5 °C and 34 ± 3 %RH, in a small and isolated clean room, located inside an air conditioned laboratory. Conditioning 41 involved positioning the filters on the filter racks, exposed to the temperature and humidity. The clean room was always locked (except when in use) to limit entry to anyone except the researchers. All tabletops in the clean room were wiped daily, and the clean room was cleaned on a weekly basis using alcohol moistened cloth. The room contained the condition racks, filter papers, weighing microbalance, as well as a separate air conditioner and two dehumidifiers to maintain the prescribed environmental conditions. Once conditioned, the filter papers were weighed using A&D BM-20 (A&D Australasia Pty Ltd, Australia) ionizing microbalance. The BM20 model electronic microbalance manufactured by A&D has been designed for precision weighing with a readability to 1 μg. The scale is equipped with a built in DC 4 point ionizing system to remove the effect of static electricity while weighing. The ionizing system is also fanless, thereby, minimizing any counter effect on the filters, especially during post weighing. This also prevents embedded particles on the filter being blown away. The microbalance also has a motor driven internal calibration system with one touch calibration and automatic response adjustment, giving it a better accuracy (A&D Australasia Pty Ltd, 2017). The microbalance was wiped with a moist cloth to remove any particles. It was positioned on a flat surface and adjusted using the bubble level so that it was sufficiently level. To maintain accurate readings, the BM-20 was calibrated using internal mass before every weighing session. The balance was “zeroed” or tared before weighing each filter paper to eliminate and prevent concurrent errors. The ionizer was switched on for 10 seconds before weighing to eliminate static electricity. The filters were ionized and the mass was noted to a readability of 1 μg after the readings had stabilized. All filters were weighed in duplicate. Following the weighing, the filters were packed in zip lock bags and stored for sampling. A similar approach was used for post sampling in which the filters were conditioned again and weighed. Once weighed, the filters were stored in marked zip locks for further analysis. 42 3.2.1.4 Calculation of PM2.5 Mass Concentration PM2.5 mass concentrations were calculated based on the formula documented by US Environmental Protection Agency (2016). The mass of PM2.5 collected on each filter was determined by Equation 1, which was then used to calculate the concentration of PM2.5 in Equation 2. (൧ݔͳͲͲͲ (1ܯ െܯெమǤఱ ൌൣݏݏܽܯ MassPM2.5 = Total mass of PM2.5 obtained in the sampling period (μg) Mf = Mass of the filter paper after the sampling period (mg) Mi = Mass of the filter paper before the sampling period (mg) ெ௦௦ುಾమǤఱ (ݐ݅݊ெమǤఱ ൌቂ ቃ (2ܽݎ݊ܿ݁݊ݐܥ ೌ 3 ConcentrationPM2.5 = Concentration of PM2.5 (μg/m ) MassPM2.5 = Total mass of PM2.5 obtained in the sampling period (μg) 3 Va = Total volume of air pulled through the filter (m ) 3.2.2 Real Time PM2.5 Concentrations Real time analysis was also done to examine the spatiotemporal distribution of PM2.5 in Suva and Lautoka City. A scattering nephelometer DataRAM (pDR-1500, Thermo Scientific, Franklin, USA) was used to investigate the diurnal patterns of PM2.5 concentrations. The pDR 1500 is an integrated type nephelometer, which measures scattering of light by particles. It operates on a wavelength of 880 nm and the output is based on calibration against Arizona Test Dust (ISO 12103-1, Powder Technology Inc, USA) (Malm & Hand, 2013; Shakya et al., 2017; Zhang et al., 2018b). 43 A 2.5 μm cut point cyclone was fitted at the pDR-1500 inlet for selective measurement of particulates of size 2.5 μm or less. The pDR was operated at a flow rate of 1.52 L min-1 with logging intervals of 5 min for 24 hours during each sampling session. The cyclone inlet was washed with distilled deionized water (DDW) every two weeks to prevent any blockage (Shakya et al., 2017; Vilcassim et al., 2014). The nephelometer was autozeroed using clean air via a HEPA filter before each sampling session to minimize errors. Clean glass microfiber filters (Whatman 934- AH, GE Healthcare, UK) were installed before autozero was done prior to each sampling. Each sampling session lasted 24 hours using both the real time analyzers. In addition, sampling sessions lasted a week in each of the two cities. 3.2 Real Time Black Carbon Concentrations 3.2.1 MicroAeth AE51 A micro-aethalometer (AE51, Aethlabs, San Francisco, USA) was used to investigate the real time concentrations of black carbon. The sampling was done alongside real time PM2.5 sampling using the pDR-1500. The underlying principle of microAeth is that the output measurements is based on quantifying the transmission of light at 880 nm (Williams & Knibbs, 2016). This is done through monitoring changes between the active areas and non-active areas of the filter. The sample (particles) are deposited in a 3 mm circular portion, referred to as the active area of the filter and monitored by a sensing channel detector. Changes to this is compared to an unused portion of the same filter (having no active flow) which is monitored by the reference channel detector (Cai et al., 2014). The assumption of such a technique of optical measuring is that the increase in optical attenuation is proportional to an increase of mass loading of BC on the active area. This forms the basis of operation of all aethalometers (Cai et al., 2013). High reproducibility and good agreement with full size rack mounted aethalometers have been observed for the microAeth in validation studies (Cai et al., 2013; Cai et al., 2014). Cai et al. (2014) reported strong correlation with R=0.98 and a slope of 1.04 ± 0.07 for 24 averages of BC between the microAeth and full size rack mounted 44 aethalometers. The microAeth also performed well at high time resolutions, giving correlations of R= 0.92 for 1 minute interval data. The microAeth was operated at a flow rate of 50 ml min-1 and logged data at a user defined interval of 5 minutes (Shakya et al., 2017) . Moreover, T60 PTFE coated borosilicate glass fiber filters (AE51 Filter Strips, Aethlabs, San Francisco, CA) were used for the micro-aethalometer and sampling was done alongside real time PM2.5 monitoring in Suva and Lautoka City. 3.3 Heavy Metal Characterization Chemical characterization was done by digesting particulate loaded filters using microwave assisted digestion and analyzing the filtrate using Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES). Microwave digestion of the filters was done at the Institute of Exact and Applied Sciences at University of New Caledonia, Noumea, New Caledonia. Trace metal and elemental analysis was performed at the Research Institute for Development (IRD) based in New Caledonia. Microwave assisted digestion technique was used to digest particulate loaded Teflon filters. Specifically, 2 inch by 8 inch filter strips were cut using plastic scissors and placed inside a 100 ml Polytetrafluoroethylene tetrafluoromethoxylene (PTFE –TFM) digestion vessel. To this, 10 ml of HNO3 (70%, Merck, Germany) and 2 ml of HCl (37%, Merck, Germany) was added (Taiwo, 2016). The vessels were then placed in a ceramic jacket and placed in the microwave housing. A set of four digestions were done each time using a microwave digestion system (Multiwave 3000, Anton Paar, Austria). The microwave functioned as follows: it was ramped up to 1000 watts for 10 min and then held steady for 30 min before being brought to 0 watts for 15 minutes to allow for cooling. The vessels were then allowed to cool to reduce residual pressure inside the vessels. The digested solution was then filtered using a disposable filter having a pore size of 0.20 μm (Minisart, Sartorius, Germany) and 10 ml was transferred using a Terumo syringe (Terumo, Philippines) into a 50 ml volumetric flask. The filtrate was made to the mark using ultra-pure water (UPW) of resistivity 18.2 ΩM/cm at 25 ˚C. The solution was then stored in a cooler for trace metal analysis. 45 Heavy metal analysis was done at the Analytical Equipment Laboratory (LAMA) under Instrumentation, Analytical Resource, Geophysical and Oceanography Observatories (IMAGO) division at the Institute of Research and Development (IRD). The IMAGO labs are ISO 9001:2015 certified since 2017. An ICP OES (Varian 730- ES, California, USA) was used to analyze the samples. More details about the ICP – OES Varian 730-ES are highlighted by Agilent Technologies (2010) and Calderon (2010). Samples were analyzed in duplicates. Four different types of blanks were included in the analysis to maintain quality. The first blank was solely distilled deionized water. The second blank, known as the method blank comprised of reagents that underwent same digestion process as the filters. The third blank was labelled as laboratory blank and made of digesting unused and unexposed filter paper. The fourth blank was labeled as field blank and made by digesting field blank filters. Field blanks were analyzed to examine possible contamination of PM2.5 filters during transportation and the handling processes, reducing the gravimetric bias (Figueroa et al., 2006; Kulshrestha et al., 2009). 3.3.1 Calculation of Metal Concentration Mean concentrations of PM2.5 was determined by Equation 4 which was derived from Equation 3 in accordance with US Environmental Protection Agency (2016). ሾሺρ௧Ȁሻሺ௫௧௧௩௨ሻሺଽሻିிሿ ܥൌ (3) C = Concentration of metal (μg / m3) μg metal / ml = Concentration of metal determined by instrumental analysis 9 = (Usable filter area (8” x 9”)) / (Exposed area of 1 strip (1” x 8”)) Fm = Average Concentration of blank filters (μg) V = Volume of air pulled through the filter (m3) 46 ሾሺ௧Ȁିிሻ௫௫௧௧௩௨ሺǤହሻ௫ସǤହሻሿ (ൌቂ ቃ ݔͳͲͲͲͲͲͲ (4 ܥ Ca = Concentration of metal (ng / m3) mg metal / l = Concentration of metal determined by instrumental analysis (ppm) 4.5 = (Usable filter area (8” x 9”)) / (Exposed area of 2 strip (2” x 8”)) Fa = Average Concentration of blank filters (ppm) V = Volume of air pulled through the filter (m3) 3.4 Statistical Analysis All statistical analyses were performed using the SPSS (IBM, USA) software and Microsoft Excel. Descriptive statistics were used to examine the mean concentrations, the deviations as well as the obtained minimum and maximum values from the gravimetric sampling. To understand the effects or relations, if any, that the meteorological parameters have on the concentrations of PM2.5, Spearman Rho correlations (one-tail) analysis was done. Hourly means were computed using the obtained real time data and bar charts were made to see the distribution and diurnal patterns of PM2.5 and black carbon. Boxplots of real time data were also made to visualize the distribution and the measures of central tendency of the real time data. Non- parametric correlations (Spearman Rho) were also used to investigate the relations between metals and enrichment factors computed to determine the source of origin. 47 4.0 Results and Discussion Overview This section presents the findings of the research. The concentration of fine particulate matter is presented here with associated discussion on meteorological parameters as pathways of dispersion and its conformity to WHO guidelines. The real time data is then provided and the diurnal variation of the PM2.5 concentrations is demonstrated. A comparison between the data obtained from the high volume sampler and the real time analyzer (pDR-1500) is also done. Finally, the heavy metal concentrations are discussed as a composition of PM2.5. 4.1 PM2.5 Concentration The concentrations of fine particulate matter are presented as arithmetic mean of the concentrations during the designated sampling periods. A majority of the regulations, standards and comparison of environmental pollutants, including air pollutants are based on arithmetic mean concentrations (Chaloulakou et al., 2003). Table 2 provides the statistical summary including the mean and median concentrations of PM2.5 for the months sampled. Table 2 Descriptive statistics of monthly PM2.5 concentrations in Suva and Lautoka City Month Mean ± SD Median Max Min (μg/m3) (μg/m3) (μg/m3) (μg/m3) Suva September 26.1 ± 17.2 23.8 71.3 9.6 (2018) October 22.9 ± 17.9 21.2 69.6 8.5 November 19.0 ± 7.56 15.7 35.2 12.2 December 18.2 ± 7.3 16.1 33.0 10.0 Lautoka January 48.7 ± 24.6 43.6 95.2 24.5 (2019) February 74.2 ± 38.8 59.0 177.6 30.1 March 72.1 ± 33.8 57.4 122.6 51.2 3 The daily mean PM2.5 concentration for Suva City was 21.6 ± 13.3 μg/m (median, 17.9 μg/m3). The daily mean weekday concentration was higher (23.8 ±14.8 μg/m3, 48 median, 21.7 μg/m3) than weekend concentrations (15.2 ± 5.4 μg/m3, median, 13.9 μg/m3). However, the daily mean fine particulate matter concentration in Lautoka City was greater than the mean concentrations in Suva City, at 67.2 ± 35.2 μg/m3 (median, 3 58.1 μg/m ). The average weekday and weekend PM2.5 concentration in Lautoka City were determined to be 71.4 ± 36.5 μg/m3 (median, 58.9 μg/m3) and 57.3 ± 32.7 μg/m3 (median, 47.1 μg/m3) respectively. Lower concentrations were typically observed for the months of November, December and January. At present, there are no national standards for fine particulate matter under the Fijian Law. Hence, the guidelines put forward by the World Health Organization can be used as the bench mark for comparison. The WHO guideline for mean daily and annual 3 3 PM2.5 concentrations are 25 μg/m and 10 μg/m , respectively (World Health Organisation, 2006). In comparison, 95% of the daily means obtained in Lautoka exceeded the WHO daily guideline with 65% more than twice the guideline. Suva had 27% of the daily means exceeding the WHO guideline. Only 7% of the daily means in Suva were below 10 μg/m3 while the lowest daily concentration recorded in Lautoka was 24.5 μg/m3 when compared to the WHO annual guideline. The mean concentration of PM2.5 over the sampling periods in both the cities exceed the WHO annual mean guidelines. 49 Figure 18 PM2.5 concentrations categorized by day at the sampling sites Mean PM2.5 concentrations in Suva City were below the WHO guideline. On average, all days of the week except Monday have PM2.5 concentrations lower than the WHO 3 guideline of 25 μg/m (Figure 18). However, PM2.5 concentrations determined in this study is approximately 3 times higher than that reported by Isley et al. (2017), who found it to be 7.4 μg/m3. As highlighted earlier, the sampling sites and altitude were different between the current study and that done by Isley et al. (2017). Moreover, while Isley et al. (2017) study highlighted the ambient and background PM2.5 concentrations, this study demonstrates the concentrations in heavy traffic areas. While the concentration for Suva is below the WHO guideline, cities with similar population such as Westminster and Santa Monica in California, USA, have fine particulate concentrations below 15 μg/m3 near roadways (Air Visual, 2019; Polidori & Fine, 2012). In addition, cities likes West Jerusalem and Haifa, which are also in developing countries like Israel have PM2.5 concentrations very similar to Suva. However, West Jerusalem and Haifa have more than twice the population size of Suva City and climate conditions that favor particulate formation (Sarnat et al., 2010; United Nations, 2018). This raises concern about the sustainable development of Suva City 50 and the high probability of fine particulates to exceed the WHO guidelines as the population of Suva grows in the near future. The average PM2.5 concentration in Lautoka City was more than 3 times higher than that in Suva City and little less than 3 times more than the WHO guideline. High variation between the average concentrations for each day of the week were also observed between Suva and Lautoka, as shown in Figure 18. Of all the weekdays, mean PM2.5 concentrations by day were the highest on Thursday (92.00 ± 74.71 3 μg/m ). The highest daily mean PM2.5 concentration recorded in Lautoka was 177.15 μg/m3. Currently, there are no literature studies on fine particulate matter concentrations in Lautoka City at the time of this study, thus, comparison cannot be made. This is the first study which has investigated PM2.5 and air pollution in general in the Western side of Fiji. Cities such as Hastings in New Zealand have lower concentrations of PM2.5 than Lautoka, despite having a similar population size. PM2.5 concentrations in Hastings peaks in winter due to domestic heating, however, mean concentrations still remain below 30 μg/m3 while the summer season mean concentrations are generally below 10 μg/m3 (Wilton et al., 2007). The city of Wagga Wagga (Australia), which experiences similar climate conditions as Lautoka, has a 24 hour mean PM2.5 concentration of around 13.4 μg/m3 (New South Wales Government, 2019), which again is lower than that observed in Lautoka City. PM2.5 concentrations in Suva and Lautoka were compared to roadside concentrations in other cities globally as presented in Figure 19. In Comparison to London, a city of approximately 2.6 million registered vehicles (Transport for London, 2013), summer 3 roadside PM2.5 concentrations (19.40 μg/m ) (Kassomenos et al., 2014) were lower than concentrations in both, Suva and Lautoka. This has important implication as London has approximately 23 times more vehicles than the total number of vehicles in Fiji, yet has lower roadside fine particulate matter concentration. Similarly, other large cities such as Athens and Madrid which have much higher 3 vehicle numbers show lower roadside PM2.5 concentrations at 38.48 μg/m and 20.63 μg/m3 respectively (Kassomenos et al., 2014) as highlighted in Figure 19. Megacities like Tokyo, Osaka, Miami, Singapore, New York and Atlanta also have annual mean concentrations below 15 μg/m3 (Cheng et al., 2016). 51 Figure 19 Roadside PM2.5 concentrations in some global cities References for the illustrated cites are as follows; Kanpur, (Sharma & Maloo, 2005); Kathmandu, Shakya et al. (2017); Chennai, Srimuruganandam and Shiva Nagendra (2012) ; Nairobi, Kinney et al. (2011); Cordoba, López et al. (2011) ; Hong Kong, Lee et al. (2006); Istanbul, Onat et al. (2013); Athens, Kassomenos et al. (2014); Singapore, Zhang et al. (2017); Cincinnati, Martuzevicius et al. (2004); Bern, Gehrig and Buchmann (2003); Madrid, Kassomenos et al. (2014); New York, Patel et al. (2010); London, Kassomenos et al. (2014); Sydney, Wadlow et al. (2019); Auckland, (Davy et al., 2017); Comparing our study elsewhere, roadside PM2.5 concentrations in Singapore were reported to be 28.9 μg/m3 (Zhang et al., 2017). This is much lower than the concentrations observed in Lautoka, a city of 5 million fewer people. Similarly, PM2.5 concentrations in Istanbul were 40.5 μg/m3 (Onat et al., 2013), which were lower than the concentrations observed in Lautoka. It is noteworthy that Istanbul, a city of 2.5 million vehicles and greater traffic congestion, showed lower roadside PM2.5 concentrations than Lautoka City. 52 The incidence of PM2.5 concentrations exceeding the WHO daily guideline was high in Lautoka City. The percentage of sampling days exceeding the WHO guideline was 95% in Lautoka in comparison to 27% in Suva. On average, all days of the week except Monday have mean PM2.5 concentrations lower than the WHO guideline in Suva City. The daily mean PM2.5 concentrations by day for all the days except for Sunday was more than twice the WHO guideline of 25 μg/m3 in Lautoka City. The PM2.5 concentrations observed for Suva and Lautoka City were used to assess the air quality and its impact on local population using the Australian standards. The Environmental Protection Authority of Victoria (EPA Victoria) provides qualitative parameters for air quality, and categorical parameters for its influence on health, based on 24 hour mean PM2.5 concentrations and are presented in Figure 4 and Figure 5. Based on the rubric (Figure 4 and Figure 5) the air quality was fair during 27% of the days in Suva City and only 4% in Lautoka City. Only 5% of the days had very poor air quality in Suva in comparison to 87% in Lautoka. Lautoka City had poor or very poor air quality for approximately 96% of the sampling days. This meant that the level of PM2.5 concentrations induced unhealthy effects on all age groups on 78% of the days and very unhealthy effects on 9% of the days (Figure 20). It is noteworthy that this study demonstrates the effect on the population spending time in the Lautoka city areas with usual traffic conditions and high pollution levels and should not be taken as the levels of which the whole population of Lautoka is exposed to. 53 Figure 20 Air quality during sampling days in Suva City (A) and Lautoka City (B); Health effects on population as per the sampling days in Suva City (C) and Lautoka City (D) Moreover, while 73% of the sampling days in Suva had moderate effect on the public health, only 4% of days in Lautoka had the same effect. Suva also saw 2% of the days with very low effect on health while Lautoka had none such days. The high frequency of poor and very poor air quality days as well as the subsequent high rates of unhealthy and very unhealthy conditions demonstrates the need for immediate action in terms of air pollution mitigation in Lautoka City. Furthermore, it was observed that the average monthly concentration in Suva City was lower in November and December in comparison to September and October. A rise in mean monthly concentrations were also observed for the months of February and March in comparison to January in Lautoka. Important ground based observations and interviews can be used to attribute these observations to the primary and high school calendar in Fiji. There are three school terms in a year in Fiji. Terms 1 and 2 are made up of 14 weeks each of classroom studies while the 3rd term has a duration of 13 weeks. The first term 54 usually begins towards the end of January and the third school term finishes in mid- November (The Government of Fiji, 2018). Around 236 515 students combined with more than 13 000 teachers countrywide attend schools using various modes of transport, accounting to approximately 30 % of the country’s population. Most of the schools are situated in urban centres and steep increase in the number of active bus trips (due to “school buses”) is seen during the school terms (Ministry of Education, 2019). Most of the students commute by buses and this causes greater traffic on the road. Bus drivers have pointed out that the traffic is much heavier and longer during school term days in comparison to non-school term days (Chand, 2019). As highlighted, fine particulate matter concentrations in Lautoka City were greater than Suva. Apart from vehicle emissions, the geographic position and climate of Lautoka are important factors which influence the mean PM2.5 concentrations. In emphasis, concentrations in Lautoka were greater than twice the WHO guideline and may pose a health risk to the exposed population. This accentuates PM2.5, and air pollution in general as a national issue. It further effaces the idea that fine particulate matter is not an issue in the South Pacific countries. 4.2 Geography and Metrological Correlation The two cities under investigation are distinct in terms of their climate and geography. The geography and climate of the sampling areas influence the formation, concentration and dispersion of pollutants such as fine particulate matter. Factors such as low amount of rainfall, lower humidity, aridity and increased thermal convection increase dust re-suspension into the atmosphere. These factors are also known to contribute to pollutant transport processes (Artı́ñano et al., 2004; Ouyang et al., 2015; Tai et al., 2010; Wang & Ogawa, 2015). To examine the relationship between the metrological variables and PM2.5 concentrations, Spearman’s Rho correlation was computed. The correlation data is shown in Table A1.1 and Table A1.2 for Suva and Lautoka respectively and can be found in Appendix A. To visualize the correlations comprehensively, a correlation plot was made for each of the cities and is used in further discussion. The correlation plot demonstrates the correlation between PM2.5 and the meteorological variables and between the 55 meteorological variables as well. Figure 21 A demonstrates the correlations found in Suva City while Figure 21 B highlights the correlations from Lautoka City. (A) (B) Figure 21 Correlation Plot of PM2.5 and Meteorological variables in Suva (A) and Lautoka City (B) 56 4.2.1 Rain Due to its geographical location, Suva City receives more rain than Lautoka. Suva is located on the South Eastern side of the main island of Viti Levu. The eastern and south eastern sides of Viti Levu are influenced by mountains and high terrains, with some peaks reaching as high as 1300 m above sea level (Pacific Climate Change Science, 2011). Figure 22 is an elevation map of the main island of Viti Levu and demonstrates areas with high altitude. South Eastern Trade winds Figure 22 Altitude Map of Viti Levu Source: topographic-map.com (2019) In general, as air masses rise over the mountains, it cools and condensation takes place. This causes the water vapor to condense into liquid and form clouds (Royal Meterological Society, 2017). Therefore, high terrains and mountainous regions experience more rain due to the increased cloud formation and orographic effects (Carpenter, 2018). Fiji experiences a dominant South Eastern trade winds as highlighted in Figure 22. These factors cause more clouds to build up on the south and eastern parts of the island (which is also the windward side) in comparison to the western and northern areas as shown by Figure 23. Hence, Suva City experiences more rainfall than Lautoka. The geographical location of Lautoka puts it on the leeward side of the island, having effects of rain shadow and therefore, receiving less precipitation (Stockham et al., 2017). 57 Figure 23 Satellite image showing typical difference in cloud build-up between the eastern and western sides of Viti Levu Source: National Aeronautical and Space Administration (2019) Historical data on rainfall also show greater amount of precipitation in Suva than in Lautoka. Suva experiences around 3000 mm of rainfall per year in comparison to around 2000 mm of rainfall in Lautoka (Fiji Meterological Service, 2019; Pacific Climate Change Science, 2014). Moreover, Fiji experiences only two seasons. The first is the warm and wet season which begins in November till April and marks higher average temperatures and greater tropical downpour. The second season is the cool and dry season which lasts between May through October, having slightly lower temperature and rainfall (Pacific Climate Change Science, 2014). Considering the amount of rainfall only on the sampling days, each sampling day in Suva received on average 12.8 mm of rainfall compared to only 5.7 mm in Lautoka. The sampling in Lautoka was done during the warm and wet season and hence the amount of rainfall is expected to be lower during the cool and dry season, possibly increasing PM2.5 concentrations. Rain has a negative effect on fine particulate matter concentrations due to wet deposition. Moderate negative correlation (rs = -0.354, p < 0.05) between the fine particulate matter concentration and total rain was found in Suva City and can be observed in Figure 21 (A). However, no significant correlation was found between rain and PM2.5 concentration in Lautoka City as demonstrated in Figure 21 (B). An examination of Table 2 also shows that the average monthly fine particulate matter 58 concentrations are actually higher for the months of September and October, which are in the cool and dry season in comparison to the months of November and December, which are in the warm and wet season. Lautoka City, which receives only two thirds the amount of rain experienced by Suva, shows higher concentration of ground level PM2.5. This demonstrates that rainfall is an important factor in lowering ambient PM2.5 concentrations. 4.2.2 Wind Wind affects the pollutant concentration, dispersion and transportation. Statistically significant negative correlation was found between fine particulate matter and wind speed in Suva City, with a correlation coefficient of -0.671, p < 0.01. Wind speed and direction over the sampling period is shown in Figure 24 A. The dominant wind direction at the Samabula site was SSW with average wind speeds of 4 to 8 km/h. However, the concentration of PM2.5 was highest when the wind had a NNW direction. No significant correlation was found between wind speed and PM2.5 in Lautoka City. Lower wind speeds were recorded at the Lautoka Bus Station with a dominant wind direction of WNW. Wind speeds were generally lower than 3 km/h as presented in Figure 24 B. Figure 24 Wind Rose of Sampling in Suva (A) and Lautoka City (B) 59 The geographical location of Fiji puts it in the path of the South Easterly trade winds, hence, SE is expected to be the dominant wind direction. However, Mataki et al. (2006) reported that while SE winds are prevalent in the cool season, wind direction is quite variable in the hot and wet season. Moreover, winds from the northern quarter are common during the hot and wet season as the South Pacific Convergence zone (SPCZ) lies just north of Fiji (Mataki et al., 2006). Lautoka is situated on the North Western side of Fiji in comparison to Suva, thereby, making it closer to the SPCZ. Therefore, winds from the North Western quadrant were seen to be dominant. The influence of the SPCZ during the sampling period may have led the dominant wind direction to be different from the general SE trade winds. Since Lautoka City is on the leeward side of the main island, it has lower wind speeds and is drier. It can also be argued that the position of Lautoka Bus Station is an important contributing factor to the higher levels of PM2.5. The bus station is surrounded by a cluster of buildings which are taller than the bus station. These building contribute to the “wall effect” (Yim et al., 2009) by blocking the flow of air. With reduced air flow, the air at the Lautoka Bus Station stagnates, accumulating pollutants such as PM2.5 in the process, thereby contributing to higher concentrations. 60 (A) (B) Figure 25 (A) Typical Backward wind trajectories arriving at Suva City and (B) movement across the Suva Peninsula 61 (A) (B) Figure 26 (A) Typical Backward wind trajectories arriving in Lautoka City and (B) movement across the Lautoka 62 Backward wind trajectories were made using the Hybrid Single Particle Langarigan Interagrated Trajectory (HYSPLIT) model to see the “the” has been inserted typcial travel paths of winds arriving in the two cities. The HYSPLIT model was developed by NOAA’s Air Resources Laboratory and is based on a hybrid between the Langragigan and Eulerian calculation methods (Stein et al., 2015). 24 hour backward trajectories were obtained at three different altitudes. This was done to observe if there was significant mixing taking place when considering the vertical wind profile and to see how the wind parcels moved at higher altitude with respect to the ground level. Wind trajectroies were calculated at 100 m, 500 m and 1000 m above ground level (AGL) as done by Talbi et al. (2018), arriving at midnight at the end of the sampling date. Typical wind parcels arriving in Suva City had travelled 27.2 km in the last hour before arriving at the sampling site and a total of 847.3 km over the 24 hour period. Wind parcels in Suva travelled largely over the ocean before arriving on land (Figure 25). It travelled only 2.9 km on land with a total time of approximately 7 minutes before arriving at the sampling site as shown in Figure 25. In contrast, wind parcels arriving at the Lautoka sampling time typically travelled 696.57 km over a 24 hour period and 16.32 km in the last hour. Wind parcels arriving in Lautoka were noted to be slower than those arriving in Suva. A notable difference is that wind parcels travel much greater distances on land in Lautoka than in Suva. While wind parcels travelled only 2.9 km and spent 7 minutes on land before arrivng at the Suva site, wind parcels arriving at lautoka site travelled for 118.9 km and spent 4.2 hours on land as shown in Figure 26 a and b. Wind parcels arriving in Lautoka City therefore have travelled over the drier side of Viti Levu, passing over mountains and large farming area in between Lautoka and Ba before arrving at the site. These wind parcels possibly carry larger concentrations of crustal elements by geological errosion and dust resuspension in comparison to Suva. Examining the movement of trajectories with respect to altitude, it can be argued that vertical mixing of wind parcels are slower in Lautoka (Figure 25a) than in Suva (Figure 26a). Wind parcels in Suva shift from approximately 900 m AGL to 500 m AGL in comparision to 600 m AGL to 500 AGL in Lautoka. Similarly, wind parcels in Suva moved from approimately 1400 AGL to 1000 m AGL in comparision to moving from approximately 1250 m AGL to 1000 m AGL in Lautoka. Vertical mixing 63 of air parcels are important as it aids in the dilution and transportation of pollutants. Slow vertical mixing and weak exchange of air masses lead to increased pollutant accumulation. The acumualtion of crustal elements and dust resuspension combined with slower vertical mixing may be important contributors to the higher PM2.5 concetrations in Lautoka City. 4.2.3 Relative Humidity Relative Humidity (RH) was monitored alongside other meteorological parameters. Statistically significant, although, a weak negative correlation was found between PM2.5 and relative humidity in Suva. The correlation coefficient between humidity and PM2.5 was -0.285 at p < 0.01, indicating a negative association as highlighted by Figure 21 (A). The negative correlation can be explained by examining the average relative humidity over the sampling period. In Suva City, the mean relative humidity was at 83.59%. As discussed earlier, conditions of relative humidity over 70% tend to favor dry deposition, thereby removing aerosols including fine particulates from being suspended in the atmosphere. This happens as greater hygroscopic growth makes the particulates too heavy to remain suspended in the atmosphere (Lou et al., 2017; Wang & Ogawa, 2015). No significant correlation was found between RH and PM2.5 concentration in Lautoka City. However, significant correlation was found between RH and rain as well as temperature. RH had a positive moderate correlation with rain and a negative moderate correlation with temperature with coefficient correlations of rs = 0.561, p <0 .01 and rs = -0.484, p < 0.01 respectively. RH and rain have previously been positively related. 4.2.4 Temperature Ambient temperature has several implications for atmospheric pollutants. It affects the deposition and the kinetics of formation including the conversion of much of the compounds present in the atmosphere. However, this study only focuses on the land surface temperature and does not investigate vertical temperature profiles and associations with fine particulate matter. Moreover, ambient temperature and its influence on particulate pollution has been noted to affect the cardiovascular function, morbidity and mortality as well as non-accidental mortality. This has made ambient temperature an important parameter to consider when examining the effects of fine 64 particulate on human health, especially in epidemiologic studies (Fang et al., 2017; Mathew et al., 2018; Ren et al., 2011). The mean temperature in Suva over the sampling period was 25.25 ± 1.47 ˚C while that in Lautoka was 2 ˚C higher with a small standard deviation, at 27.68 ± 0.67 ˚C. As shown in Figure 21, statistically significant correlation between temperature and PM2.5 concentrations was found only in Lautoka. Significant correlations between temperature and rain in Suva City and relative humidity in Lautoka were also observed. However, no correlation between temperature and fine particulate matter concentrations was found Suva, 4.3 Real Time Trends in PM2.5 and BC Real time monitoring of PM2.5 and BC was undertaken during the field campaign to understand the diurnal patterns. PM2.5 and BC were monitored at 4 different locations to analyze the temporal and spatial variabilities. The World Health Organization prescribes PM2.5 to be the primary pollutant of interest when monitoring air pollution and its effect on public health. However, it strongly suggests that BC should be monitored in combination with PM2.5 to provide better insights into health repercussions of air pollution (Janssen, 2012). Moreover, exposure to BC has previously been noted to increase much more significantly during commuting in comparison to PM2.5 (Adams et al., 2002; Zuurbier et al., 2010). Elemental Carbon (EC) is sometimes measured as an alternative or a surrogate to BC, however, EC and BC may not be used interchangeably (Briggs & Long, 2016) . In the present study, BC and PM2.5 concentrations were monitored using microAeth and pDR- 1500 respectively. The mean concentrations over 5 minute intervals were recorded for both PM2.5 and BC. Using this data, hourly averages were computed and presented in Figure 27 for Samabula, Suva City Bus Station, Lautoka City and Reservoir Road site respectively. The use of hourly averages lowers the excessive influence of out of scenario source emissions and helps to comprehend ambient concentrations in normal conditions. Hourly averages have widely been used elsewhere in literature (Harrison et al., 1997; Perez & Gramsch, 2016; Ruuskanen et al., 2001; Shakya et al., 2017; Steinle et al., 2015). 65 Figure 27 Diurnal Variation of PM2.5 and BC concentrations in Samabula (A), Suva City Bus Station (B), Reservoir Road (C) and Lautoka City (D) over the sampling period Samabula and Suva City Bus Station display similar diurnal patterns of PM2.5 and BC variability. Morning peaks start from 5 am till 11 am while the evening peak occurs between 3 pm till 9 pm. The concentrations of PM2.5 determined by the pDR-1500 at 3 both of these sites are generally below 15 μg /m , with mean PM2.5 concentration of 7.7 ± 4.6 μg /m3 and 6.7 ± 5.5 μg /m3 for Samabula and Suva City Bus Station respectively. At the Suva bus station, 5 minute averages as high as 295.9 μg /m3 were 3 noted with maximum displays up to 1747.5 μg /m . Black carbon concentrations for both sites generally remained below 5 μg /m3 with exceptions during peak traffic hours. The mean BC concentrations were 3.9 ± 2.9 μg /m3 and 2.6 ± 2.7 μg /m3 for Samabula and Suva Bus Station respectively. The diurnal variation profile is similar to that of the PM2.5 variation as presented in Figure 27 (A) 27 (B) with BC concentrations increasing with PM2.5 from 05 00 to 11 00 hours and from 1500 to 21 00 hours in the evening. 66 It was observed that the PM2.5 concentrations were at the lowest during different times of the day between the two sites. The lowest concentration in Samabula occurs between midnight to very early hours of the morning between 00 00 hrs to 03 00 hours. Contrarily, PM2.5 concentrations are at their lowest levels during midday, between 11 00 to 13 00 hours in Suva City. A previous study done in Suva City by Isley et al. (2017) also reported PM2.5 concentrations to be at the lowest between 11 00 and 14 00 hours. Isley et al. (2017) noted that wind speed had a dominant influence, having higher wind speed during midday and conversely lower PM2.5 concentration. The difference in the diurnal pattern, particularly, that of lowest concentration between Samabula and Suva city bus station indicates that wind speed exerts lesser influence on the Samabula site in comparison to Suva City. It can be argued that the source of emission, which in this case is road traffic, exerts a greater dominance on the concentration of PM2.5 than the meteorological parameters. Hence, concentration of fine particulates in the Samabula area is higher with increasing traffic and much lower when traffic is almost non-existent in the early hours of the morning. The third site was based in the Nauluvatu Community at Reservoir Road. The site was selected to see the influence of the traffic on a residential community adjacent to the 3 road. The average PM2.5 and BC concentrations were 5.2 ± 8.4 μg /m and 2.4 ± 2.3 μg /m3, respectively. Stronger peaks were seen during the evening in comparison to the early morning periods in the community. There is a greater difference between PM2.5 and BC in the early morning hours which suggests emission sources which may be non-combustion driven. This could be due to indoor emission sources such as mosquito coils. Mosquito coils are quite common in the small PICs and used extensively due to the low cost and high availability. Although cheap, mosquito coils produce high PM2.5 emissions thereby increasing chances of morbidity (Salvi et al., 2014). Moreover, higher peaks are seen till late into the evening. While there is an influence of traffic, community based open burning of rubbish must be acknowledged as contributing factors. Despite the main island of Viti Levu having 4 waste dump sites and a landfill (The Government of Fiji, 2011), open burning is a common occurrence. This is also observed in areas where municipality waste collection services are provided. As such, rubbish is usually openly burnt in the afternoon and left to smoulder till the evening, as observed in the Nauluvatu Community. 67 Larger peaks during the afternoon rush hours were also observed for Lautoka City. The mean real time PM2.5 concentration in Lautoka City obtained through the pDR 1500 was 9.8 ± 11.07 μg /m3. This is lower than the concentration determined by gravimetric analysis. The difference between real time data and that obtained by using a high volume sampler is discussed later. Real time concentrations of fine particulate matter during the evening rush hour rose up to 290 μg /m3 for 5 minute averages and up to 1489 μg /m3 for the 10 second displays at the Lautoka Bus Station. Black carbon also followed a similar trend and while concentrations generally remained below 10 μg /m3, values up to 97 μg /m3 were recorded for 5 minute intervals during the evening rush hour. This suggests that the increase in PM2.5 concentration is a result of increased combustion activity, particularly, through motor vehicle emissions. A rise in PM2.5 concentrations were observed in hourly averages between 16 00 to 17 00 hours (Figure 27) in Lautoka. The BC concentrations follow the PM2.5 concentrations with an average of 4.02 ± 4.7 μg /m3. This can be attributed to the increase in the number of buses actively using the Lautoka Bus Station. During the evening rush hours, “extra buses” are put on routes to meet the demand of buses. As observed, many of the buses are old with visible smoke emissions as shown by Figure 28. These buses line up at the bus stop during the evening rush hour. Typically, a bus waits for 20 minutes before it can find space to board passengers which usually lasts for another 20 to 30 minutes. It was noted that the engines remain running during the waiting and boarding period, possibly elevating emissions in a short span of time at the bus station. 68 Figure 28 Visible smoke emissions from buses in Lautoka City The Lautoka Bus Station is surrounded by taller buildings which provide cover and act as windbreakers (Figure 29) and gives rise to the “Wall Effect” as discussed earlier (Yim et al., 2009). This is also evident when considering average wind speeds of only 2 km/h were noted for the sampling site. In addition, it can also be argued that the Lautoka bus station is running beyond its operational capacity. No re-designing has been done to cater for the increase in number of buses in more than 40 years (Singh, 2019). Using data collected through interviews, the average number of buses operated by each bus company and the trips made is provided in Table 3. 69 Table 3 Number of buses operated and trips made by companies at the Lautoka Bus Station (Source: (Mani, 2019) Bus Company Number of Bus Number of trips made operated everyday by each bus Classic Buses 94 19 Lautoka General Transport 25 15 Kader Buses Ltd 20 10 Akbar Buses 10 1 Khan Buses 9 6 Pacific Buses (Nadi-Lautoka) 5 8 Pacific Buses (Express) 12 1 Inter-Cities Buses 6 1 Nadi General Transport 4 6 Fiji Transport 6 4 Sunset 5 1 Sunbeam 8 1 The Lautoka Bus Station accommodates approximately 2544 trips made by all the buses combined every day. The high density of old buses, overcrowding at the bus station, visible dust re-suspension due to the dry and hot climate conditions are possible contributing factors to the high PM2.5 concentrations. Figure 29 Location of Lautoka Bus station and Surround buildings 70 4.4 High Volume Sampler vs Real Time Data Two methods were employed to assess the fine particulate matter concentrations. Gravimetric analysis was used as the primary method to determine fine particulate matter concentrations and investigate correlations with meteorological parameters. Gravimetric analysis is central to both, the European and the US reference methods for monitoring ambient particulate matter (Tasić et al., 2012) and has been widely used. Real time analysers such as the pDR 1500 was used to examine the diurnal variation patterns of PM2.5 concentrations. Moreover, high volume sampler has also been used in parallel to gravimetric methods as it gives sampling advantages over low volume samplers. Adams et al. (2001) reported that the low flow rate of traditional personal particle samplers is not able to attain sufficient sample mass over short sampling time for accurate gravimetric analysis. Mean concentrations given by pDR are lower than that determined through gravimetric analysis in the present study. This can be expected when considering the limitations of optical based instruments which are discussed below and have been highlighted elsewhere in literature (Briggs et al., 2008; Fischer & Koshland, 2006; Tasić et al., 2012). Real time particle counters have been noted to report lower concentrations than gravimetric analysis using low volume samplers (LVS). Tasić et al. (2012) reported that the Osiris particle monitor (using light scattering technology as the pDR 1500) underestimated the concentration of PM2.5 by as much as 63% in comparison to the reference LVS. Moreover, nephelometers show strong distortions in comparison to gravimetric data if the relative humidity, even briefly exceeds 95% and even “robust statistical measures of central tendency” is not able to correct this non-systematic deviation (Fischer & Koshland, 2006). Similarly, studies such as that by Babich et al. (2000) found BC measurements taken by an aethalometer to be approximately 24% less in comparison to EC concentrations determined through thermal analysis of filters. Fischer and Koshland (2006) also argued that correction factors should be adjusted according to different measurement conditions rather than just using a single correction factor. Particle monitors that use light scattering as the operational basis may underestimate concentrations due to incomplete detection of lower size category particulates. For 71 instance, Tasić et al. (2012) states that these instruments accurately relay particulate mass concentrations only if the size distribution and composition is similar to the calibration standard. Hence, it may systematically underestimate the very finer and ultra-fine components (Briggs et al., 2008). This attribute may inadequately define the concentrations of fine particulate matter near certain emission sources such as traffic emissions. This is because motor vehicles are the most dominant source of both fine and ultra-fine particles in urban areas (Zhu et al., 2002). In emphasis, particulates emitted from diesel engines have a mean diameter in the range of 0.06 -0.12 μm while gasoline engines emit even smaller particles with mean diameter size between 0.04 – 0.08 μm (Harris & Maricq, 2001). Fresh emissions from automobiles are generally in the ultrafine range, below 0.1 μm (Pey et al., 2009).Ultra- fine particles may account for as much as 80% of particle count concentrations in urban areas and therefore deserve attention (Zhu et al., 2002). Optical based instruments are only able to measure concentrations up to a certain particle size based on the selected wavelength. For instance, the pDR 1500 measures particles with a minimum size of 0.1 μm and disregards smaller particles (Thermo Fisher Scientific, 2008). The ultrafine particles, although undetected by optical based instruments, accumulate and gain mass with time. On the contrary, the high volume sampler considers all particles smaller than or equal to 2.5 μm, including ultra-fine particles. As such, gravimetric sampling shows higher concentrations of fine particulates in comparison to the pDR 1500 as shown in this study. The difference in mass concentrations further supports the literature that traffic emissions are important sources of ultra-fine particulate matter. 4.5 Elemental Composition of PM2.5 4.5.1 Element Concentrations Elemental composition is typically determined to investigate toxicity and source of emission of particulates. This has important implications for PM2.5 as metals that are generally emitted from combustion related activities occur in the fine and ultra-fine particle ranges and spend greater time in the atmosphere in comparison to larger sized particles (US Environmental Protection Agency, 2007). Moreover, these elements can be regarded as signatures of different combustion sources and have been extensively 72 used for source apportionment studies to determine the contribution of different sources to the overall PM2.5 concentrations. Elements that are commonly used to delineate different sources are provided in Table 4. Table 4 Common signature elements and sources Element Source Study Al, Si, Ca, Fe, Mg, Na, K, Crustal, Geological, Soil (Hieu & Lee, 2010; Onat Si et al., 2013; Shakya et al., 2017) Ba, Cu, Zn, Ca, Mg, Fe, Traffic Exhaust (Hieu & Lee, 2010; Liu et Ni, C, Cr, Pb, V, Br, Ba, Emissions, Brake pad al., 2018; Onat et al., Sb, Ca abrasion, tire wear and 2013; Shakya et al., 2017) tear and mechanical grinding, Lubrication, Oil Combustion Si, Fe, Al, Mn, Ca Construction (Hieu & Lee, 2010; Jeong et al., 2019; Shakya et al., 2017; Wu et al., 2007) Fe, Zn, Cr, Mn, Ti, Al, Ni, Industrial Emissions (Hieu & Lee, 2010; Wu et V al., 2007) Ca, Mg, K, Na Sea Aerosols (Hieu & Lee, 2010; Wu et al., 2007) The common occurrence of some metals from multiple sources, however, makes it difficult to identify the exact source. Liu et al. (2018) highlighted that noticeable discrepancies in source profiles of heavy metals are found in many studies. Hence, different weights are given to commonly occurring elements in many source apportionment studies. In the present study, concentrations of 22 elements were determined. Most of the elements were metals, however, important non-metals such as sulphur and phosphorus were also analysed. The concentration of the metals, in comparison with studies conducted on roadsides throughout the globe is present in Table 5. Concentrations are reported with one standard deviation (1 σ). Notably, concentrations do not have a normal distribution and hence the standard deviations are indicative of the variance. High concentrations of crustal elements, Ba and Zn were found in this study. Elements such as Al, Ba, Ca, K, Na and Zn were observed to have concentrations greater than 1 73 μg/m3 in both the cities. Ca had the highest concentration in the two cities with concentrations 19.5% higher in Suva than in Lautoka. Concentrations of Al in the two cities were similar, however, Lautoka City demonstrated higher concentrations of K and Na by 25% and 60% respectively than in Suva City. Presence of high amounts of Al, Ca, K, Na in the particulates highlight the strong contributions from natural sources including but not limited to geological erosion and dust suspension (Shakya et al., 2017). High concentrations of Ca, Na and K, combined with comparable concentrations of Mg may also demonstrate the influence of marine aerosols (Hieu & Lee, 2010). Marine aerosols have previously been suggested as the major contributor to PM2.5 concentrations in Suva by Isley et al. (2017). This is followed by smoke emissions through the use of fossil fuel and secondary sulphate formations by vehicle emissions (Isley et al., 2018). 74 Table 5 Elemental concentrations (ng/m3) in current study compared to roadside concentrations in other global cities Ele Our Study Suva Our Study Kathmand Leece New Istanbul Singapore Macau Karachi ment Lautoka u (Shakya (Italy) Jersey (Onat et al., (Zhang et al., (Song et (Mansha s et al., (Cesari et (Xia & 2013) 2017) al., 2014) et al., 2017) al., 2016) Gao, 2012) 2011) Al 3700± 2900 3500± 2600 2060 ± 27.3 602.1 ± 286.66 ± 69 ± 84 2240 ± 1770 191.5 83.63 2960 As 2.5 ± 3.6 2.4 ± 1.7 10 ± 10 0.75 ± 0.86 3.50 ± 0.73 Ba 2400 ± 2100 1300 ± 1100 750 ± 630 241.9 ± 18.63 ± 5.95 119.4 Ca 9900 ± 6300 8300 ± 6400 2240 ± 437.5 ± 146.76 ± 120 ± 90 3351 ± 1740 172.7 84.38 1912 Cd 0.2 ± 0.1 30 ± 20 0.32 ± 0.29 0.107 0.15 ± 0.007 5 ± 3.9 Co 0.3 ± 0.2 0.3 ± 0.2 0.146 0.52 ± 0.68 0.14 ± 007 37 ± 19 2.1 ± 1.4 Cr 3.7 ± 2.4 5.7 ± 4.1 30 ± 20 0.80 ± 0.43 0.786 121.7 ± 7.29 ± 3.10 84 ± 40 40.7 Cu 4 ± 6 9.9 ± 5.6 30 ± 20 3.40 ± 3.98 39.01 19.6 ± 5.66 16.46 ± 4.98 13 ± 5 58 ± 33 Fe 420 ± 220 2300 ± 930 2160 ± 14.22 ± 86.6 117.3 ± 257.32 ± 160 ± 3150 ± 1740 6.65 43.8 75.04 100 725 K 1200 ± 1100 1400 ± 1900 1970 ± 157.2 ± 107.4 ± 660.21 ± 150 ± 60 2656 ± 100 83.72a 54.3 195.80 1396 Mg 640 ± 290 660 ± 420 260 ± 230 34.5 ± 15.3b 66.3 ± 36.5 10 ± 0.00b 33 ± 34 715 ± 418 Mn 4.6 ± 4.6 41 ± 19 50 ± 30 1.83 42.1 ± 51.2 24.24 ± 11.03 19 ± 11 132 ± 159 Mo 0.5 ± 0.3 1.08 ± 0.44 14 ± 9.3 Na 2100 ± 2100 3300 ± 3700 520 ± 740 389.2 ± 429.5 ± 220 ± 40c 140 ± 2047 ± 292.1c 1473.8 100 1434 Ni 1.1 ± 0.6 1.2± 0.7 20 ± 10 7.79 8.08 ± 3.26 40 ± 18 75 P 14 ± 10 44 ± 20 140± 80 24 ± 10.2 Pb 5.2 ± 10 4.2 ± 2.3 30 ± 20 9.44 ± 9.64 3.26 29.67 ± 6.92 38 ± 13 368 ± 479 S 390 ± 180 390 ± 140 2250 ± 110.5 ± 1260 ± 1260 49.4 450 Si 80 ± 45 170± 130 5670 ± 5441 ± 190 ± 4480 1901 200 Sr 89 ± 73 51 ± 44 10.0 ± 4.50 1.48 ± 0.43 V 5.4 ± 3.8 8.5 ± 3.4 10 ± 10 2.84 2.54 ± 2 9.5 ± 5.2 Zn 2700 ± 1900 3800 ± 2500 350 ± 460 6.63 ± 8.35 14.7 384.7 ± 119.25 ± 542 ± 245 197.6 30.22 a K+ ; b Mg2+ ; c Na+ 76 Ca had the highest concentration of all the elements analysed in this study in both the cities. Isley et al. (2018) also determined a source characterised by high Ca emissions in Suva. This was taken as a separate factor in the nine-factor Positive Matrix Factorisation (PMF) model used by Isely and given the label of “Industry Ca”. High concentrations of Ca could have arisen from quarry activities taking place at Lami and Nasinu (Isley et al., 2018). While typical scenario based trajectories are provided in chapter 4.2.2, atypical scenarios such as that demonstrated in Figure 30 should be considered. Figure 30 Backward wind trajectories arriving from the SW and NW quadrants 77 Figure 30 highlights scenarios in which wind parcels arrive at the sampling site from the South and North West quadrants. On days with such wind movement, the air parcels spend a greater time over land before arriving at the sampling site. Air parcels also travel over Lami town and the heavily industrialised areas in between Suva City and Lami. In emphasis, factories of the two cement producing companies in Fiji are located in these areas, and they are responsible for almost 100% (80% Pacific Cement Limited and 20% by Tengy Cement) of cement production in Fiji (Lacanivalu, 2017). These factories in combination with construction works throughout the Suva zone may also be significant contributors to the overall Ca concentrations. Similarly, it can be argued that another cement factory of Tengy Cement, located 2.6 km away from the sampling site in Lautoka may be an important contributor of Ca in Lautoka City. Apart from Ca, higher concentrations of Al, As, Ba and Sr were observed in Suva than in Lautoka City as highlighted in Figure 31. Concentration of all other elements, specifically, Co, Cr, Cu, Fe, K, Mg, Mn, Ni, Na, P, S, Si, V and Zn were higher in Lautoka City than in Suva City. Zn V Sr Si S Pb P Ni Na Mn Mg K Fe Cu Cr Co Ca Ba As Al 00.51 Lautoka Suva Figure 31 Element to element ratios highlights elements that have greater concentration in Lautoka City in comparison to Suva City and vice-versa 78 Elements such as Fe, Ni, Zn, S, Cu, and Sr, Cr, V and Mg have been used as traffic markers. Ni and V are typically produced by oil combustion in automobiles. Zn, Cu, Cr, Fe, Mn and Ba have been associated with fuel additives, brake abrasion and mechanical wear and tear (Liu et al., 2018). Zn has also been related to exhaust emissions and tire wear (Shakya et al., 2017). Some studies suggest that Ca, Mg and Fe is also emitted by diesel vehicles (Sharma et al., 2005; Squizzato et al., 2018). Metals such as Fe and Mg have been known to contribute to heterogeneous secondary aerosol formation on the surface of particles, thereby decreasing the overall air quality (Fu et al., 2016). Traffic related elements, with an exception of Ba and Sr were higher in Lautoka. Lautoka City also reported greater concentrations of S, suggesting traffic related emissions to be an important source of PM2.5. Concentrations of Fe and Zn were higher by 458% and 43%, respectively in Lautoka. This can be related to greater number of aged vehicles which suffer higher rates of mechanical wear and tear including metallic abrasion. High concentrations of Fe, Ca, Si, Na, K, Cu and Zn, especially in Lautoka City also suggest contributions from crustal sources including road dust suspension. Cu has largely been used to characterise road dust in many studies (Liu et al., 2018). Concentration of Si in Lautoka City was greater by 112% in comparison to Suva City. While Si is largely recognised as crustal, high concentrations of Si have previously been reported in gold mining dust (Annegarn et al., 1987). Considering this, the Vatukola Gold Mine may be a contributing source to Si concentrations found in Lautoka as it lies on the edge of typical moving air parcels (trajectories provided in Chapter 4.2.2) towards Lautoka. Higher concentration of K, Zn, Cr and Co may also indicate greater industrial related emissions in Lautoka than Suva. A heavily industrialised zone is situated at around 78.2˚ North at a distance of 500 m away from the sampling site in Lautoka City. Numerous metallurgy workshops including engineering firms and bus company garages are situated in this zone. Metallurgy workshops and related activities such as welding and milling are considerable sources of Fe, Zn, Cr and Mn (Hieu & Lee, 2010), all of which are higher in concentration at Lautoka in comparison to Suva. As presented in Chapter 4.2.2, wind parcels travel close to Ba town before arriving in Lautoka City. Potassium (K) is also a good inorganic tracer for biomass burning 79 (Sulong et al., 2017). Agricultural and refuse related burning, especially at peri-urban locations have been a long term issue in Lautoka and may also be contributing sources to concentration of K (The Government of Fiji, 2011). The concentrations of metals determined in Suva and Lautoka are similar to roadside concentrations determined in urban centres in other countries, with exception of Ca, Na, Zn, Al and Ba (Table 5). Concentrations of S were lower than those reported in Kathmandu and Macau by approximately 450% and 220% respectively, however, were greater than 3 times the concentrations determined in Istanbul. The concentration of As were similar to that found in Istanbul and Singapore. Cd was only detected in Suva City and concentrations were similar to Lecce, New Jersey and Singapore and remained much lower than those found in Kathmandu and Karachi. Similar concentrations of Co and Cr but lower concentration of Cu were noted when the two cities were compared with other global cities as per Table 5. The concentration of Fe in Suva was similar to that found in Istanbul but much lower than that found in Lautoka and Kathmandu. High concentrations of Na, Zn and Mg were found in Suva and Lautoka in comparison to other cities. Mn concentrations reported in Suva were only similar to New Jersey, albeit much lower than Lautoka, Kathmandu, Singapore, Macau and Karachi. Mo concentration was also reported to be lower concentration in Suva while it was undetected in Lautoka. P and Si concentrations were similar in both the cities and were comparable to other cities in shown in Table 5, however were lower than concentrations in Kathmandu. 4.5.2 Element Correlations Correlations were determined to investigate the existence of relationships between different elements. A non-parametric 2 tailed Spearman’s Rho correlation was done and the results for the two cities are presented in Table 6 and Table 7, respectively. The correlation between metals demonstrates common emission sources (Thomaidis et al., 2003; Zheng et al., 2016). In Suva City, significant correlation between crustal elements were noted. Correlation existed between Al and Ca, Fe and Si at 99% level of confidence. Similarly, strong correlation was also noted between Na and K. A greater number of significant correlation between traffic related elements were observed. Ba was positively 80 correlated with Ca, Cr, Mg at 95% level of confidence and with Cd, Mo, Sr at 99% level of confidence suggesting similar emission sources. Distinctive elements such as V was also observed to have significant correlation with other traffic related markers, particularly, Ca, Cr, Fe, Mg, Mo, Ni, and S reaffirming direct traffic emissions as the major sources of PM2.5 in Suva City. Typical industrial markers such as Zn was also noted to have significant correlations with Cu. 81 Table 6 Correlation Coefficient between elements in Suva (* p = 0.05) (** p = 0.01) Al As Ba Ca Cd Co Cr Cu Fe K Mg Mn Mo Na Ni P Pb S Si Sr V Zn Al 1.00 As .396* 1.00 Ba 0.42 0.30 1.00 Ca .927** 0.24 .415* 1.00 Cd 1.000** 0.70 1.000** 0.80 1.00 Co 0.10 -0.57 -0.60 0.66 0.50 1.00 Cr .873** 0.11 .504* .809** -.900* 0.09 1.00 Cu -0.09 -0.12 -0.16 -0.14 -0.21 0.60 -0.03 1.00 Fe .346** -0.16 -0.11 .326** -0.18 .750* .393** .291* 1.00 K 0.32 -0.29 0.40 0.18 -0.50 0.03 0.45 0.51 0.19 1.00 Mg .911** .476** .482* .865** 1.000** -0.43 .788** -0.03 .335** 0.11 1.00 Mn 0.04 -0.07 -0.29 -0.06 0.00 .750* 0.11 .394** .854** 0.19 0.06 1.00 Mo 0.33 0.36 1.000** 0.58 0.43 -0.36 0.30 0.63 0.38 1.00 Na -0.27 0.38 0.20 -0.24 0.40 -0.20 -0.18 .362* 0.03 .964** -0.12 0.21 - 1.00 1.000** Ni 0.01 -0.27 -0.54 -0.04 0.77 .800** -0.07 0.23 .373* 0.37 -0.14 .332* -0.31 0.12 1.00 P .315* 0.18 0.25 .347** 0.60 .883** .376** .358** .615** 0.47 .324** .505** 0.25 .385* 0.31 1.00 Pb 0.11 0.21 0.01 0.17 - 0.22 .329* 0.23 .291* 0.14 0.48 0.16 -0.05 .403** 1.00 1.000** S 0.00 0.04 -0.14 -0.15 -0.25 0.07 0.04 .485** 0.19 -0.34 -0.01 0.18 0.33 0.20 0.17 .303** 0.22 1.00 Si .557** 0.14 0.21 .528** 0.50 -0.20 .691** 0.01 0.21 -0.40 .476** 0.09 0.27 - -0.19 0.11 0.06 -0.15 1.00 0.29 Sr .748** .552* .570** .753** 1.000** -0.40 .673** -0.30 0.06 -0.20 .651** -0.10 0.14 - - 0.24 0.11 -0.25 .388** 1.00 0.32 .638** V .429** 0.29 0.34 .518** 0.75 0.45 .426** 0.11 .290** 0.24 .456** 0.03 .750* - .339* .376** 0.22 .335** 0.17 .392** 1.00 0.03 Zn 0.19 0.39 0.21 0.17 1.000** -0.26 0.27 0.05 0.14 0.44 .337* 0.13 0.80 0.19 -0.05 0.23 - 0.11 .383* 0.02 0.24 1.00 0.22 82 Table 7 Correlation Coefficient between elements in Lautoka Al As Ba Ca Cd Co Cr Cu Fe K Mg Mn Mo Na Ni P Pb S Si Sr V Zn Al 1.00 As 0.07 1.00 Ba 0.30 0.17 1.00 Ca .788** 0.00 0.17 1.00 Cd Co -0.21 -0.34 0.50 -0.41 1.00 Cr .516** -0.06 0.21 .531** 0.13 1.00 Cu 0.32 -0.11 -0.02 0.09 .560* 0.16 1.00 Fe 0.25 0.00 -0.16 -0.03 .665* -0.02 .569** 1.00 K 0.60 0.10 1.000 0.60 - .943** -0.03 0.20 1.00 ** 1.000 ** Mg .663** -0.04 0.17 .918** -0.36 .642** 0.13 -0.01 0.54 1.00 Mn 0.17 0.04 -0.22 -0.12 .863** -0.13 .554** .975** 0.20 -0.06 1.00 Mo Na 0.18 0.04 0.46 0.36 -0.12 0.17 0.20 0.17 .943** 0.40 0.16 1.00 Ni 0.16 -0.13 -0.40 0.26 0.26 0.24 0.04 -0.18 0.37 0.26 -0.19 0.08 1.00 P 0.19 -0.12 0.03 -0.03 .604* -0.08 .510** .843** 0.03 0.10 .863** 0.39 -0.13 1.00 Pb .418** 0.10 0.41 .524** 0.53 .582** .451** 0.13 .886* .609** 0.07 .584** 0.33 0.19 1.00 S 0.17 0.11 -0.19 -0.08 0.55 -0.14 .487** .688** 0.31 0.03 .765** 0.27 -0.01 .718** 0.14 1.00 Si .493** -0.06 0.20 .475** 0.32 .372* .315* .497** -0.14 .373* .365* 0.34 -0.05 .332* .441** 0.06 1.00 Sr 0.28 -0.06 .857** 0.41 0.02 -0.19 0.09 1.000 0.18 0.08 0.53 -0.04 0.12 0.17 -0.01 0.16 1.00 ** V .500** -0.02 0.15 .337* .736** 0.22 .658** .733** 0.77 .377* .697** 0.38 0.09 .635** .466** .701** .451** 0.17 1.00 Zn .699* -0.36 1.000 .713** -0.50 .685* -0.05 0.13 .682* 0.05 -0.50 -0.20 0.18 .622* 0.00 .629* -0.50 0.38 1.00 ** 83 Significant correlations between crustal elements were also found in Lautoka City. Al had significant positive correlations with Ca, Cr, and Si at 99% level of confidence. Si also demonstrated good correlation between elements such as Ca, Cr, Cu, Fe, Mg, Mn and P. P also had significant correlations with Cu and Fe. The stronger correlation between the crustal elements suggests that crustal elements have greater contributions to the overall PM2.5 in Lautoka than in Suva. Notably, significant correlation between Na and K similar to Suva City was found in Lautoka City. This also suggests significant contribution of marine aerosols in both the cities. Traffic markers also demonstrated good correlation. S had significant positive correlation with Cu, Fe, Mn, P and V at 99% confidence level and at 95% level of confidence with Zn. This is reflective of both vehicular exhaust and non-exhaust emissions. The correlation between Fe and Cu suggests mechanical wear and tear including brake and tire abrasions while S, Mn and V are typical markers of direct exhaust emissions. In Lautoka City, significant correlations were observed for elements characterizing industrial emissions. Significant correlations between Zn and Ba, Ca, Cr, Mg, Al, Pb were noted. Correlations between K and Ba, Cr, Na, and V also existed at 99% level of confidence. This suggests significant contributions from industrial activity and burning. Examining the correlations between signature elements between the two cities, it can be argued that direct traffic emissions are the leading source of pollutants in Suva City. However, while the influence of traffic emissions are an important source, Lautoka City also has large contributions from industrial activity and crustal elements when compared to Suva, resulting in greater particulate matter loading. This argument is further substantiated by including the BC concentrations of the two cities. While PM2.5 concentrations in Lautoka City are greater than three times the average PM2.5 concentration of Suva City, the BC concentrations are approximately the same at 4.0 μg/m3 and 3.9 μg/m3 for Lautoka and Suva (Samabula site) city respectively. This implies that direct combustion contributions are approximately the same for the two cities and therefore, other factors including crustal and industrial emissions are significant contributors to the PM2.5 concentration in Lautoka. 84 4.5.3 Enrichment Factor The enrichment factor (EF) technique is widely used to assess potential sources of aerosols (Shakya et al., 2017). This technique compares the ratio of the element concentration to a reference element in the aerosols with the concentration of that element to the reference element in the upper continental crust (UCC) (Rastogi & Sarin, 2009). The EF was computed according to Equation 5. Al followed by Fe is commonly used as the reference element. Al is regarded as an index of mineral dust and shows good correlation with crustal elements. ቀ ೣൗቁ ಲ ುಾ ܧܨ ൌ (5) ቀ ೣൗቁ ಲ ೆ Where Cx is the concentration of the element Generally, elements having EF values ≤ 1 are regarded as predominantly having crustal origins. Trace metals that have EF’s between 1 and 5 are deemed to be originating from both crustal and anthropogenic sources. Elements with EF values t 5 are primarily emitted through anthropogenic activities (Hsu et al., 2016; Hsu et al., 2010) . Some studies also suggest that elements having EF values less than 5 are predominantly derived from mineral aerosols (Rastogi & Sarin, 2009) and values greater than 10 suggest anthropogenic origins (Shakya et al., 2017). Enrichment factor values were calculated for all the elements including trace metals which were analysed to ascertain the dominant source of emission. EF’s of elements in both the cities are presented in Figure 32. 85 Figure 32 Enrichment Factor (EF) of elements in Suva and Lautoka cities Elements such as Cd, Co, K, Fe, Mg, Mn, Ni, P, and Si showed EF values closer to 1, indicating that these elements were largely derived from crustal sources and natural phenomenon such as dust suspension and geological surface erosions. This is also consistent with the correlations found for these elements and Al. Metals including Cr, Cu, and Na have EF’s greater than 1 but below 5. This suggests that these elements are also largely of crustal origin with some anthropogenic influence. On the contrary, elements such as Ca, Sr, V, Pb showed EF’s close to or greater than 5 but below 10, suggesting anthropogenic emissions as the major sources with some crustal contribution. As, Ba, Mo, S and Zn had higher EF’s suggesting predominant contributions from anthropogenic emissions. While Mo was not detected in Lautoka City, Mo and Ba showed significant correlation in Suva City suggesting the same source of emissions. EF of Cu in Lautoka City was higher than Suva City with a value of 9.2. This suggested predominant anthropogenic emissions. Cu in Lautoka City had significant positive correlation with S (0.487, p=0.01) and traffic is an important source of both these elements. Moreover, elements such as As, Ba, S and Zn had EF’s much greater than 10, implying anthropogenic sources. Zn and Ba were also noted to have strong 86 correlations in Lautoka City, likely to have originated from direct exhaust emissions and tire wear (Hieu & Lee, 2010; Shakya et al., 2017). This reaffirms traffic emissions to be an important source in both the cities. Elements which had the highest concentration difference (Figure 31) between Lautoka and Suva City were Na, Fe, Si, and Mn. These elements were noted to have EF’s less than or close to 1, indicating crustal origins. This further confirms crustal sources to be the dominant contributor to the PM2.5 concentrations in Lautoka City. 4.6 Implications and Policy Comparisons A few approaches could be taken up by Fiji to remedy traffic related emissions, which could serve as a model to other PICs. Currently, the Land Transport Authority (LTA) of Fiji, within the bounds of Land Transport Act of 1998 and Land Transport (Vehicle Registration and Construction) Regulations (2000), has the authority to issue Traffic Infringement Notices (TIN’s) and or Vehicle Defect Notices for vehicles that emit excessive smoke for greater than 10 seconds (High Court of Fiji, 2004; Parliment of the Republic of Fiji, 1998). The current smoke opacity standard in Fiji is 50%. Even with a lower standard as such, around 200 vehicles per month fail smoke opacity test at authorized testing centers (Pratibha, 2017). Rules surrounding on-road vehicle smoke emissions are very rarely enforced (Isley et al., 2016; Secretariat of the Pacific Regional Environment Programme, 2014; Sevura, 2005) in Fiji. Due to the lapse in policy and enforcement, poorly maintained old vehicles (primarily heavy duty diesel vehicles or HDDV) with high emission rates remain in service. This can also be observed considering the high BC concentrations in both the cities. Older diesel vehicles have shown significantly higher emission factors for BC in comparison to new vehicles (Ježek et al., 2015). Comparing smoke emission standards, Singapore has a tighter standard at 40% Hartridge Smoke Unit (HSU) for all vehicles. Moreover, the Environmental Protection and Management (Vehicle Emission) Regulations makes it an offence for any vehicle to emit visible smoke while on the road in Singapore (National Environment Agency, 2019). New Zealand also has a rigorous snap acceleration test for old diesel vehicles, setting the standard to be no more than 25% (Ministry of Transport, 2007). Australia has emission standards analogous to the Euro V emission standards, while a review is 87 in progress to shift to Euro VI guidelines (Australian Government, 2018). The Euro V standards include limits for NOx, HC (hydrocarbons), CO (Carbon Monoxide) and particulate matter (Australian Government, 2018). If Fiji is to attain a sustainable transport sector, then higher standards should be considered as future initiatives by LTA. However, present goals should include training technical staff with expertise in assessing vehicles emissions. Further investments should be made in portable and relevant technologies so that meaningful emission on road checks can be carried out. This will be instrumental in reducing the number of high emitting old vehicles and encourage timely maintenance of in-service vehicles thereby lowering contributions of the transport sector. While monitoring the existing fleet, PICs including Fiji should strengthen laws and regulations surrounding motor vehicle importation in order to protect and improve air quality. Implementing age restrictions for second hand vehicles is an effective start to reducing emissions. From 2019, used diesel and petrol vehicles imported into Fiji must be Euro IV compliant and be no more than 5 years old. Similarly, Liquefied Petroleum Gas (LPG) and special use vehicles should also be Euro IV compliant but are allowed a maximum age of 8 years (Fiji Revenue and Customs Service, 2019). Cities such as Beijing have noted a decrease in air pollutants including BC by replacing old buses with new Euro IV compliant and Compressed Natural Gas (CNG) based buses. Other PICs such as Samoa have also imposed stronger age restriction of 8 years for vehicle importation, however, effectiveness of this policy in Samoa remains a question as old vehicles continue to be imported (Sanerivi, 2017; Tong, 2016). The import age regulations in Fiji are now similar to that of New Zealand, however, Singapore has a much more defined and strict vehicle importation policy where imported vehicles can be no more than 3 years from first registration or manufacture (Land Transport Authority, 2017). Reducing sulfur in fuel is also an effective method to lower emissions. Studies done through modelling show reductions in PM2.5 concentrations by reducing the sulfur content in fuels (Leelasakultum et al., 2012). This option has also proved to be beneficial in reducing maritime transport emissions by lowering SO2 emissions from ships and consequently reducing both primary and secondary aerosol formation 88 (Contini et al., 2015). Blumberg et al. (2003) argued that by reducing the sulfur content in fuels, developing countries can allay mounting human health impacts with increasing vehicles numbers. Moreover, it can also reduce the burden associated with cleaning up existing vehicles. Heavy sulfur based fuels are important sources of particulates in the atmosphere. Many countries have adopted near zero (<10 ppm) fuel sulfur concentrations. Australia currently allows 50 ppm and 10 ppm in petrol and diesel respectively (Australian Government, 2019). New Zealand has standardized 10 ppm for both petrol and diesel since 2018 (New Zealand Government, 2017). Singapore has implemented fuel standards restrictions in line with Euro VI standards, limiting sulfur content to 10 ppm for heavy duty vehicles (National Environment Agency, 2019). Fiji has adopted standards parallel to Australia, with sulfur content limits of 50 ppm and 10 ppm for petrol and diesel, respectively, since 2019. Moreover, the reduction of sulfur in petrol from 50 ppm to 10 ppm will come into effect from 2021 (Government of Fiji, 2019). While these limits were standardized almost a decade later than Australia, it will be a critical measure in reducing transport based emissions. Another approach to strengthen the transport sector and reduce emissions is to assess each vehicles before it can be imported. Fiji has introduced offshore JEVIC (Japan Export Vehicle Inspection Center) inspections for second hand vehicle imports from Japan, New Zealand and Australia, commencing from the 1st of October, 2019 (Land Transport Authority, 2019). Inspections as such have been carried out in New Zealand with further requirements including emissions and fuel consumption certifications (NZ Transportation Agency, 2019). In comparison, Singapore vehicle importation policy has a greater progressive approach to safeguard environmental and public health. Apart from smoke emissions, recently introduced Vehicle Emission Scheme sets out standard for CO2, CO, NOX and particulate matter emissions (Land Transport Authority, 2017). Furthermore, depending on the amount of emissions that the vehicle makes, rebate and penalties have been implemented. These policies encourage people to purchase vehicles with as low emission rates as possible to qualify for a rebate. Moreover, vehicle certifications, apart from safety mechanisms also include diesel smoke testing and fuel consumption testing (Land Transport Authority, 2017). Certifications as such should also be included in the PICs to prevent importation of vehicles which are not economically viable in terms of fuel consumption and 89 performance. This will also be a screening measure to prevent the PICs from becoming a dumping ground for second hand vehicles. Vehicles with a better fuel economy and lower emissions will aid in improving air quality in the PICs. The transport sector of Singapore can act as a model for the developing PICs. Singapore has a unique system that regulates the number of vehicles. To own a vehicle, prospective owners need to bid for a Certificate of Entitlement (COE), which lets the owner use and own a vehicle for 10 years with options to renew the COE thereafter (Ministry of Transport, 2019). This encourages the use of public transportation and effectively regulates vehicle population. While such polices would be difficult to implement in the PICs given the financial limitations, it serves as a guidance to greater strategic development plans. Regulating the transport sector and encouraging public and non-motorized transportation may also raise benefits for the struggling health sectors. Non-motorized transportation may reduce obesity issues (Giles-Corti et al., 2010), an epidemic from which the PICs heavily suffer from (World Health Organisation, 2010). Reducing emissions from the transport sector will also act to reduce greenhouse gas emissions and pollutants including BC. Chapman (2007) highlights that the transport sector is one of the few industrial sectors in which emissions are still growing, accounting for 26 % of global CO2 emissions. Climate change also affects air quality by altering local weather patterns including wind and rain, thereby, influencing the distribution of air pollutants (Younger et al., 2008). Younger et al. (2008) also suggests opting for alternative transport options such as non-motorized transport will reduce transport related greenhouse gas emissions and improve air quality coupled with health benefits. Moreover, to have such policies in place in future, the PICs must first act to improve current infrastructure including public transportation. Implementing a stronger monitoring network to assess the current fleet may prove to be successful with faster results. While Fiji has taken steps that may work in favor of reducing vehicle emissions, localized assessments and characterization of roadside PM2.5 should be done to comprehend the effect of high roadside concentrations on the health of the Fijian population. Other PICs should also follow and improve current existing vehicle importation and use policies if sustainable transport in balance with environment and health sectors is to be achieved in the Pacific. 90 5.0 Summary, Conclusion and Recommendations This study was done to determine the concentrations of PM2.5 in heavy traffic areas in Suva and Lautoka City and to examine the diurnal concentration patterns and elemental composition of PM2.5. At the end of this study, all of the aforementioned objectives have been met, including developing the local capacity to monitor air quality. The daily mean PM2.5 concentration for Suva and Lautoka City were 21.6 ± 13.3 μg/m3 (median, 17.9 μg/m3) and 67.2 ± 35.2 μg/m3 (median, 58.1 μg/m3) respectively. Weekdays showed higher concentrations than weekends. Lower concentrations were also typically observed for the months of November, December and January. Roadside concentrations of PM2.5 in Lautoka City are comparably higher than many urban centres in developed and developing countries. Elevated PM2.5 in Lautoka may have serious public health implications and warrants further air quality related research in Fiji. The influence of meteorological parameters on the concentrations of PM2.5 were also examined in this study. Briefly, rain was found to be negatively correlated (rs = -0.354, p < 0.05) with PM2.5 in Suva City. Similarly, wind speed and relative humidity were found to have negative correlations with PM2.5 concentrations in Suva with a correlation coefficient of -0.671, p < 0.01 and -0.285, p < 0.05 respectively. Only temperature was noted to have a correlation with PM2.5 concentrations in Lautoka. Temperature was positively correlated with PM2.5 concentrations with a correlation coefficient of 0.358, p < 0.05 in Lautoka. This study further supports the literature that rainfall, high wind speed and high relative humidity are important meteorological factors that help in the deposition of PM2.5, thereby reducing its ambient concentrations. Diurnal trends of PM2.5 and black carbon were similar in Samabula, Suva City bus station and Lautoka City sites. This reaffirms that the PM2.5 is largely emitted from combustion driven sources such as vehicle emissions in these areas. The concentrations of PM2.5 determined by the pDR-1500 in Samabula, Suva City bus station, Lautoka City and Reservoir road community were 7.7 ± 4.6 μg /m3, 6.7 ± 5.5 91 μg /m3, 9.8 ± 11.07 μg /m3 and 5.2 ± 8.4 μg /m3 for Samabula, Suva bus station, Lautoka City and Reservoir road community respectively. Similarly, BC concentrations for the sites were 3.9 ± 2.9 μg /m3 and 2.6 ± 2.7 μg /m3, 4.02 ± 4.7 μg /m3 and 2.4 ± 2.3 μg /m3 in Samabula, Suva City bus station, Lautoka City and Reservoir road community respectively. While the diurnal trends are similar between PM2.5 and BC at the first 3 sites, it differs slightly during the early morning hours at the Reservoir road side. During this time, there is an increase in PM2.5 but not in BC concentrations. This could be due to indoor sources such as mosquito coils which are widely used in many homes in Fiji. This study recommends an investigation into indoor air quality in Fiji and the Pacific Island Countries in general. A notable difference was observed in the concentrations reported through the gravimetric analysis using the high volume sampler and the real time instruments. Optical based instruments such as the pDR 1500 monitors concentrations within a range of size only. Particles smaller than certain sizes (100 nm in this case) are not included. This attribute of real time instruments that work on light scattering basis may provide limitations for it to be used in environments with high amount of ultrafine particles. Moreover, the difference in the concentrations reported using the two different techniques reaffirms the influence of "fresh emissions”, a result of direct and indirect emissions from vehicles with high ultrafine particle concentrations. Concentrations of 22 elements, mostly metals were also analysed. Concentrations of Al, Ba, Ca, Fe (Lautoka), K, Na and Zn were greater than 1000 ng/m3. Concentrations of Fe (Suva), Mg, Mn (Lautoka), P, S, Si, Sr were in the range of 11 ng/m3 to 999 ng/m3. As, Cr, Cu, Mn (Suva), Ni, Pb, V had concentrations below 10 ng/m3 while Cd, Co and Mo had concentration less than 1 ng/m3. Co was not detected in Lautoka. A greater number of correlations between traffic related elements were observed in comparison to crustal elements in Suva. Distinctive elements such as V was also observed to have significant correlation with other traffic related markers, particularly, Ca, Cr, Fe, Mg, Mo, Ni, and S suggesting greater influence of direct emission sources in Suva City. In contrast, while there is strong influence of traffic emissions in Lautoka, strong correlations between crustal elements were also observed. This suggests that there is a greater contribution of crustal sources in Lautoka. BC concentrations can be used to supplement this. While the concentrations of PM2.5 between Suva and Lautoka vary by more than 3 times, the BC concentrations are 92 almost the same suggesting similar traffic contributions. Hence, crustal sources were dominant in elevating PM2.5 concentrations in Lautoka. The Enrichment Factor (EF) technique was also used to identify the most likely origin of the elements that were analysed. Using the EF, elements including Cd, Co, K, Fe, Mg, Mn, Ni, P, and Si can be said to be derived from crustal and natural sources, primarily through dust resuspension and natural surface erosion. Cr, Cu, and Na were noted to be largely having crustal origin with some anthropogenic influence. EF’s of elements such as Ca, Sr, V, Pb suggests the source to be predominantly anthropogenic with some crustal contribution. As, Ba, Mo, S and Zn demonstrated higher EF’s, originating from anthropogenic emissions. Concentrations of traffic related PM2.5 were greater than twice the WHO daily guidelines in Lautoka. This poses a health risk to the exposed populations and warrants mitigation actions. Moreover, the PM2.5 concentrations in Suva are likely to exceed the WHO guideline if emissions are not regulated. Strengthening environmental polices is a fundamental approach in improving air quality. While this study provides data and recommendation to supplement emission control policies, it strongly recommends further air quality studies to be done in other areas of Fiji. The sampling time, particularly, in Lautoka should be increased to assess the impact of emissions from sugarcane production and field burning. Analysis of inorganic ions and carbonaceous compounds should also be included in further research to aid in the comprehension of the composition of PM2.5. This can also be used in source apportionment studies to ascertain the emission sources and contribution of each source to the overall PM2.5 in Suva and Lautoka. Scanning Electron Microscope (SEM) analysis can also be done to examine the morphological qualities of particulates which would further aid in understanding the source of the particles and the possible health effects. 93 Bibliography A&D Australasia Pty Ltd. (2017). Microbalance. Retrieved 11/06/2018, from https://www.andweighing.com.au/products-service/scientific-balances/bm- series-micro-analytical-balances Adams, H. S., Nieuwenhuijsen, M. J., Colvile, R. N., McMullen, M. A. S., & Khandelwal, P. (2001). Fine particle (PM2.5) personal exposure levels in transport microenvironments, London, UK. Science of The Total Environment, 279(1), 29-44. doi: https://doi.org/10.1016/S0048-9697(01)00723-9 Adams, H. S., Nieuwenhuijsen, M. J., Colvile, R. N., Older, M. J., & Kendall, M. (2002). Assessment of road users’ elemental carbon personal exposure levels, London, UK. Atmospheric Environment, 36(34), 5335-5342. doi: https://doi.org/10.1016/S1352-2310(02)00637-4 Adar, S. D., Filigrana, P. A., Clements, N., & Peel, J. L. (2014). Ambient Coarse Particulate Matter and Human Health: A Systematic Review and Meta- Analysis. Current environmental health reports, 1(3), 258-274. doi: 10.1007/s40572-014-0022-z Agilent Technologies. (2010). Agilent 700 Series Optical Emission Spectrometers. Retrieved 22/11/2019, from http://photos.labwrench.com/equipmentManuals/7566-3082.pdf Air Visual. (2019). Westminister Air Quality Index and PM2.5 Pollution. Retrieved 21/05/2019, 2019, from https://www.airvisual.com/usa/california/westminster Amato, F., Cassee, F. R., Denier van der Gon, H. A. C., Gehrig, R., Gustafsson, M., Hafner, W., Harrison, R. M., Jozwicka, M., Kelly, F. J., Moreno, T., Prevot, A. S. H., Schaap, M., Sunyer, J., & Querol, X. (2014). Urban air quality: The challenge of traffic non-exhaust emissions. Journal of Hazardous Materials, 275, 31-36. doi: https://doi.org/10.1016/j.jhazmat.2014.04.053 Annegarn, H. J., Storms, H., van Grieken, R. E., & Booth-Jones, P. A. (1987). Composition and size of individual particles from a gold mine atmosphere. 94 Mining Science and Technology, 5(2), 111-119. doi: https://doi.org/10.1016/S0167-9031(87)90345-8 Artı́ñano, B., Salvador, P., Alonso, D. G., Querol, X., & Alastuey, A. (2004). Influence of traffic on the PM10 and PM2.5 urban aerosol fractions in Madrid (Spain). Science of The Total Environment, 334-335, 111-123. doi: https://doi.org/10.1016/j.scitotenv.2004.04.032 Asian Development Bank. (2017). Republic of Fiji: Hospital and Health Service Review. Retrieved 15/10/2019, from https://events.development.asia/system/files/materials/2017/10/201710-fiji- hospital-and-health-service-review.pdf Australian Government. (2018). Vehicle Emission Standards. Retrieved 19/09/2019, from https://www.infrastructure.gov.au/vehicles/environment/emission/index.aspx Australian Government. (2019). Fuel Quality Standards. Retrieved 21/09/2019, from http://www.environment.gov.au/protection/fuel-quality/standards Aw, J., & Kleeman, M. J. (2003). Evaluating the first‐order effect of intraannual temperature variability on urban air pollution. Journal of Geophysical Research: Atmospheres, 108(D12). Babich, P., Davey, M., Allen, G., & Koutrakis, P. (2000). Method Comparisons for Particulate Nitrate, Elemental Carbon, and PM2.5 Mass in Seven U.S. Cities. Journal of the Air & Waste Management Association, 50(7), 1095-1105. doi: 10.1080/10473289.2000.10464152 Bachmann, J. (2007). Will the Circle Be Unbroken: A History of the U.S. National Ambient Air Quality Standards. Journal of the Air & Waste Management Association, 57(6), 652-697. doi: 10.3155/1047-3289.57.6.652 Baumann, R., Krzyzanowski, M., & Chicherin, S. (2006). Framework Plan for the Development of Monitoring of Particlate Matter in EECCA. In W. R. O. f. Europe (Ed.). Denmark: World Health Organisation. 95 Bergstra, A. D., Brunekreef, B., & Burdorf, A. (2018). The effect of industry-related air pollution on lung function and respiratory symptoms in school children. Environ Health, 17(1), 30. doi: 10.1186/s12940-018-0373-2 Blumberg, K. O., Walsh, M. P., & Pera, C. (2003). Low-sulfur gasoline & diesel: the key to lower vehicle emissions. Available online [April 24] at:< http://www. theicct. org/documents/Low-Sulfur_Exec_Summ_ICCT_2003. pdf. Briggs, D. J., de Hoogh, K., Morris, C., & Gulliver, J. (2008). Effects of travel mode on exposures to particulate air pollution. Environment International, 34(1), 12- 22. doi: https://doi.org/10.1016/j.envint.2007.06.011 Briggs, N. L., & Long, C. M. (2016). Critical review of black carbon and elemental carbon source apportionment in Europe and the United States. Atmospheric Environment, 144, 409-427. doi: https://doi.org/10.1016/j.atmosenv.2016.09.002 Brown, A. S., Yardley, R. E., Quincey, P. G., & Butterfield, D. M. (2006). Studies of the effect of humidity and other factors on some different filter materials used for gravimetric measurements of ambient particulate matter. Atmospheric Environment, 40(25), 4670-4678. doi: https://doi.org/10.1016/j.atmosenv.2006.04.028 Cai, J., Yan, B., Kinney, P. L., Perzanowski, M. S., Jung, K.-H., Li, T., Xiu, G., Zhang, D., Olivo, C., Ross, J., Miller, R. L., & Chillrud, S. N. (2013). Optimization approaches to ameliorate humidity and vibration related issues using the microAeth black carbon monitor for personal exposure measurement. Aerosol science and technology : the journal of the American Association for Aerosol Research, 47(11), 1196-1204. doi: 10.1080/02786826.2013.829551 Cai, J., Yan, B., Ross, J., Zhang, D., Kinney, P. L., Perzanowski, M. S., Jung, K., Miller, R., & Chillrud, S. N. (2014). Validation of MicroAeth® as a Black Carbon Monitor for Fixed-Site Measurement and Optimization for Personal Exposure Characterization. Aerosol and Air Quality Research, 14(1), 1-9. doi: 10.4209/aaqr.2013.03.0088 96 Caiazzo, F., Ashok, A., Waitz, I. A., Yim, S. H. L., & Barrett, S. R. H. (2013). Air pollution and early deaths in the United States. Part I: Quantifying the impact of major sectors in 2005. Atmospheric Environment, 79, 198-208. doi: https://doi.org/10.1016/j.atmosenv.2013.05.081 Calderon, V. (2010). Analysis of Environment Samples with Agilent 730-ES Following USEPA Guidelines. Retrieved 22/11/2019, from https://www.agilent.com/cs/library/applications/IO-038.pdf Carpenter, M. E. (2018). How Do Mountains affect Precipitation? Retrieved 25/06/2019, 2019, from https://sciencing.com/do-mountains-affect- precipitation-8691099.html Cesari, D., Donateo, A., Conte, M., Merico, E., Giangreco, A., Giangreco, F., & Contini, D. (2016). An inter-comparison of PM2.5 at urban and urban background sites: Chemical characterization and source apportionment. Atmospheric Research, 174-175, 106-119. doi: https://doi.org/10.1016/j.atmosres.2016.02.004 Chaloulakou, A., Kassomenos, P., Spyrellis, N., Demokritou, P., & Koutrakis, P. (2003). Measurements of PM10 and PM2.5 particle concentrations in Athens, Greece. Atmospheric Environment, 37(5), 649-660. doi: https://doi.org/10.1016/S1352-2310(02)00898-1 Chan, Y. C., Simpson, R. W., McTainsh, G. H., Vowles, P. D., Cohen, D. D., & Bailey, G. M. (1999). Source apportionment of visibility degradation problems in Brisbane (Australia) using the multiple linear regression techniques (Vol. 33). Chand, A. (2018). Ministry of Health, World Health Organisation Joint Statement on NCD Rates in Fiji. Retrieved 13/09/19, from https://www.health.gov.fj/?p=7042 Chand, S. (2019). Traffic Congestion in Cities in Fiji. In S. Mani (Ed.). Chapman, L. (2007). Transport and climate change: a review. Journal of Transport Geography, 15(5), 354-367. doi: https://doi.org/10.1016/j.jtrangeo.2006.11.008 97 Chen, J. (2014). Impact of Relative Humidity and Water Soluble Constituents of PM2.5 on Visibility Impairment in Beijing, China. Aerosol Air Quality Research, 14, 260-268. doi: 10.4209/aaqr.2012.12.0360 Chen, S.-Y., Chan, C.-C., & Su, T.-C. (2017). Particulate and gaseous pollutants on inflammation, thrombosis, and autonomic imbalance in subjects at risk for cardiovascular disease. Environmental Pollution, 223, 403-408. doi: https://doi.org/10.1016/j.envpol.2017.01.037 Cheng, Z., Luo, L., Wang, S., Wang, Y., Sharma, S., Shimadera, H., Wang, X., Bressi, M., de Miranda, R. M., Jiang, J., Zhou, W., Fajardo, O., Yan, N., & Hao, J. (2016). Status and characteristics of ambient PM2.5 pollution in global megacities. Environment International, 89-90, 212-221. doi: https://doi.org/10.1016/j.envint.2016.02.003 Chow, J. C., Watson, J. G., Edgerton, S. A., & Vega, E. (2002). Chemical composition of PM2.5 and PM10 in Mexico City during winter 1997. Science of The Total Environment, 287(3), 177-201. doi: https://doi.org/10.1016/S0048- 9697(01)00982-2 Chuersuwan, N., Nimrat, S., Lekphet, S., & Kerdkumrai, T. (2008). Levels and major sources of PM2.5 and PM10 in Bangkok Metropolitan Region. Environment International, 34(5), 671-677. doi: https://doi.org/10.1016/j.envint.2007.12.018 Cohen, A. J., Ross Anderson, H., Ostro, B., Pandey, K. D., Krzyzanowski, M., Künzli, N., Gutschmidt, K., Pope, A., Romieu, I., Samet, J. M., & Smith, K. (2005). The Global Burden of Disease Due to Outdoor Air Pollution. Journal of Toxicology and Environmental Health, Part A, 68(13-14), 1301-1307. doi: 10.1080/15287390590936166 Contini, D., Gambaro, A., Donateo, A., Cescon, P., Cesari, D., Merico, E., Belosi, F., & Citron, M. (2015). Inter-annual trend of the primary contribution of ship emissions to PM2.5 concentrations in Venice (Italy): Efficiency of emissions mitigation strategies. Atmospheric Environment, 102, 183-190. doi: https://doi.org/10.1016/j.atmosenv.2014.11.065 98 Davis, M. E., Blicharz, A. P., Hart, J. E., Laden, F., Garshick, E., & Smith, T. J. (2007). Occupational exposure to volatile organic compounds and aldehydes in the U.S. trucking industry. Environmental science & technology, 41(20), 7152- 7158. doi: 10.1021/es071041z Davy, P., Trompetter, W. J., Ancelet, T., & Markwitz, A. (2017). Source Apportionment and trend analysis of air particulate matter in the Auckland region Auckland Council Technical Report. New Zealand: Institute of Geological and Nuclear Sciences. Dawson, J. P., Adams, P. J., & Pandis, S. N. (2007). Sensitivity of PM2.5 to climate in the Eastern US: a modeling case study. Atmos. Chem. Phys., 7(16), 4295-4309. doi: 10.5194/acp-7-4295-2007 De Marco, A., Amoatey, P., Khaniabadi, Y. O., Sicard, P., & Hopke, P. K. (2018). Mortality and morbidity for cardiopulmonary diseases attributed to PM2.5 exposure in the metropolis of Rome, Italy. European Journal of Internal Medicine. doi: https://doi.org/10.1016/j.ejim.2018.07.027 Deng, H., Eckel, S. P., Liu, L., Lurmann, F. W., Cockburn, M. G., & Gilliland, F. D. (2017a). Particulate matter air pollution and liver cancer survival. International Journal of Cancer. Journal International du Cancer, 141(4), 744-749. doi: http://dx.doi.org/10.1002/ijc.30779 Deng, X., Feng, N., Zheng, M., Ye, X., Lin, H., Yu, X., Gan, Z., Fang, Z., Zhang, H., Gao, M., Zheng, Z.-j., Yu, H., Ding, W., & Qian, B. (2017b). PM2.5 exposure- induced autophagy is mediated by lncRNA loc146880 which also promotes the migration and invasion of lung cancer cells. Biochimica et Biophysica Acta (BBA) - General Subjects, 1861(2), 112-125. doi: https://doi.org/10.1016/j.bbagen.2016.11.009 Deo, R. (2018, 9/07/2018). Traffic Congestion frustrates PSV operators, News, The Fiji Times. Retrieved from https://www.fijitimes.com/traffic-congestion- frustrates-psv-operators/ 99 Ecotech. (2013). HiVol 3000 High Volume Air Sampler User Manual. Retrieved 06/06/2018, Version 1.8, from https://www.ecotech.com/wp- content/uploads/2015/02/HiVol-3000-Manual.pdf Ecotech. (2017). HiVol 3000 High Volume Air Sampler Spec Sheet. Retrieved 09/06/2018, from https://www.ecotech.com/wp- content/uploads/2015/02/ECOTECH-HiVol-3000-spec-sheet-20171130.pdf Environment Protection Authority Victoria. (2018). PM2.5 Particles in the Air. Air. Retrieved 05/09/18, from https://www.epa.vic.gov.au/your- environment/air/air-pollution/pm25-particles-in-air Esworthy, R. (2015). Air quality: EPA'S 2013 changes to the particulate matter (PM) standard. Retrieved 02/04/2019, from https://fas.org/sgp/crs/misc/R42934.pdf European Environment Agency. (2019). Air Quality in Europe - 2019 Report. Retrieved 17/09/2018, from https://www.developmentaid.org/api/frontend/cms/uploadedImages/2019/10/ Air-quality-in-europe_2019-final.pdf Fang, X., Fang, B., Wang, C., Xia, T., Bottai, M., Fang, F., & Cao, Y. (2017). Relationship between fine particulate matter, weather condition and daily non- accidental mortality in Shanghai, China: A Bayesian approach. PLOS ONE, 12(11), e0187933. doi: 10.1371/journal.pone.0187933 Ferm, M., & Sjöberg, K. (2015). Concentrations and emission factors for PM2.5 and PM10 from road traffic in Sweden. Atmospheric Environment, 119, 211-219. doi: https://doi.org/10.1016/j.atmosenv.2015.08.037 Figueroa, D. A., Rodríguez-Sierra, C. J., & Jiménez-Velez, B. D. (2006). Concentrations of Ni and V, other heavy metals, arsenic, elemental and organic carbon in atmospheric fine particles (PM2.5) from Puerto Rico. Toxicology and Industrial Health, 22(2), 87-99. doi: 10.1191/0748233706th247oa Fiji Bureau of Statistics. (2018a). 2017 Population and Housing Census: Release 1 Age, Sex, Geography and Economic Activity 2017 Population and Housing Census (Vol. 1). Fiji: Fiji Burea of Statistics. 100 Fiji Bureau of Statistics. (2018b). Distribution of Vehicles Registered in Fiji. Retrieved 03/11/2018, from https://www.statsfiji.gov.fj/index.php/statistics/other-statistics/registered- vehicles Fiji Meterological Service. (2019). Fiji Climate Summary May 2019. Retrieved 06/06/2019, from http://www.met.gov.fj/index.php?page=climatedatalatest#May%202019clima teSum2019.06.07%2013.47.26.pdf Fiji Revenue and Customs Authority. (2016). Revenue Measure Introduced in the 2016/2017 Budget Address Fiji Revenue and Customs Authority Public Circular (Vol. No. 4 of 2016). Fiji. Fiji Revenue and Customs Authority. (2017). The Fijian Government Tax Policies: 2017-2018 Budget Highlights. Retrieved 29/09/2018, from https://www.frcs.org.fj/wp-content/uploads/2015/11/2016-2017-Budget- Circular.pdf Fiji Revenue and Customs Service. (2019). Motor Vehicle Landing Cost. Retrieved 30/07/19, from https://www.frcs.org.fj/our-services/customs/doing-business- in-fiji/motor-vehicles/ Fiji Roads Authority. (2014). Greater Suva Transport Strategy. Retrieved 09/10/2018, from http://www.fijiroads.org/capital-work/ Fiordelisi, A., Piscitelli, P., Trimarco, B., Coscioni, E., Iaccarino, G., & Sorriento, D. (2017). The mechanisms of air pollution and particulate matter in cardiovascular diseases. Heart Failure Reviews, 22(3), 337-347. doi: http://dx.doi.org/10.1007/s10741-017-9606-7 Fischer, S. L., & Koshland, C. P. (2006). Field performance of a nephelometer in rural kitchens: effects of high humidity excursions and correlations to gravimetric analyses. Journal Of Exposure Science And Environmental Epidemiology, 17, 141. doi: 10.1038/sj.jes.7500486 101 Fu, X., Wang, S., Chang, X., Cai, S., Xing, J., & Hao, J. (2016). Modeling analysis of secondary inorganic aerosols over China: pollution characteristics, and meteorological and dust impacts. Scientific reports, 6, 35992. Gehrig, R., & Buchmann, B. (2003). Characterising seasonal variations and spatial distribution of ambient PM10 and PM2.5 concentrations based on long-term Swiss monitoring data. Atmospheric Environment, 37(19), 2571-2580. doi: https://doi.org/10.1016/S1352-2310(03)00221-8 Gharibvand, L., Beeson, W. L., Shavlik, D., Knutsen, R., Ghamsary, M., Soret, S., & Knutsen, S. F. (2017). The association between ambient fine particulate matter and incident adenocarcinoma subtype of lung cancer. Environmental Health, 16. doi: http://dx.doi.org/10.1186/s12940-017-0268-7 Giles-Corti, B., Foster, S., Shilton, T., & Falconer, R. (2010). The co-benefits for health of investing in active transportation. New South Wales Public Health Bulletin, 21(6), 122-127. doi: https://doi.org/10.1071/NB10027 Gillies, J. A., Gertler, A. W., Sagebiel, J. C., & Dippel, W. A. (2001). On-Road Particulate Matter (PM2.5 and PM10) Emissions in the Sepulveda Tunnel, Los Angeles, California. Environmental science & technology, 35(6), 1054-1063. doi: 10.1021/es991320p Government of Fiji. (2019). Trade Standards (Fuel Standards) (Ammendemend) Order 2018. Retrieved 23/09/19, from https://www.fiji.gov.fj/getattachment/2288162d-04d5-4ab2-8778- 392ab0dee5af/LN-76-Trade-Standards-(Fuel-Standards)-(Amendment).aspx Grote, M., Williams, I., Preston, J., & Kemp, S. (2016). Including congestion effects in urban road traffic CO2 emissions modelling: Do Local Government Authorities have the right options? Transportation Research Part D: Transport and Environment, 43, 95-106. doi: https://doi.org/10.1016/j.trd.2015.12.010 Guerra, S. A., Lane, D. D., Marotz, G. A., Carter, R. E., Hohl, C. M., & Baldauf, R. W. (2006). Effects of Wind Direction on Coarse and Fine Particulate Matter Concentrations in Southeast Kansas. Journal of the Air & Waste Management Association, 56(11), 1525-1531. doi: 10.1080/10473289.2006.10464559 102 Guo, Y., Zeng, H., Zheng, R., Li, S., Pereira, G., Liu, Q., Chen, W., & Huxley, R. (2017). The burden of lung cancer mortality attributable to fine particles in China. Science of The Total Environment, 579, 1460-1466. doi: https://doi.org/10.1016/j.scitotenv.2016.11.147 Gutiérrez-Avila, I., Rojas-Bracho, L., Riojas-Rodríguez, H., Kloog, I., Just, A. C., & Rothenberg, S. J. (2018). Cardiovascular and Cerebrovascular Mortality Associated With Acute Exposure to PM2.5 in Mexico City. Stroke, 49(7), 1734-1736. doi: 10.1161/strokeaha.118.021034 Harris, S. J., & Maricq, M. M. (2001). Signature size distributions for diesel and gasoline engine exhaust particulate matter. Journal of Aerosol Science, 32(6), 749-764. doi: https://doi.org/10.1016/S0021-8502(00)00111-7 Harrison, R. M., Deacon, A. R., Jones, M. R., & Appleby, R. S. (1997). Sources and processes affecting concentrations of PM10 and PM2.5 particulate matter in Birmingham (U.K.). Atmospheric Environment, 31(24), 4103-4117. doi: https://doi.org/10.1016/S1352-2310(97)00296-3 Harrison, R. M., Jones, A. M., & Lawrence, R. G. (2004). Major component composition of PM10 and PM2.5 from roadside and urban background sites. Atmospheric Environment, 38(27), 4531-4538. doi: https://doi.org/10.1016/j.atmosenv.2004.05.022 Haworth, S., & Starkey, P. (2017). Greeing Fiji's Bus Fleet With Bus Renewal and Scrappage Plan: Findings and Reccomendations. Fiji: Asian Development Bank. Heal, M. R., Hibbs, L. R., Agius, R. M., & Beverland, I. J. (2005). Total and water- soluble trace metal content of urban background PM10, PM2.5 and black smoke in Edinburgh, UK. Atmospheric Environment, 39(8), 1417-1430. doi: https://doi.org/10.1016/j.atmosenv.2004.11.026 HEI Panel on the Health Effects of Traffic Related Air Pollution. (2010). Traffic Related Air Pollution: A Critical Review of the Literature on Emissions, Exposure, and Health Effects. HEI Special Report (Vol. 17). Boston, U.S.A Health Effects Institue. 103 Heo, J. B., Hopke, P. K., & Yi, S. M. (2009). Source apportionment of PM2.5 in Seoul, Korea. Atmos. Chem. Phys., 9(14), 4957-4971. doi: 10.5194/acp-9-4957-2009 Hieu, N. T., & Lee, B.-K. (2010). Characteristics of particulate matter and metals in the ambient air from a residential area in the largest industrial city in Korea. Atmospheric Research, 98(2), 526-537. doi: https://doi.org/10.1016/j.atmosres.2010.08.019 High Court of Fiji. (2004). State v Land Transport Authority, Ex: parte Fiji Taxi Union. Fiji: High Court of Fiji. Hime, N., Marks, G., & Cowie, C. (2018). A comparison of the health effects of ambient particulate matter air pollution from five emission sources. International journal of environmental research and public health, 15(6), 1206. Hitchins, J., Morawska, L., Wolff, R., & Gilbert, D. (2000). Concentrations of submicrometre particles from vehicle emissions near a major road. Atmospheric Environment, 34(1), 51-59. doi: https://doi.org/10.1016/S1352- 2310(99)00304-0 Hodan, W. M., & Barnard, W. R. (2004). Evaluating the contribution of PM2. 5 precursor gases and re-entrained road emissions to mobile source PM2. 5 particulate matter emissions. Retrieved 04/04/2019, from https://pdfs.semanticscholar.org/29f2/923b16b1e496233b6de6fe2b1bb13261 ba39.pdf?_ga=2.205276180.58838575.1589339769- 2046212258.1589339769 Horne, B. D., Joy, E. A., Hofmann, M. G., Gesteland, P. H., Cannon, J. B., Lefler, J. S., Blagev, D. P., Korgenski, E. K., Torosyan, N., Hansen, G. I. J. A. j. o. r., & medicine, c. c. (2018). Short-term elevation of fine particulate matter air pollution and acute lower respiratory infection. (ja). Hsu, C.-Y., Chiang, H.-C., Lin, S.-L., Chen, M.-J., Lin, T.-Y., & Chen, Y.-C. (2016). Elemental characterization and source apportionment of PM10 and PM2.5 in 104 the western coastal area of central Taiwan. Science of The Total Environment, 541, 1139-1150. doi: https://doi.org/10.1016/j.scitotenv.2015.09.122 Hsu, S. C., Liu, S. C., Tsai, F., Engling, G., Lin, I. I., Chou, C. K. C., Kao, S. J., Lung, S. C. C., Chan, C. Y., Lin, S. C., Huang, J. C., Chi, K. H., Chen, W. N., Lin, F. J., Huang, C. H., Kuo, C. L., Wu, T. C., & Huang, Y. T. (2010). High wintertime particulate matter pollution over an offshore island (Kinmen) off southeastern China: An overview. Journal of Geophysical Research: Atmospheres, 115(D17). doi: 10.1029/2009jd013641 Hu, S., Fruin, S., Kozawa, K., Mara, S., Paulson, S. E., & Winer, A. M. (2009). A wide area of air pollutant impact downwind of a freeway during pre-sunrise hours. Atmospheric Environment, 43(16), 2541-2549. doi: https://doi.org/10.1016/j.atmosenv.2009.02.033 Isley, C. F., Nelson, P., & Taylor, M. (2016). Managing air quality in Suva, Fiji. European Journal of Sustainable Development, 5, 242-254. doi: 10.14207/ejsd.2016.v5n4p242 Isley, C. F., Nelson, P. F., Taylor, M. P., Mani, F. S., Maata, M., Atanacio, A., Stelcer, E., & Cohen, D. D. (2017). PM2.5 and aerosol black carbon in Suva, Fiji. Atmospheric Environment, 150, 55-66. doi: https://doi.org/10.1016/j.atmosenv.2016.11.041 Isley, C. F., Nelson, P. F., Taylor, M. P., Stelcer, E., Atanacio, A. J., Cohen, D. D., Mani, F. S., & Maata, M. (2018). Reducing mortality risk by targeting specific air pollution sources: Suva, Fiji. Science of The Total Environment, 612, 450- 461. doi: https://doi.org/10.1016/j.scitotenv.2017.08.225 Isley, C. F., & Taylor, M. P. (2018). Air quality management in the Pacific Islands: A review of past performance and implications for future directions. Environmental Science & Policy, 84, 26-33. doi: https://doi.org/10.1016/j.envsci.2018.02.013 Jacobson, M. Z. (2001). Strong radiative heating due to the mixing state of black carbon in atmospheric aerosols. Nature, 409, 695. doi: 10.1038/35055518 105 https://www.nature.com/articles/35055518#supplementary-information Janssen, N. (2012). Health Effects of Black Carbon. Denmark: World Health Organisation. Janssen, N. A. H., Hoek, G., Simic-Lawson, M., Fischer, P., Bree, L. v., Brink, H. t., Keuken, M., Atkinson, R. W., Anderson, H. R., Brunekreef, B., & Cassee, F. R. (2011). Black Carbon as an Additional Indicator of the Adverse Health Effects of Airborne Particles Compared with PM10 and PM2.5. Environmental health perspectives, 119(12), 1691-1699. doi: doi:10.1289/ehp.1003369 Jeong, C.-H., Wang, J. M., Hilker, N., Debosz, J., Sofowote, U., Su, Y., Noble, M., Healy, R. M., Munoz, T., Dabek-Zlotorzynska, E., Celo, V., White, L., Audette, C., Herod, D., & Evans, G. J. (2019). Temporal and spatial variability of traffic-related PM2.5 sources: Comparison of exhaust and non-exhaust emissions. Atmospheric Environment, 198, 55-69. doi: https://doi.org/10.1016/j.atmosenv.2018.10.038 Jeong, C. H., Wang, J. M., & Evans, G. J. (2016). Source Apportionment of Urban Particulate Matter using Hourly Resolved Trace Metals, Organics, and Inorganic Aerosol Components. Atmos. Chem. Phys. Discuss., 2016, 1-32. doi: 10.5194/acp-2016-189 Ježek, I., Katrašnik, T., Westerdahl, D., & Močnik, G. (2015). Black carbon, particle number concentration and nitrogen oxide emission factors of random in-use vehicles measured with the on-road chasing method. Atmos. Chem. Phys., 15(19), 11011-11026. doi: 10.5194/acp-15-11011-2015 Juswanto, W. (2016). Economic growth and sustainable development in the Pacific Island countries. Retrieved 02/07/2018, from https://www.adb.org/publications/economic-growth-and-sustainable- development-pacific-island-countries Karagulian, F., Belis, C. A., Dora, C. F. C., Prüss-Ustün, A. M., Bonjour, S., Adair- Rohani, H., & Amann, M. (2015). Contributions to cities' ambient particulate matter (PM): A systematic review of local source contributions at global level. 106 Atmospheric Environment, 120, 475-483. doi: https://doi.org/10.1016/j.atmosenv.2015.08.087 Karagulian, F., Van Dingenen, R., Janssens-Maenhout, G., Crippa, M., Guizzardi, D., & Dentener, F. (2016). Attribution of anthropogenic PM2.5 to emission sources: A global analysis of source-receptor model results and measured source-apportionment data. JRC Technical Reports. doi: 10.2760/344371 Kassomenos, P. A., Vardoulakis, S., Chaloulakou, A., Paschalidou, A. K., Grivas, G., Borge, R., & Lumbreras, J. (2014). Study of PM10 and PM2.5 levels in three European cities: Analysis of intra and inter urban variations. Atmospheric Environment, 87, 153-163. doi: https://doi.org/10.1016/j.atmosenv.2014.01.004 Kelly, F. J., & Fussell, J. C. (2015). Air pollution and public health: emerging hazards and improved understanding of risk. Environmental geochemistry and health, 37(4), 631-649. doi: 10.1007/s10653-015-9720-1 Ketzel, M., Omstedt, G., Johansson, C., Düring, I., Pohjola, M., Oettl, D., Gidhagen, L., Wåhlin, P., Lohmeyer, A., Haakana, M., & Berkowicz, R. (2007). Estimation and validation of PM2.5/PM10 exhaust and non-exhaust emission factors for practical street pollution modelling. Atmospheric Environment, 41(40), 9370-9385. doi: https://doi.org/10.1016/j.atmosenv.2007.09.005 Kinney, P. L., Gichuru, M. G., Volavka-Close, N., Ngo, N., Ndiba, P. K., Law, A., Gachanja, A., Gaita, S. M., Chillrud, S. N., & Sclar, E. (2011). Traffic impacts on PM2.5 air quality in Nairobi, Kenya. Environmental Science & Policy, 14(4), 369-378. doi: https://doi.org/10.1016/j.envsci.2011.02.005 Krzyzanowski, M., & Cohen, A. (2008). Update of WHO air quality guidelines. Air Quality, Atmosphere & Health, 1(1), 7-13. doi: 10.1007/s11869-008-0008-9 Kulshrestha, A., Satsangi, P. G., Masih, J., & Taneja, A. (2009). Metal concentration of PM2.5 and PM10 particles and seasonal variations in urban and rural environment of Agra, India. Science of The Total Environment, 407(24), 6196- 6204. doi: https://doi.org/10.1016/j.scitotenv.2009.08.050 107 Kumar, A. (2017, 20/05/2017). Loss in Productivity Linked to Traffic Jams, Fiji Sun. Retrieved from http://fijisun.com.fj/2017/05/20/loss-in-productivity-linked- to-traffic-jams/ Kurt, O. K., Zhang, J., & Pinkerton, K. E. (2016). Pulmonary health effects of air pollution. Current opinion in pulmonary medicine, 22(2), 138-143. doi: 10.1097/MCP.0000000000000248 Lacanivalu, L. (2017). A-G: Cement can be imported; zero duty rates. Retrieved 05/11/2019, from https://www.pressreader.com/fiji/fiji- sun/20170609/282449938999278 Land Transport Authority. (2017). Procedures of Importation and Registration of Cars in Singapore. Retrieved 19/09/19, from https://www.onemotoring.com.sg/content/dam/onemotoring/pdf_2017/Car.pd f Land Transport Authority. (2019). Importation of Vehicle and Quality Standards Protecting Fijian Consumers. Retrieved 15/09/19, from https://lta.com.fj/news/2019/09/11/importation-of-vehicles-and-quality- standards-protecting-fiji-consumers Lee, P. K. H., Brook, J. R., Dabek-Zlotorzynska, E., & Mabury, S. A. (2003). Identification of the Major Sources Contributing to PM2.5 Observed in Toronto. Environmental science & technology, 37(21), 4831-4840. doi: 10.1021/es026473i Lee, S. C., Cheng, Y., Ho, K. F., Cao, J. J., Louie, P. K. K., Chow, J. C., & Watson, J. G. (2006). PM1.0 and PM2.5 Characteristics in the Roadside Environment of Hong Kong. Aerosol Science and Technology, 40(3), 157-165. doi: 10.1080/02786820500494544 Leelasakultum, K., Mölders, N., Tran, H. N. Q., & Grell, G. A. (2012). Potential Impacts of the Introduction of Low-Sulfur Fuel on PM2.5 Concentrations at Breathing Level in a Subarctic City. Advances in Meteorology, 2012, 16. doi: 10.1155/2012/427078 108 Liu, Q., Xu, C., Ji, G., Liu, H., Shao, W., Zhang, C., Gu, A., & Zhao, P. (2017). Effect of exposure to ambient PM(2.5) pollution on the risk of respiratory tract diseases: a meta-analysis of cohort studies. Journal of biomedical research, 31(2), 130-142. doi: 10.7555/JBR.31.20160071 Liu, Y., Xing, J., Wang, S., Fu, X., & Zheng, H. (2018). Source-specific speciation profiles of PM2.5 for heavy metals and their anthropogenic emissions in China. Environmental Pollution, 239, 544-553. doi: https://doi.org/10.1016/j.envpol.2018.04.047 Ljubimova, J. Y., Braubach, O., Patil, R., Chiechi, A., Tang, J., Galstyan, A., Shatalova, E. S., Kleinman, M. T., Black, K. L., & Holler, E. (2018). Coarse particulate matter (PM2.5–10) in Los Angeles Basin air induces expression of inflammation and cancer biomarkers in rat brains. Scientific reports, 8(1), 5708. doi: 10.1038/s41598-018-23885-3 López, M. L., Ceppi, S., Palancar, G. G., Olcese, L. E., Tirao, G., & Toselli, B. M. (2011). Elemental concentration and source identification of PM10 and PM2.5 by SR-XRF in Córdoba City, Argentina. Atmospheric Environment, 45(31), 5450-5457. doi: https://doi.org/10.1016/j.atmosenv.2011.07.003 Lou, C., Liu, H., Li, Y., Peng, Y., Wang, J., & Dai, L. (2017). Relationships of relative humidity with PM2.5 and PM10 in the Yangtze River Delta, China. Environmental Monitoring and Assessment, 189(11), 582. doi: 10.1007/s10661-017-6281-z Lu, C., Li, Y., & Cheng, G. (2017). Effect of short-term exposure to ambient air particulate matter on incidence of delirium in a surgical population. Scientific Reports (Nature Publisher Group), 7, 1-7. doi: http://dx.doi.org/10.1038/s41598-017-15280-1 Mainka, A., & Zajusz-Zubek, E. (2019). PM1 in Ambient and Indoor Air—Urban and Rural Areas in the Upper Silesian Region, Poland. Atmosphere, 10(11), 662. Malm, W. C., & Hand, J. L. (2013). Comparisons Between Continous and Integrated Mass Measurements USA: Cooperative Institute for Reasearch in the Atmosphere 109 Mani, S. (2019). Bus Company Schedules Mansha, M., Ghauri, B., Rahman, S., & Amman, A. (2012). Characterization and source apportionment of ambient air particulate matter (PM2.5) in Karachi. Science of The Total Environment, 425, 176-183. doi: https://doi.org/10.1016/j.scitotenv.2011.10.056 Marcazzan, G. M., Vaccaro, S., Valli, G., & Vecchi, R. (2001). Characterisation of PM10 and PM2.5 particulate matter in the ambient air of Milan (Italy). Atmospheric Environment, 35(27), 4639-4650. doi: https://doi.org/10.1016/S1352-2310(01)00124-8 Martuzevicius, D., Grinshpun, S. A., Lee, T., Hu, S., Biswas, P., Reponen, T., & LeMasters, G. (2008). Traffic-related PM2.5 aerosol in residential houses located near major highways: Indoor versus outdoor concentrations. Atmospheric Environment, 42(27), 6575-6585. doi: https://doi.org/10.1016/j.atmosenv.2008.05.009 Martuzevicius, D., Grinshpun, S. A., Reponen, T., Górny, R. L., Shukla, R., Lockey, J., Hu, S., McDonald, R., Biswas, P., Kliucininkas, L., & LeMasters, G. (2004). Spatial and temporal variations of PM2.5 concentration and composition throughout an urban area with high freeway density—the Greater Cincinnati study. Atmospheric Environment, 38(8), 1091-1105. doi: https://doi.org/10.1016/j.atmosenv.2003.11.015 Masri, S., Kang, C.-M., & Koutrakis, P. (2015). Composition and sources of fine and coarse particles collected during 2002-2010 in Boston, MA. Journal of the Air & Waste Management Association (1995), 65(3), 287-297. doi: 10.1080/10962247.2014.982307 Massey, D., Masih, J., Kulshrestha, A., Habil, M., & Taneja, A. (2009). Indoor/outdoor relationship of fine particles less than 2.5μm (PM2.5) in residential homes locations in central Indian region. Building and Environment, 44(10), 2037- 2045. doi: https://doi.org/10.1016/j.buildenv.2009.02.010 Mataki, M., Koshy, K. C., & Lal, M. (2006). Baseline Climatology of Viti Levu (Fiji) and Current Climatic Trends. Pacific Science, 60(1), 49-68, 20. 110 Mathew, A. V., Yu, J., Guo, Y., Byun, J., Chen, Y. E., Wang, L., Liu, M., Bard, R. L., Morishita, M., Huang, W., Li, J., Harkema, J. R., Rajagopalan, S., Pennathur, S., & Brook, R. D. (2018). Effect of Ambient Fine Particulate Matter Air Pollution and Colder Outdoor Temperatures on High-Density Lipoprotein Function. American Journal of Cardiology, 122(4), 565-570. doi: 10.1016/j.amjcard.2018.04.061 Mazidi, M., & Speakman, J. R. (2017). Ambient particulate air pollution (PM2.5) is associated with the ratio of type 2 diabetes to obesity. Scientific Reports (Nature Publisher Group), 7, 1-8. doi: http://dx.doi.org/10.1038/s41598-017- 08287-1 Ministry of Education, H. a. A. (2019). Education at a Glance - 2018. Retrieved 25/05/2019, 2019, from http://www.education.gov.fj/ Ministry of Health and Medical Services. (2017). Asthma. Retrieved 13/09/19, from https://www.health.gov.fj/?page_id=656 Ministry of Transport. (2007). Land Transport Rule: Vehcile Exhaust Emissions 2007 (Vol. 33001/2). Ministry of Transport. (2019). Vehicle Ownership. Retrieved 19/09/19, from https://www.mot.gov.sg/About-MOT/Land-Transport/Motoring/Vehicle- Ownership/ Morgan, J. (2015). Country in Focus: turning the tide of diabetes in Fiji. The Lancet Diabetes & Endocrinology, 3(1), 15-16. doi: https://doi.org/10.1016/S2213- 8587(14)70240-2 National Aeronautical and Space Administration. (2019). World View. Retrieved 25/06/2019, 2019, from https://worldview.earthdata.nasa.gov/?p=geographic&l=VIIRS_SNPP_Corre ctedReflectance_TrueColor(hidden),MODIS_Aqua_CorrectedReflectance_Tr ueColor,MODIS_Terra_CorrectedReflectance_TrueColor(hidden),Reference _Labels,Reference_Features,Coastlines(hidden)&t=2017-04-09- 111 T00%3A00%3A00Z&z=3&v=176.2753260597992,- 18.728742086060006,179.6503260597992,-17.122540914185006 National Environment Agency. (2019). Air Pollution. Retrieved 19/09/19, from https://www.nea.gov.sg/our-services/pollution-control/air-pollution/air- pollution-regulations New South Wales Government. (2019). Air Quality Data. Retrieved 23/05/2019, 2019, from https://www.environment.nsw.gov.au/aqms/hourlydata.htm New Zealand Government. (2017). Engine Fuel Regulations Specifcations 2011. Retrieved 19/09/19, from http://www.legislation.govt.nz/regulation/public/2011/0352/latest/whole.html #DLM4044783 NZ Transportation Agency. (2019). Step 1: Standards. Retrieved 19/09/19, from https://www.nzta.govt.nz/vehicles/importing-a-vehicle/step-1-standards/ OECD. (2016). The Economic Consequences of Outdoor Air Pollution. from https://www.oecd-ilibrary.org/content/publication/9789264257474-en Onat, B., Sahin, U. A., & Akyuz, T. (2013). Elemental characterization of PM2.5 and PM1 in dense traffic area in Istanbul, Turkey. Atmospheric Pollution Research, 4(1), 101-105. doi: https://doi.org/10.5094/APR.2013.010 Ott, D. K., Cyrs, W., & Peters, T. M. (2008). Passive measurement of coarse particulate matter, PM10–2.5. Journal of Aerosol Science, 39(2), 156-167. doi: https://doi.org/10.1016/j.jaerosci.2007.11.002 Ouyang, W., Guo, B., Cai, G., Li, Q., Han, S., Liu, B., & Liu, X. (2015). The washing effect of precipitation on particulate matter and the pollution dynamics of rainwater in downtown Beijing. Science of The Total Environment, 505, 306- 314. doi: https://doi.org/10.1016/j.scitotenv.2014.09.062 Pacifc Community. (2017). Energy Efficiency in Land Transport in Pacific Islands Pacific Energiser. Fiji: Pacific Community. 112 Pacific Climate Change Science. (2011). Current and Future Climate of the Fiji Islands International Climate Change Adaptation Initiative Fiji. Pacific Climate Change Science. (2014). Climate Variability, Extremes and Change in the Western Tropical Pacific: New Science and Updated Country Reports 2014. Australia: Pacific Climate Change Science. Pall Corporation. (2018). Pallflex Air Monitoring Filters. from https://shop.pall.com/us/en/laboratory/environmental-qc/airborne- contaminants-analysis/pallflex-air-monitoring-filters-zidgri78max Parliment of the Republic of Fiji. (1998). The Land Transport Act. from http://www.paclii.org/fj/legis/num_act/lta1998180.rtf Patel, M. M., Chillrud, S. N., Correa, J. C., Hazi, Y., Feinberg, M., KC, D., Prakash, S., Ross, J. M., Levy, D., & Kinney, P. L. (2010). Traffic-Related Particulate Matter and Acute Respiratory Symptoms among New York City Area Adolescents. Environmental health perspectives, 118(9), 1338-1343. doi: doi:10.1289/ehp.0901499 Pearson, J. F., Bachireddy, C., Shyamprasad, S., Goldfine, A. B., & Brownstein, J. S. (2010). Association Between Fine Particulate Matter and Diabetes Prevalence in the U.S. Diabetes Care, 33(10), 2196-2201. doi: 10.2337/dc10-0698 Perez, P., & Gramsch, E. (2016). Forecasting hourly PM2.5 in Santiago de Chile with emphasis on night episodes. Atmospheric Environment, 124, 22-27. doi: https://doi.org/10.1016/j.atmosenv.2015.11.016 Perrino, C., Canepari, S., & Catrambone, M. (2013). Comparing the Performance of Teflon and Quartz Membrane Filters Collecting Atmospheric PM: Influence of Atmospheric Water. Aerosol and Air Quality Research, 13, 137-147. doi: 10.4209/aaqr.2012.07.0167 Petzold, A., Ogren, J. A., Fiebig, M., Laj, P., Li, S. M., Baltensperger, U., Holzer- Popp, T., Kinne, S., Pappalardo, G., Sugimoto, N., Wehrli, C., Wiedensohler, A., & Zhang, X. Y. (2013). Recommendations for reporting "black carbon" 113 measurements. Atmos. Chem. Phys., 13(16), 8365-8379. doi: 10.5194/acp-13- 8365-2013 Pey, J., Querol, X., Alastuey, A., Rodríguez, S., Putaud, J. P., & Van Dingenen, R. (2009). Source apportionment of urban fine and ultra-fine particle number concentration in a Western Mediterranean city. Atmospheric Environment, 43(29), 4407-4415. doi: https://doi.org/10.1016/j.atmosenv.2009.05.024 Pirozzi, C. S., Jones, B. E., VanDerslice, J. A., Zhang, Y., Paine, R., III, & Dean, N. C. (2018). Short-Term Air Pollution and Incident Pneumonia: A Case– Crossover Study. Annals of the American Thoracic Society, 15(4), 449-459. doi: http://dx.doi.org/10.1513/AnnalsATS.201706-495OC Polidori, A., & Fine, P. (2012). Ambient Concetrations of Criteria and Air Toxic Pollutants in Close Proximity To a Freeway with Heavy-Duty Diesel Traffic. USA: South Coast Air Quality Managment District. Pratibha, J. (2017, 23/04/2017). LTA Place Measures to Cut Visible Emissions, Fiji Sun. Retrieved from https://fijisun.com.fj/2017/04/23/lta-place-measures-to- cut-visible-emissions/ Pun, V. C., Manjourides, J., & Suh, H. (2017). Association of Ambient Air Pollution with Depressive and Anxiety Symptoms in Older Adults: Results from the NSHAP Study. Environmental Health Perspectives (Online), 125(3), 342. Qin, Y., Kim, E., & Hopke, P. K. (2006). The concentrations and sources of PM2.5 in metropolitan New York City. Atmospheric Environment, 40, 312-332. doi: https://doi.org/10.1016/j.atmosenv.2006.02.025 Rastogi, N., & Sarin, M. M. (2009). Quantitative chemical composition and characteristics of aerosols over western India: One-year record of temporal variability. Atmospheric Environment, 43(22), 3481-3488. doi: https://doi.org/10.1016/j.atmosenv.2009.04.030 Rattigan, O. V., Dirk Felton, H., Bae, M.-S., Schwab, J. J., & Demerjian, K. L. (2010). Multi-year hourly PM2.5 carbon measurements in New York: Diurnal, day of 114 week and seasonal patterns. Atmospheric Environment, 44(16), 2043-2053. doi: https://doi.org/10.1016/j.atmosenv.2010.01.019 Ren, C., O'Neill, M. S., Park, S. K., Sparrow, D., Vokonas, P., & Schwartz, J. (2011). Ambient temperature, air pollution, and heart rate variability in an aging population. American journal of epidemiology, 173(9), 1013-1021. doi: 10.1093/aje/kwq477 Roy, R., & Braathen, N. A. (2017). The Rising Cost of Ambient Air Pollution thus far in the 21st Century. doi: doi:https://doi.org/10.1787/d1b2b844-en Royal Meterological Society. (2017). Mountain Weather Retrieved 25/06/2019, 2019, from https://www.metlink.org/other-weather/weather-hazards/mountain- weather/ Ruuskanen, J., Tuch, T., Ten Brink, H., Peters, A., Khlystov, A., Mirme, A., Kos, G. P. A., Brunekreef, B., Wichmann, H. E., Buzorius, G., Vallius, M., Kreyling, W. G., & Pekkanen, J. (2001). Concentrations of ultrafine, fine and PM2.5 particles in three European cities. Atmospheric Environment, 35(21), 3729- 3738. doi: https://doi.org/10.1016/S1352-2310(00)00373-3 Salvi, D., Limaye, S., Muralidharan, V., Londhe, J., Madas, S., Rao, A., Juvekar, S., Biswal, S., & Salvi, S. (2014). Burning of mosquito coils (MCs) produce high levels of indoor PM2.5 and CO, which are associated with adverse respiratory health effects, and are significantly altered by indoor ventilation patterns. European Respiratory Journal, 44(Suppl 58), P4957. Sanerivi, S. (2017). "Corupt Practises" Alleged in Samoa's Second Hand Car Industry Pacific Islands Report. Hawai'i: East West Wing Centre. Sarnat, J. A., Moise, T., Shpund, J., Liu, Y., Pachon, J. E., Qasrawi, R., Abdeen, Z., Brenner, S., Nassar, K., Saleh, R., & Schauer, J. J. (2010). Assessing the spatial and temporal variability of fine particulate matter components in Israeli, Jordanian, and Palestinian cities. Atmospheric Environment, 44(20), 2383- 2392. doi: https://doi.org/10.1016/j.atmosenv.2010.04.007 115 Secretariat of the Pacific Regional Environment Programme. (2014). Fiji's State of Environment Report 2013. Samoa: Secretariat of the Pacific Regional Environment Programme. Sevura, J. (2005). Promotion of Environmentally Sustainable Transportation in the Pacific Islands: Fiji's Land Transport Sector from an Energy Perspective. Fiji: South Pacific Applied Geoscience Commission. Sezer Turalıoğlu, F., Nuhoğlu, A., & Bayraktar, H. (2005). Impacts of some meteorological parameters on SO2 and TSP concentrations in Erzurum, Turkey. Chemosphere, 59(11), 1633-1642. doi: https://doi.org/10.1016/j.chemosphere.2005.02.003 Shakya, K. M., Rupakheti, M., Shahi, A., Maskey, R., Pradhan, B., Panday, A., Puppala, S. P., Lawrence, M., & Peltier, R. E. (2017). Near-road sampling of PM2. 5, BC, and fine-particle chemical components in Kathmandu Valley, Nepal. Atmos. Chem. Phys., 17(10), 6503-6516. doi: 10.5194/acp-17-6503- 2017 Sharma, M., Agarwal, A. K., & Bharathi, K. V. L. (2005). Characterization of exhaust particulates from diesel engine. Atmospheric Environment, 39(17), 3023-3028. doi: https://doi.org/10.1016/j.atmosenv.2004.12.047 Sharma, M., & Maloo, S. (2005). Assessment of ambient air PM10 and PM2.5 and characterization of PM10 in the city of Kanpur, India. Atmospheric Environment, 39(33), 6015-6026. doi: https://doi.org/10.1016/j.atmosenv.2005.04.041 Singh, P. (2019). Assesment of Bus Count and Number of Daily Trip (S. Mani, Trans.). In S. Mani (Ed.). Song, S., Wu, Y., Zheng, X., Wang, Z., Yang, L., Li, J., & Hao, J. (2014). Chemical characterization of roadside PM2.5 and black carbon in Macao during a summer campaign. Atmospheric Pollution Research, 5(3), 381-387. doi: https://doi.org/10.5094/APR.2014.044 116 Squizzato, S., Masiol, M., Rich, D. Q., & Hopke, P. K. (2018). A long-term source apportionment of PM2.5 in New York State during 2005–2016. Atmospheric Environment, 192, 35-47. doi: https://doi.org/10.1016/j.atmosenv.2018.08.044 Srimuruganandam, B., & Shiva Nagendra, S. M. (2012). Application of positive matrix factorization in characterization of PM10 and PM2.5 emission sources at urban roadside. Chemosphere, 88(1), 120-130. doi: https://doi.org/10.1016/j.chemosphere.2012.02.083 Stein, A. F., Draxler, R. R., Rolph, G. D., Stunder, B. J. B., Cohen, M. D., & Ngan, F. (2015). NOAA’s HYSPLIT Atmospheric Transport and Dispersion Modeling System. Bulletin of the American Meteorological Society, 96(12), 2059-2077. doi: 10.1175/bams-d-14-00110.1 Steinle, S., Reis, S., Sabel, C. E., Semple, S., Twigg, M. M., Braban, C. F., Leeson, S. R., Heal, M. R., Harrison, D., Lin, C., & Wu, H. (2015). Personal exposure monitoring of PM2.5 in indoor and outdoor microenvironments. Science of The Total Environment, 508, 383-394. doi: https://doi.org/10.1016/j.scitotenv.2014.12.003 Stockham, A. J., Schultz, D. M., Fairman, J. G., & Draude, A. P. (2017). Quantifying the Rain-Shadow Effect: Results from the Peak District, British Isles. Bulletin of the American Meteorological Society, 99(4), 777-790. doi: 10.1175/BAMS- D-17-0256.1 Sulong, N. A., Latif, M. T., Khan, M. F., Amil, N., Ashfold, M. J., Wahab, M. I. A., Chan, K. M., & Sahani, M. (2017). Source apportionment and health risk assessment among specific age groups during haze and non-haze episodes in Kuala Lumpur, Malaysia. Science of The Total Environment, 601-602, 556- 570. doi: https://doi.org/10.1016/j.scitotenv.2017.05.153 Tai, A. P. K., Mickley, L. J., & Jacob, D. J. (2010). Correlations between fine particulate matter (PM2.5) and meteorological variables in the United States: Implications for the sensitivity of PM2.5 to climate change. Atmospheric Environment, 44(32), 3976-3984. doi: https://doi.org/10.1016/j.atmosenv.2010.06.060 117 Taiwo, A. M. (2016). Source apportionment of urban background particulate matter in Birmingham, United Kingdom using a mass closure model. Aerosol Air Qual. Res, 16, 1244-1252. Talbi, A., Kerchich, Y., Kerbachi, R., & Boughedaoui, M. (2018). Assessment of annual air pollution levels with PM1, PM2.5, PM10 and associated heavy metals in Algiers, Algeria. Environmental Pollution, 232, 252-263. doi: https://doi.org/10.1016/j.envpol.2017.09.041 Tallon, L. A., Manjourides, J., Pun, V. C., Salhi, C., & Suh, H. (2017). Cognitive impacts of ambient air pollution in the National Social Health and Aging Project (NSHAP) cohort. Environment International, 104, 102-109. doi: https://doi.org/10.1016/j.envint.2017.03.019 Tao, J., Zhang, L., Engling, G., Zhang, R., Yang, Y., Cao, J., Zhu, C., Wang, Q., & Luo, L. (2013). Chemical composition of PM2.5 in an urban environment in Chengdu, China: Importance of springtime dust storms and biomass burning. Atmospheric Research, 122, 270-283. doi: https://doi.org/10.1016/j.atmosres.2012.11.004 Targino, A. C., Gibson, M. D., Krecl, P., Rodrigues, M. V. C., dos Santos, M. M., & de Paula Corrêa, M. (2016). Hotspots of black carbon and PM2.5 in an urban area and relationships to traffic characteristics. Environmental Pollution, 218, 475-486. doi: https://doi.org/10.1016/j.envpol.2016.07.027 Tasić, V., Jovašević-Stojanović, M., Vardoulakis, S., Milošević, N., Kovačević, R., & Petrović, J. (2012). Comparative assessment of a real-time particle monitor against the reference gravimetric method for PM10 and PM2.5 in indoor air. Atmospheric Environment, 54, 358-364. doi: https://doi.org/10.1016/j.atmosenv.2012.02.030 Tecer, L. H., Alagha, O., Karaca, F., Tuncel, G., & Eldes, N. (2008a). Particulate Matter (PM2.5, PM10-2.5, and PM10) and Children's Hospital Admissions for Asthma and Respiratory Diseases: A Bidirectional Case-Crossover Study. Journal of Toxicology and Environmental Health, Part A, 71(8), 512-520. doi: 10.1080/15287390801907459 118 Tecer, L. H., Süren, P., Alagha, O., Karaca, F., & Tuncel, G. (2008b). Effect of Meteorological Parameters on Fine and Coarse Particulate Matter Mass Concentration in a Coal-Mining Area in Zonguldak, Turkey. Journal of the Air & Waste Management Association, 58(4), 543-552. doi: 10.3155/1047- 3289.58.4.543 The Fijian Government. (2010). Vehicle Duty Drops. Retrieved 04/05/2018, from http://www.fiji.gov.fj/Media-Center/Press-Releases/Vehicle-duty-drops.aspx The Global Asthma Network. (2018). The Global Asthma Report 2018. Retrieved 08/08/2019, from http://www.globalasthmareport.org/resources/global_asthma_report_2018.pdf The Government of Fiji. (2011). National Solid Waste Management Strategy. In U. D. Ministry of Local Government, Housing and Environment (Ed.). Fiji: Ministry of Local Government, Urban Development, Housing and Environment. The Government of Fiji. (2012). Environment Stakeholders Meet to discuss Air Pollution. Fiji: The Government of Fiji. The Government of Fiji. (2018). 2019 School Terms. Retrieved 25/05/2019, 2019, from https://www.fiji.gov.fj/About-Fiji/School-Holidays.aspx The World Bank. (2018). Overview: The World Bank in Pacific Islands. Retrieved 23/11/2018, 2018, from https://www.worldbank.org/en/country/pacificislands/overview#3 Thermo Fisher Scientific. (2008). Mie pDR 1500: Instruction Manual. In Thermo Fisher Scientific (Ed.). USA: Thermo Fisher Scientific. Thomaidis, N. S., Bakeas, E. B., & Siskos, P. A. (2003). Characterization of lead, cadmium, arsenic and nickel in PM2.5 particles in the Athens atmosphere, Greece. Chemosphere, 52(6), 959-966. doi: https://doi.org/10.1016/S0045- 6535(03)00295-9 Thorpe, A., & Harrison, R. M. (2008). Sources and properties of non-exhaust particulate matter from road traffic: A review. Science of The Total 119 Environment, 400(1), 270-282. doi: https://doi.org/10.1016/j.scitotenv.2008.06.007 Tong, L. (2016). Samoa to stop importing cars more than 8 years old. Pacifc Islands Report. Retrieved 19/09/19, from http://www.pireport.org/comment/26352#comment-26352 topographic-map.com. (2019). Viti Levu. Retrieved 24/06/2019, 2019, from https://en-gb.topographic-map.com/maps/d52t/Viti-Levu/ Transport for London. (2013). How many cars are there in London and who owns them? In T. f. London (Ed.), Road Task Force - Technical Notes London: Transport for London. Tsapakis, M., Lagoudaki, E., Stephanou, E. G., Kavouras, I. G., Koutrakis, P., Oyola, P., & von Baer, D. (2002). The composition and sources of PM2.5 organic aerosol in two urban areas of Chile. Atmospheric Environment, 36(23), 3851- 3863. doi: https://doi.org/10.1016/S1352-2310(02)00269-8 Tucker, W. G. (2000). An overview of PM2.5 sources and control strategies. Fuel Processing Technology, 65-66, 379-392. doi: https://doi.org/10.1016/S0378- 3820(99)00105-8 United Nations. (2018). World Economic Situation and Prospects 2018. Retrieved 23/07/2019, from https://www.un.org/development/desa/dpad/wp- content/uploads/sites/45/publication/WESP2018_Full_Web.pdf United Nations Environment Programme. (2015). Air Quality Policies - Republic of Fiji. Retrieved 24/07/2019, from https://wedocs.unep.org/bitstream/handle/20.500.11822/17194/Fiji.pdf?seque nce=1&isAllowed=y US Environmental Protection Agency. (2004). The Particle pollution report: current understanding of air quality and emissions through. (E. Office of Air Quality Planning and Standards, Monitoring, and Analysis Division, Trans.): United States Environmental Protection Agency. 120 US Environmental Protection Agency. (2007). Framework for Metal Risk Assesment United States of America: US Environmental Protection Agency. US Environmental Protection Agency. (2016). Quality Assurance Guidance Document 2.12 - Monitoring PM2.5 in Ambient Air Using Designated Reference or Class 1 Equivalent Methods: US Environmental Protection Agency. US Environmental Protection Agency. (2017a). Quality Assurance Handbook for Air Pollution Measurment Systems Vol 2 Analytical Methods (pp. 3). United States: US Environmental Protection Agency. US Environmental Protection Agency. (2017b). US CFR 40 Appendix L to Part 50: Reference Method for the Determination of FIne Particulate Matter as PM2.5 in the Atmosphere Part 50 - National Primary and Secondary Ambient Air Quality Standards (Vol. 2). United States: Office of the Federal Register. Viidanoja, J., Sillanpää, M., Laakia, J., Kerminen, V.-M., Hillamo, R., Aarnio, P., & Koskentalo, T. (2002). Organic and black carbon in PM2.5 and PM10: 1 year of data from an urban site in Helsinki, Finland. Atmospheric Environment, 36(19), 3183-3193. doi: https://doi.org/10.1016/S1352-2310(02)00205-4 Vilcassim, M. J. R., Thurston, G. D., Peltier, R. E., & Gordon, T. (2014). Black carbon and particulate matter (PM2.5) concentrations in New York City's subway stations. Environmental science & technology, 48(24), 14738-14745. doi: 10.1021/es504295h Vojtisek-Lom, M., Pechout, M., Dittrich, L., Beránek, V., Kotek, M., Schwarz, J., Vodička, P., Milcová, A., Rossnerová, A., Ambrož, A., & Topinka, J. (2015). Polycyclic aromatic hydrocarbons (PAH) and their genotoxicity in exhaust emissions from a diesel engine during extended low-load operation on diesel and biodiesel fuels. Atmospheric Environment, 109, 9-18. doi: https://doi.org/10.1016/j.atmosenv.2015.02.077 Wadlow, I., Paton-Walsh, C., Forehead, H., Perez, P., Amirghasemi, M., Guérette, É.- A., Gendek, O., & Kumar, P. (2019). Understanding Spatial Variability of Air Quality in Sydney: Part 2—A Roadside Case Study. Atmosphere, 10(4), 217. 121 Wang, J., Hu, Z., Chen, Y., Chen, Z., & Xu, S. (2013). Contamination characteristics and possible sources of PM10 and PM2.5 in different functional areas of Shanghai, China. Atmospheric Environment, 68, 221-229. doi: https://doi.org/10.1016/j.atmosenv.2012.10.070 Wang, J., & Ogawa, S. (2015). Effects of Meteorological Conditions on PM2.5 Concentrations in Nagasaki, Japan. International journal of environmental research and public health, 12(8), 9089-9101. doi: 10.3390/ijerph120809089 Wang, Y., Zhuang, G., Zhang, X., Huang, K., Xu, C., Tang, A., Chen, J., & An, Z. (2006). The ion chemistry, seasonal cycle, and sources of PM2.5 and TSP aerosol in Shanghai. Atmospheric Environment, 40(16), 2935-2952. doi: https://doi.org/10.1016/j.atmosenv.2005.12.051 Watson, J. G., Chow, J. C., Moosmuller, H., Frank, N., & Pitchford, M. (1998). Guidance for Using Continous Monitors in PM2.5 Monitoring Networks. USA: US Environmental Protection Agency. Williams, R. D., & Knibbs, L. D. (2016). Daily personal exposure to black carbon: A pilot study. Atmospheric Environment, 132, 296-299. doi: https://doi.org/10.1016/j.atmosenv.2016.03.023 Wilton, E., Davy, P., & Smith, J. (2007). Source Identification and Apportionment of PM10 and PM2.5 in Hastings and Auckland. New Zealand: National Institute for Water and Atmospheric Research. Winckelmans, E., Vrijens, K., Tsamou, M., Janssen, B. G., Saenen, N. D., Roels, H. A., Kleinjans, J., Lefebvre, W., Vanpoucke, C., de Kok, T. M., & Nawrot, T. S. (2017). Newborn sex-specific transcriptome signatures and gestational exposure to fine particles: findings from the ENVIRONAGE birth cohort. Environmental Health, 16. doi: http://dx.doi.org/10.1186/s12940-017-0264-y World Bank. (2016). The Cost of Air Pollution; Strengthening the Economic Case for Action. Washington, DC: Institue for Health Metrics and Evaluation. World Health Organisation. (2006). Air Quality Guidelines: Global Update 2005. 122 World Health Organisation. (2010). Pacific Islander Pay Heavy price for Abandoning Traditional Diet. Bulletin of the World Health Organsiation, 88(7). World Health Organisation. (2018). World Health Statistic 2018: Monitoring Health for the SDG's, sustainable development goals. Geneva. Wu, C.-F., Woodward, A., Li, Y.-R., Kan, H., Balasubramanian, R., Latif, M. T., Sahani, M., Cheng, T.-J., Chio, C.-P., Taneepanichskul, N., Kim, H., Chan, C.- C., Yi, S.-M., Withers, M., & Samet, J. (2017). Regulation of fine particulate matter (PM2.5) in the Pacific Rim: perspectives from the APRU Global Health Program. Air Quality, Atmosphere & Health, 10(9), 1039-1049. doi: 10.1007/s11869-017-0492-x Wu, Y.-S., Fang, G.-C., Lee, W.-J., Lee, J.-F., Chang, C.-C., & Lee, C.-Z. (2007). A review of atmospheric fine particulate matter and its associated trace metal pollutants in Asian countries during the period 1995–2005. Journal of Hazardous Materials, 143(1), 511-515. doi: https://doi.org/10.1016/j.jhazmat.2006.09.066 Xia, L., & Gao, Y. (2011). Characterization of trace elements in PM2.5 aerosols in the vicinity of highways in northeast New Jersey in the U.S. east coast. Atmospheric Pollution Research, 2(1), 34-44. doi: https://doi.org/10.5094/APR.2011.005 Xing, Y.-F., Xu, Y.-H., Shi, M.-H., & Lian, Y.-X. (2016). The impact of PM2.5 on the human respiratory system. Journal of thoracic disease, 8(1), E69-E74. doi: 10.3978/j.issn.2072-1439.2016.01.19 Xu, X., Nie, S., Ding, H., & Hou, F. F. (2018). Environmental pollution and kidney diseases. Nature Reviews Nephrology, 14, 313. doi: 10.1038/nrneph.2018.11 Yim, S. H. L., Fung, J., Lau, A., & Kot, S. C. (2009). Air ventilation impacts of the “wall effect” resulting from the alignment of high-rise buildings. Atmospheric Environment, 43, 4982–4994. doi: 10.1016/j.atmosenv.2009.07.002 Younan, D., Tuvblad, C., Franklin, M., Lurmann, F., Li, L., Wu, J., Kiros, B., Baker, L. A., & Chen, J.-C. (2018). Longitudinal Analysis of Particulate Air Pollutants 123 and Adolescent Delinquent Behavior in Southern California. Journal of Abnormal Child Psychology, 46(6), 1283-1293. doi: http://dx.doi.org/10.1007/s10802-017-0367-5 Younger, M., Morrow-Almeida, H. R., Vindigni, S. M., & Dannenberg, A. L. (2008). The Built Environment, Climate Change, and Health: Opportunities for Co- Benefits. American Journal of Preventive Medicine, 35(5), 517-526. doi: https://doi.org/10.1016/j.amepre.2008.08.017 Zhang, H., Xu, T., Zong, Y., Tang, H., Liu, X., & Wang, Y. (2015). Influence of Meteorological Conditions on Pollutant Dispersion in Street Canyon. Procedia Engineering, 121, 899-905. doi: https://doi.org/10.1016/j.proeng.2015.09.047 Zhang, J., Liu, J., Ren, L., Wei, J., Duan, J., Zhang, L., Zhou, X., & Sun, Z. (2018a). PM2.5 induces male reproductive toxicity via mitochondrial dysfunction, DNA damage and RIPK1 mediated apoptotic signaling pathway. Science of The Total Environment, 634, 1435-1444. doi: https://doi.org/10.1016/j.scitotenv.2018.03.383 Zhang, J., P. Marto, J., & Schwab, J. (2018b). Exploring the Applicability and Limitations of Selected Optical Scattering Instruments for PM Mass Measurement. Zhang, K., & Batterman, S. (2013). Air pollution and health risks due to vehicle traffic. The Science of the total environment, 450-451, 307-316. doi: 10.1016/j.scitotenv.2013.01.074 Zhang, S., Breitner, S., Cascio, W. E., Devlin, R. B., Neas, L. M., Diaz-Sanchez, D., Kraus, W. E., Schwartz, J., Hauser, E. R., Peters, A., & Schneider, A. (2018c). Short-term effects of fine particulate matter and ozone on the cardiac conduction system in patients undergoing cardiac catheterization. Particle and Fibre Toxicology, 15. doi: http://dx.doi.org/10.1186/s12989-018-0275-z Zhang, Z.-H., Khlystov, A., Norford, L. K., Tan, Z.-K., & Balasubramanian, R. (2017). Characterization of traffic-related ambient fine particulate matter (PM2.5) in an Asian city: Environmental and health implications. Atmospheric 124 Environment, 161, 132-143. doi: https://doi.org/10.1016/j.atmosenv.2017.04.040 Zheng, M., Cass, G. R., Schauer, J. J., & Edgerton, E. S. (2002). Source Apportionment of PM2.5 in the Southeastern United States Using Solvent- Extractable Organic Compounds as Tracers. Environmental science & technology, 36(11), 2361-2371. doi: 10.1021/es011275x Zheng, X., Xu, X., Yekeen, T. A., Zhang, Y., Chen, A., Kim, S. S., Dietrich, K. N., Ho, S.-M., Lee, S.-A., & Reponen, T. (2016). Ambient air heavy metals in PM2. 5 and potential human health risk assessment in an informal electronic- waste recycling site of China. Aerosol Air Qual Res, 16(2), 388-397. Zhu, W., Luo, L., Cheng, Z., Yan, N., Lou, S., & Ma, Y. (2018). Characteristics and contributions of biogenic secondary organic aerosol tracers to PM2.5 in Shanghai, China. Atmospheric Pollution Research, 9(2), 179-188. doi: https://doi.org/10.1016/j.apr.2017.09.001 Zhu, Y., Hinds, W. C., Kim, S., Shen, S., & Sioutas, C. (2002). Study of ultrafine particles near a major highway with heavy-duty diesel traffic. Atmospheric Environment, 36(27), 4323-4335. doi: https://doi.org/10.1016/S1352- 2310(02)00354-0 Zuurbier, M., Hoek, G., Oldenwening, M., Lenters, V., Meliefste, K., van den Hazel, P., & Brunekreef, B. (2010). Commuters' exposure to particulate matter air pollution is affected by mode of transport, fuel type, and route. Environmental health perspectives, 118(6), 783-789. doi: 10.1289/ehp.0901622 125 Appendix Appendix A Table A1.1 Spearman’s Rho correlation between the metrological variables and PM2.5 in Suva 126 Table A1.2 Spearman’s Rho correlation between the metrological variables and P 127 128 129 130