University of Southern California Dissertations and Theses

Temporal, spatial and toxicological characteristics of coarse particulate matter in an urban area and relation to sources and regulations

(USC Thesis Other)

Temporal, spatial and toxicological characteristics of coarse particulate matter in an urban area and relation to sources and regulations

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content TEMPORAL, SPA TIAL AND TOXICOLOGICAL CHARACTERI STICS OF COARSE PA RTICULA TE MA TTER IN AN URBAN AREA AND RELATION TO SOURCES AND REGULATIONS by Kalarn Cheung A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL Copyright 2012 UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfilhnent of the Requirements for the Degree DOCTOR OF PHI LOS OPHY (ENVIRONMENTAL ENGINEERING) August 2012 Kalarn Cheung Dedication To my grandparents for their unconditional love and support. 11 Acknowledgements This thesis arose, in part, out of the four years of research that has been conducted since I joined the aerosol lab at the University of Southern California (USC) in 2008. It is my great pleasure to convey my sincere thanks to all the people for their help along with my humble acknowledgement. In the first place, I express deep sense of gratitude to my advisor Prof essor Constantinos Sioutas for his unflinching encouragement and guidance in my doctoral research work. His exceptional scie ntific intuition and passion in aerosol research has nourished my intellectual maturity and inspired my growth as a scientist. I would like to thank him for providing the extraordinary experiences throughout my graduate study. I would also like to thank the members of my guidance committee, Prof essor Ronald C Henry, Prof essor Jiu-chiuan Chen and Prof essor Scott Fruin for their thoughtful suggestions on my research work. My sincere thanks are also due to Prof essor James J. Schauer, who always share insightful comments and contributed in various ways to the development of my disse rtation. Research in the aerosol lab is always a team effort. It has been a great pleasure to share my graduate studies and life with the former and current colleagues at USC: Dr. Katharine Moore, Dr. Andrea Polidori, Dr. Bangwoo Han, Dr. Maria Cruz Minguillon, Dr. Subhasis Biswas, Dr. Mohammad Arhami, Dr. Zhi Ning, Dr. Vishal Verma, Dr. Payam Pakbin, Neelakshi Hudda, Winnie Kam, Nancy Daher, James Liacos, Dongbin Wang and Sina Hasheminassab. Special thanks to Dr. Zhi Ning for his friendship and help in my graduate study. Without their support, involvement and compassions for work, this disse rtation would not have been possible. Lastly, I would like to thank my fa mily for their persistent confidence and tolerance on me. 111 Table of Contents Dedication ........................................................................................................................................ ii Acknowledgements ......................................................................................................................... iii List of Tables ................................................................................................................................ viii List of Figures .................................................................................................................................. x Abstract ......................................................................................................................................... xiv Chapter 1 Int roduction ..................................................................................................................... 1 1.1. Background ...................................................................................................................... 1 1.2. Characteristics of Particulate Matter ................................................................................ 2 1.2.1. Particle size .............................................................................................................. 2 1.2.2. Particle mass ............................................................................................................ 3 1.2.3. Particle number ........................................................................................................ 3 1.2.4. Health effects ........................................................................................................... 4 1.3. Rationale ofthe Present Study ......................................................................................... 5 1.3.1. Motivation and objectives ........................................................................................ 5 1.4. Thesis Overview .............................................................................................................. 9 Chapter 2 Spatial and Temporal Variation of Chemical Composition and Mass Closure of Ambient Coarse Particulate Matter (PM10. 2.5) in the Los Angeles Area ........................................ 11 2.1. Abstract. ......................................................................................................................... 11 2.2. Introduction .................................................................................................................... 12 2.3. Methodology .................................................................................................................. 13 2.3.1. Site description and sampling time ........................................................................ 13 2.3.2. Sampling equipment and setup .............................................................................. 14 2.3.3. Chemical analys is .................................................................................................. 15 2.4. Results and Discussion .................................................................................................. 16 2.4. 1. Meteorology ........................................................................................................... 16 2.4.2. Data overview ........................................................................................................ 18 lV 2.4.3. CPM mass reconstruction methodology ................................................................ 20 2.4.4. CPM mass chemical composition .......................................................................... 24 2.4.5. Coefficients of divergence calculations for chemical species concentrations ....... 34 2.5. Summary and Conclusions ............................................................................................ 36 2.6. Acknowledgements ........................................................................................................ 36 Chapter 3 Seasonal and Spatial Variations of Individual Organic Compounds of Coarse Particulate Matter in the Los Angeles Basin .................................................................................. 37 3.1. Abstract .......................................................................................................................... 37 3. 2. Introduction .................................................................................................................... 3 8 3.3. Methodology .................................................................................................................. 39 3.3.1. Chemical analys is .................................................................................................. 39 3.4. Results and Discussion .................................................................................................. 40 3.4. 1. Overview ................................................................................................................ 40 3.4.2. PAl-Is ...................................................................................................................... 42 3.4.3. Hopanes and steranes ............................................................................................. 44 3.4.4. n-Alkanes ............................................................................................................... 46 3.4.5. n-Alkanoic acids .................................................................................................... 49 3.4.6. Other carboxylic acids ........................................................................................... 53 3.4. 7. Levogluc osan ......................................................................................................... 54 3.5. Summary and Conclusions ............................................................................................ 56 3.6. Acknowledgements ........................................................................................................ 57 Chapter 4 Historical Trends in the Mass and Chemical Species Concentrations of Coarse Particulate Matter (PM) in the Los Angeles Basin and Relation to Sources and Air Quality Regulations .................................................................................................................................... 58 4. 1. Abstract .......................................................................................................................... 58 4.2. Introduction .................................................................................................................... 59 4.3. Methodology .................................................................................................................. 60 4.4. Results and Discussion .................................................................................................. 65 4.4. 1. Trends in CPM mass concentrations ...................................................................... 65 4.4.2. Trends in chemical CPM components ................................................................... 71 4.5. Summary and Conclusions ............................................................................................ 80 4.6. Acknowledgments .......................................................................................................... 81 v Chapter 5 Diurnal Trends in Coarse Particulate Matter Composition in the Los Angeles Basin .. 82 5.1. Abstra ct. ......................................................................................................................... 82 5.2. Introduction .................................................................................................................... 83 5.3. Methodology .................................................................................................................. 84 5.3.1. Sites description ..................................................................................................... 84 5.3.2. Sampling time and setup ........................................................................................ 84 5.3.3. Chemical analys es .................................................................................................. 85 5.4. Results and Discussion .................................................................................................. 86 5.4. 1. Meteorology ........................................................................................................... 86 5.4.2. Coarse PM component model and data overview .................................................. 87 5.4.3. Diurnal profiles ...................................................................................................... 90 5.4.4. Seasonal and spatial correlations ........................................................................... 99 5.4.5. Chloride depletion ................................................................................................ 101 5.5. Summary and Conclusions .......................................................................................... 104 5.6. Acknowledgements ...................................................................................................... 105 Chapter 6 Diurnal Trends in Oxidative Potential of Coarse Particulate Matter in the Los Angeles Basin and Their Relation to Sources and Chemical Composition ................................. 106 6.1. Abstra ct. ....................................................................................................................... 106 6.2. Introduction .................................................................................................................. 106 6.3. Methodology ................................................................................................................ 108 6.3. 1. Chemical analys es ................................................................................................ 108 6.4. Results and Discussion ................................................................................................ 110 6.4. 1. Overview .............................................................................................................. 110 6.4.2. Water solubility of elements ................................................................................ 112 6.4.3. ROS activity ......................................................................................................... 113 6.4.4. Asso ciation between ROS activity and water- soluble elements .......................... 115 6.4.5. Comparisons with other studies ........................................................................... 119 6.5. Summary and Conclusions .......................................................................................... 122 6.6. Acknowledgements ...................................................................................................... 122 Chapter 7 Conclusions ................................................................................................................. 123 7. 1. Characteristics of Coarse Particles ............................................................................... 123 7. 1.1. Mass concentration .............................................................................................. 123 Vl 7.1.2. Chemical composition .......................................................................................... 124 7. 1.3. Health effects ....................................................................................................... 125 7.2. Discussions and Recommendations ............................................................................. 125 7.2.1. Limitations of current investigation ..................................................................... 125 7.2.2. Implications on epidemiological studies .............................................................. 126 7.2.3. Recommendations for future research ................................................................. 127 7.2.4. Recommendations on coarse particle regulation .................................................. 128 Bibliography ................................................................................................................................ 130 Vll List of Tables Table 2-1: Selected meteorological parameters by location and season . ....................................... 17 Table 2-2: Statistical summary of annual concentrations (ng/m 3 ) of selected chemical species and gravimetric CPM mass concentrations .................................................................. 19 Table 2-3 : Annual mass fractions (%) (mean ± standard deviation) of CPM components . ........... 26 Table 2-4: Regre ssion analysis of organic carbon (OC) vs. 4 main coarse PM categories and selected species in: (a) spring and summer and (b) fall and winter . ............................ 29 Table 2-5: Correlation coefficient (R) between organic carbon (OC) and soil dust tracers of Fe and Ti at the 10 sampling sites .. ................................................................................... 30 Table 3-1: Summary table of organic component concentration (ng/m 3 ) at the 10 sampling sites in: (a) summer and (b) winter. BDL denotes below detection limit. .. .......................... 42 Table 3-2: Correlation coefficients (R) among individual organic component, compound class, organic carbon ( OC), elemental carbon (EC), selected elements and meteorological paramet ers. Species with less than half data points with detected levels are excluded in this analys is . ............................................................................. 46 Table 4- 1: Sampling location, sampling time and fr equency, sampling instrument and method, analytical method and other background information of the seven studies analyzed .. ......................................................................................................... 63 Table 4-2: Linear regres sion analysis of PM data (PM10 fr om 1988 to 2009, PM10. 2.5 and PM 2.5 fr om 1999 to 2009) in: (a) downtown Los Angeles, (b) Long Beach and (c) Rive rside. Values in parent heses represent standard errors of the slope and intercept. ...................................................................................................................... 66 Table 5-l: Selected meteorological parameters at the 3 sampling sites in: (a) summer and (b) winter . .......................................................................................................................... 87 V111 Table 5-2: Av erage diurnal concentration (flg/m 3 ) of chemical components at the three sampling sites in: (a) summer and (b) winter (± indicates uncertainties calculated based on the analytical uncertainties and uncertainti es from blank corrections) . ........ 89 Table 5-3: Correlation coefficient between selected species in: (a) summer and (b) winter .. ........ 99 Table 5-4: Correlation coefficient between selected species in: (a) Lancaster (LAN) ; (b) Los Angeles (USC) and (c) River side (RIV) .. ...................................................... 100 Table 6-1: Coefficient of determination (R 2 ) between ROS activity and selected water-soluble (WS) elements .. ................................................................................... 11 6 Table 6-2: Summary of studies that employed the same in-vitro bioassay as our study to examine oxidative potential in ambient particulate matter . ....................................... 120 IX List of Figures Figure 2-1: Map of the 10 sampling sites . ..................................................................................... 14 Figure 2-2: Linear regre ssion of overall reconstructed vs. gravimetric mass concentrati ons. Values in parent heses represent standard errors of the slope and intercept. ............... 23 Figure 2-3: Linear regre ssion of reconstructed vs. gravimetric mass concentrations in: (a) spring; (b) summer; (c) fall and (d) winter. Values in parent heses represent standard errors of the slope and intercept. .................................................................. 23 Figure 2-4: Chemical composition and gravimetric mass concentration by location in: (a) spring; (b) summer; (c) fall and (d) winter ........................................................... 24 Figure 2-5: Crustal enrichment factors of individual elements at Los Angeles, Long Beach, Riverside and Lancaster. The reference element is Al ................................................ 27 Figure 2-6: Ammonium, non-sea salt sulf ate and nitrate concentrations by location in: (a) spring; (b) summer; (c) fall; and (d) winter ........................................................... 32 Figure 2-7: Coefficients of divergence (COD) for all sites (excluding Lancaster) in: (a) spring and summer and (b) fall and winter . .......................................................... 34 Figure 3-1: Concentration (ng/m 3 ) ofPAHs with (a) MW = 228 g/mole and (b) MW = 252 g/mole. The level of detection (LOD) of the compound class is shown as a refer ence line ............................................................................................ 43 Figure 3-2: Concentration (ng/m 3 ) of sum of hopanes. The level of detection (LOD) of the compound class is shown as a reference line . ........................................................... .4 5 Figure 3-3: Concentration (ng/m 3 ) of sum of n-alkanes. The level of detection (LOD) of the compound class is shown as a reference line . ........................................................... .4 8 X Figure 3-4: Concentration distribution (per sum of n-alkanes, C19-C38) of coarse particulate matter for n-alkanes in: (a) summer and (b) winter ................................................... .4 9 Figure 3-5: Concentration (ng/m 3 ) ofn -alkanoic acids, as a sum ofC15-C30. The level of detection (LOD) of the compound class is shown as a reference line ........................ 50 Figure 3-6: Concentration distribution (per sum of n-alkanoic acids, C15-C30) of coarse particulate matter for n-alkanoic acid in: (a) summer and (b) winter . ........................ 52 Figure 3-7: Concentration (nglm 3 ) of: (a) pinonic acid, (b) palmitoleic acid and (c) oleic acid. The level of detection (LOD) of each organic compound is shown as a reference line . ............................................................................................................. 54 Figure 3-8: Concentration (nglm 3 ) of levoglucosan. The level of detection (LOD) of levoglucosan is shown as a refer ence line .................................................................. 56 Figure 4-1: Map of the samplings sites in downtown Los Angeles (DLAand USC), Long Beach (NLB, HUD, LBCC and S3) and Ri verside (RUB, VBR, UCR and RIV). Sites operated and maintained by the SCAQMD are represented in triangles .................... 61 Figure 4-2: Annual concentrations of: (a) PM10 from 1988 to 2009, and (b) CPM I PM2.5 fr om 1999-2009 in downtown Los Angeles ............................................................... 68 Figure 4-3: Annual concentrations of: (a) PM10 from 1988 to 2009, and (b) CPM I PM2 .5 fr om 1999-2009 in Long Beach ................................................................................. 69 Figure 4-4: Annual concentrations of: (a) PM10 from 1988 to 2009, and (b) CPM I PM2 .5 fr om 1999-2009 in River side ..................................................................................... 71 Figure 4-5: CPM concentrations of: (a) mass, organic and elemental carbon and inorganic ions, (b) elements of crustal origins, and (c) elements of anth ropogenic origins in downtown Los Angeles . Error bars show standard errors ofthe average when available ..................................................................................................................... 72 xi Figure 4-6: CPM concentrations of: (a) mass, organic and elemental carbon and inorganic ions, (b) elements of crustal origins, and (c) elements of anthropogenic origins in Long Beach. Error bars show standard errors of the average when available ........ 72 Figure 4-7: CPM concentrations of: (a) mass, organic and elemental carbon and inorganic ions, (b) elements of crustal origins, and (c) elements of anth ropogenic origins in Rive rside. Error bars show standard errors of the average when available ............ 73 Figure 5-1: Map of the 3 sampling sites ....................... 0 ........................ o ....................... o ............... 84 Figure 5-2: Diurnal profiles (normalization to the 24-hr average) of CPM mass .......................... 90 Figure 5-3: Carbon monoxide (CO) levels at: (a) Los Angeles-USC; (b) Lancaster-LAN; (c) Riverside-RJ\1 ............................................................................................................. 92 Figure 5-4: Diurnal profiles of: (a) crustal materials and trace elements; (b) vehicle abrasion; (c) water soluble organic carbon and (d) sea salt. Normalization to the 24-hr average is presented .................................................................................................... 94 Figure 5-5: Diurnal profiles of: (a) ammonium; (b) nitrate and (c) non-sea salt sulfate. Normalization to the 24-hr average is presented ........................................................ 97 Figure 5-6: Chloride depletion and nitrate replacement scatter-plots in a) summer and b) winter. Dark legends refer to the sum of [Cl-] + [N03- ], and gray legends refer to [Cl-]. ............................................................................................................ 101 Figure 5-7: Diurnal profile of excess ammonium ion concentration (nglm3) and molar equivalent ratios of [Cl-] I [Na+] and sum of [Cl-] and [N03-] I [Na+] in: (a) summer and (b) winter . ....................................................................................... 104 Figure 6-1: Diurnal profile of chemical composition in: (a) summer and (b) winter. Error bars represent analytical uncertainties ........ 0 ....................... 0 ....................... 0..... . . . . . . . . 111 Xll Figure 6-2: Water solubility of selected metals and elements, calculated across three sites, four periods and two seasons. 1s t quarti le, median and 3r d quartile is shown. Species with > 25% of data points under detection limit (2 x total uncertainti es) were excluded ........................................................................................................... 11 3 Figure 6-3: Diurnal profile ofROS activity on a: (a) air volume basis and (b) PM mass basis .. ........................................................................................................................ 11 4 Figure 6-4: Correlations between measured ROS activity and water- soluble Cu in: (a) summer and (b) winter . .......................................................................................................... 11 7 Xlll Abstract To advance our understanding on the relationship between the sources, chemical composition and toxicity of coarse part icles, and help the regulatory community to desigu cost-eff ective control strat egies, two comprehensive investigations were conducted in the Los Angeles Basin fr om 2008 to 2010 to characterize the phys ico-chemical and toxicological properties of ambient coarse particulate matter (CPM) . The first study features of a year-long sampling campaigu at 10 sampling sites throughout the basin in an attempt to study the spatial and seasonal characteristics of ambient coarse part icles. An intensive study, focusing on the diurnal trends, was conducted at 3 sampling sites to examine how the change in meteorological conditions throughout a day may affect the source strength and fo rmation mechanisms of coarse mode aerosols. Overall, the group of crustal materials and other trace elements was the most dominant component of CPM, accounting for approximately half of the total reconstructed CPM mass. The contributi ons varied fr om 41% to 61% across the 10 samplings sites, with higher fractions in the inland areas . Organic materi als, accounting for around 20% of the CPM mass, well correlated with crustal materials and the soil dust tracers ofTi and Fe, suggesting that humic substance is a major source of organic materials in coarse part icles. This is confirmed by further investigation on individual organic compound s, which revealed that the predominant organic constituents in the coarse size fraction, including n-alkanoic acids and medium molecular weight (MW) n-alkanes (C25 to C35), were highly assoc iated with crustal materials. Sea salt, on the other hand, accounted for an average of 9% of the reconst ruct CPM mass, with higher levels in spring and summer when the onshore wind prevailed. Nitrate, the most dominant inorganic species in the X!V coarse size fraction, arose predominantly fr om the depletion of sea salt, and thereby mostly followed the temporal and spatial patterns of sea salt aer osols. Constituents with primary anth ropogenic origins, such as elemental carbon, hopanes and stera nes, experienced very low concen trati ons. On the other hand, natural sources such as mineral dust and the associated biota, and to a lesser extent fresh and aged sea salt, constituted to the ma jority of ambient CPM mass in the Los Angeles Basin. In summer, ambient coarse particles were mostly re-suspended by wind, and their levels were generally higher in the midday and afternoon sampling periods, in parallel with the higher wind speeds. In winter, the levels of CPM were generally lower, with the exception of stagnation condition s, when particles accumulated in the atmosphere in episodes of low atmospheric dilution. During winter nighttime, the mixing height was lowest. Turbulences induced by vehicular movements became a dominant re-suspension mechanism of coarse mode aerosols, resulting in high levels of mineral and road dust at near-freeway sites. Overall, the contributions of inorganic species to CPM mass were generally higher overnight, suggesting that the lower temperature and higher relative humidity at nightti me fa vored the fo rmation of these ions in the coarse mode. To evaluate the toxic activity induced by CPM, a cellular assay was used to quantify the generation of reactive oxygen species (ROS). The ROS activity experienced a significant diurnal variation, with higher levels in summer than winter. During summertime, higher ROS activity was observed in the midday I afternoon periods, while the peak activity occurred overnight in winter. Using linear regression analys is, the ROS activity was highly associ ated with the water-soluble fraction of four elements (V, Pd, Cu and Rh), which have primary anth ropogenic origins in the coarse size fraction. Based on the results of this investigation, coarse particles XV generated by anthropogenic activiti es, although low in mass concentra tions, are the key drivers of ROS fo rmation, and therefore more targeted control strategies may be needed to better protect the public health from these toxic CPM sources . XV! Chapter 1 Introduction 1.1. Background Particulate matter (PM) is defined as a suspension of liquid droplets or solid matter in a gas or liquid. PM is an extr emely important component in the ambient atmosphere. It contributes to smog fo rmation and visibility degradation in urban atmo spheres, and influences the surface albedo by decreasing the amount of heat reaching the ground (Seinfeld and Pandis, 2006). Part icles also play significant roles in biogeochemical cycles in the atmosphere, serving as reaction substrates and carriers for sorbed chemical and /or biological species (Dentener et a!., 1996). More importantly, numerous epidemiological studies have shown ass ociations between adverse health effects and ambient PM levels (Dockery et a!., 1993; McConnell et a!., 1999; Sram et a!., 1999; Stayner et a!., 1998). Therefore, understa nding the phys ico-chemical and toxicological characteristics of PM is essential to evaluate its environmental and health consequenc es. At mospheric PM is a complex mixture of many classes of chemical constituents, with size ranging fr om a few nanometers (urn ) to tens of micrometers (f!m) in aerodynamic diameter. Sources of ambient PM can be both natural and anthropogenic. Natural sources include windb ome dust, sea spray, fo rest fir es, and volcanic emissions (Seinfeld and Pandis, 2006). Combustion activities such as burning of fossil fu els and biomasses are the ma jor anthropogenic sources in an urban atmosphere. These particles are also known as primary pollutants since they are emitted directly into the atmosphere. On the other hand, particles can be fo rmed by secondary processes, such as the photo-oxidation of S02 and N02 to form sulfate and nitrate part icles, respec tively. Once emitted to the atmosphere, aerosol propert ies, including size and chemical composition, are sub ject to transportation and transformation by various chemical and physical processes (Hinds, 1999). 1.2. Characteristics of Particulate Matter 1.2.1. Particle size Particle size is the most important parameter in describing particle characteristics including fo rmation and removal mechanisms, atmospheric lifetime, as well as the site of deposition in human respiratory tract. There are three distinct size modes of particles based on an observed particle size distributi on: (a) the coarse mode, particle with an aerodynamic diameter between 2.5 and 10 f!m; (b) the accumulation mode, particle with an aerodynamic diameter between 0.1 and 2.5 f!ID; and (c) the ultrafme mode, particle with an aerodynamic diameter between 0.005 and 0. 1 flill · The three size modes diff er in their means of fo rmation and removal, chemical composition, deposition, and optical properties (Sei nfeld and Pandi s, 2006). Ultraf me particles primarily originate from combustion sources , as well as homogeneous or heter ogeneous nucleation of atmospheric species. Ultrafine particles are removed from the mode predomina ntly through physical process of coagulation with larger part icles. Accumu lation mode particles compose a siguificant amount of PM mass and surface area. These particles are fo rmed through coagulation oful trafine mode part icles, and growing of existing particles from condensation of gases. The accumulation mode particles have inefficient removal mechanisms since they are too small to be settled by gravity and too large to coagulate into larger part icles, resulting in long residence times (Seinfeld and Pandis, 2006). Lastly, coarse mode aerosols are generated primarily by mechanical processes such as grinding, erosion, and wind re-suspension. These particles have relatively high 2 gravitati onal settling velocities, allowing them to settle out of the atmosphere in a relatively short amount of time. Coarse particles are also known to react with atmospheric gases. For examp le, fo rmation of coarse particulate nitrate from reaction of nitric acid with mineral dust and sea salt particles was observed in many studies (Goodman et a!., 2000; Usher et a!., 2003; Zhuang et a!., 1999). Unlike the fo rmation of volatile compounds such as ammonium nitrate in the fine mode, the irrever sible reactions between atmospheric gases and coarse part icles, coupled with the lower residence time of coarse particles decrease the atmospheric lifetime and serve as an important removal pathway of gaseous pollutants . 1.2.2. Particle mass Particle mass is an important parameter in terms of particle measurement. Particle mass concentration is used for regulatory purposes in the United States, and is dominated by the accumulation and coarse mode particles (Hinds, 1999). Ultrafine mode part icles, on the other hand, comprise a relatively small percent of mass concentration in ambient atmospheres due to their small sizes. 1.2.3. Particle number Particle number is another important characteristic that emerged in the field of PM measurement in the last decade. In contrast with particle mass, particle number concentration is dominated by ultrafine particles (UFPs ). UFPs constitute to a majority of ambient particle number concentrations but only a small fraction of particle mass. Primary vehicular emissions and photochemical reactions are the ma jor sources of particle number concentrations in an urban atmosphere (Fine et a!., 2004b ; Kulmala et a!., 2004). Although not currently used for regulation in the United States, particle number emission limit was introduced in the Euro 5/6 standards in 3 recognition of the higher toxicity of ultrafine particles compared with particles in accumulation and coarse mode (Delfino et a!., 2005; Li et a!., 2003). 1.2.4. Health eff ects The impact of ambient PM on human health is the most prominent motivation that drives aerosol researches. Numerous epidemiological studies linked elevated PM levels to various adverse health outcom es, including premature deaths (Hoek et a!., 2002), respiratory and cardiovascular diseases (Gauderrnan et a!., 2007; Pope and Dockery, 2006), and neurodegenerative disorders (Peters et a!., 2006). A few other studies also presented negative impacts of PM exposure on liver functions (Folkrnann et a!., 2007) and reproductive systems (Sram et a!., 1999). Both in-vitro and in-vivo studies demonstrated associ ations between particulate exposure to respiratory inflammation, mitochondrial damage, and lung cancer (Castranova et a!., 2001; Costa and Dreher, 1997; Klei rnn an et a!., 2007; Li et a!., 2003). Whi le some recent studies presented higher toxicity of ultrafine particles relative to the larger particles in the accumulation and coarse mode in urban atmospheres (Hu et a!., 2008; Li et a!., 2003), a few epidemiological studies have shown ass ociations between coarse particles with adverse health outcomes in areas where higher fr actions of coarse PM are fo und (Lipsett et a!., 2006; Smith et a!., 2000). Particle size, governing the site of particle deposition along the respiratory tract, has an important role in health effects induced by PM exposure. Particles that are larger than 10 f!m are usually not the point of interest in health studies due to their short atmospheric lifet imes . CPM ranging fr om 2.5 to 10 f1ID is generally referr ed as "inhalable coarse particles", and primarily deposits on the upper respiratory tract including the nasal airway and phary ngeaVlaryngeal 4 regions . A small fraction of coarse particles might also penetrate and deposit further in the respiratory tract. Particles in accumulation mode have low removal efficiency on the nasal airwa y, and generally penetrate deeper to the tracheobronchial tree and puhnonary regions . Ultrafine partic les, on the other hand, have higher deposition efficiency in the nasal and alveolar regions . Some recent studies have also shown the presence of ultrafine particles in the brain and central nervous system (Oberdorster et a!., 2004). In spite of recent advancements in PM toxicity research, the relationships between aerosol properties (e.g., size, number, volatili ty, surface area, and composition) and specific health end points are not well unders tood and remain an area of active research. The production of reactive oxygen species (ROS) and the consequent generation of oxidative stress is proposed to play a direct role in PM-induced adverse health outcomes (Castro and Freeman, 2001; Tao et a!., 2003). Oxidative stress refers to an imbalance within cells where more ROS is produced than eliminated, and high levels of oxidative stress can result in damage within a cell that can harm the cell's functions and lead to apoptosis (Pelicano et a!., 2004). Both chemical and biological assays are available to quantitatively characterize the production of ROS, thereby allowing the examination ofROS activity with aerosols of various sizes and compositions . 1.3. Rationale of the Present Study 1.3.1. Motivation and ob jectives Ambient PM is regulated under the Clean Air Act using size specific mass concentration standa rds. The National Ambient Air Quality Standards (NAAQS) for ambient PM were first established in 1971, when the standard of total suspended particle (TSP) was introduced. They have evolved over the years as a result of emerging epidemiological and toxicological studies that 5 linked elevated airborne PM mass concentration of different size fractions to a variety of adverse health outcomes as discussed in Section 1.2 .4. In 1987, the PM 10 (part icles smaller than 10 fUll in diameter) standard was used to replace the TSP (Total Suspended Particulate) standard. The PM 2.5 (particles smaller than 2.5 fUll in diameter) standard was introduced in 1997 to specifically regulate fine particles. The 1997 standards also intended to regulate the "inhalable coarse partic les" ranging from 2.5 to 10 fUll (herein referred as CPM, PMcFl · However, most epidemiological studies reported only findings for PM10 and PM 2.5, and they were mostly conducted in urban areas where the levels of fme particles were higher than those of coarse particles (Abbey et a!., 1995; Dockery and Pope, 1994; Dockery et a!., 1993; Schwartz et a!., 1994). Very limited studies provided clear quantitative evidence to link coarse particles to adverse health eff ects. Due to the abovementioned reasons, the Enviro rnnental Protection Agency (EPA) concluded at that time to continue to use PM10 standards to control thoracic coarse part icles. In the last few years, the difference between coarse and fine particle has become more explicitly understo od and appreciated. There is a growing, but still limited, evidence of health effects induced by coarse part icles. A number of epidemiological studies were conducted in areas of lower PM2 .5 I PM10 ratios such as Rena, NV and Phoenix, AZ (Chen et a!., 2000; Lipsett et a!., 2006; Smith et a!., 2000), allowing a more relevant evaluation of health effe cts imposed by CPM. Some relatively inexpensive in-vivo and in-vitro tests developed in recent years also become readily available to stimulate PM-induced toxicological responses, which provide a sc ientific basis for understanding the pathological pathway of PM-induced health eff ects . Furthe rmore, since the PM 2.5 standard was in place in 1997, PM 2.5levels have come down and thereby leading to more comparable levels of fine and coarse PM even in urban areas . As a result, there has been 6 a ma jor consideration on esta blishing a new PM 10 _2_5 standard to specifically regnlate coarse particles. The potential change of PM standard has been controv ersiaL There are concerns regarding the inconsist ent CPM-induced adverse health outcomes demonstrated in health studi es. The new standard may impact urban and rural areas differently because of the different sources and size distribution of PM in these areas . The implementation of a new PM 10 _2_5 standard would also devalue the historical PM 10 measurements and probably require additional monitoring stations and equipment Currently, CPM is regulated under the PM 10 standa rds. PM 10 consists of both fme and coarse partic les, of which the sources, fo rmation, and removal mechanisms are entirely different Due to higher sett ling velocity and distinct origins, CPM is more affe cted by local sources, and is more heterogeneous compared with particles in the fine mode. Thus, the temporal and spatial variation of coarse part icles, as well as the chemical composition could be different from place to place. Therefore, regulating CPM using PM10 standards might not be appropriate in cont rolling this thoracic coarse fraction ofPM10. Additionally, limited literature has studied coarse particles specifically, in contrast to the extensive monitoring network currently in place for PM10 and PM 2_5_ PM10 _2_5 measurements were often obtained using the diff erence between PM10 and PM 2_5, which introduced higher uncertainti es, particularly in areas with high PM2 _5 I PM10 ratio_ As a result, sig nificantly less is known about the phys ico-chemical and toxicological characteristics of CPM. The distinct chemical composition in the fine and coarse fractions could also contribute to potentially different health out comes. Since these parameters are important in distinguishing the characteristics and health effects of CPM from those of PM10 or PM2_5, it is desirable to have comprehensive investigations on CPM based on independent ambient measurements_ 7 In order to provide a scientific basis to develop cost- effe ctive regulations to regulate CPM, two comprehensive investigations were conducted in Southern California to examine the phys ical, chemical and toxicological characteristics of coarse part icles. The first proj ect involved a year-long sampling campaigu to collect ambient coarse particles in 10 sampling sites across the Los Angeles Basin, and the subsequent physical and chem ical analys es of these samples. To better understand the toxicity of coarse partic les, the production of reactive oxygen species (ROS) was used to characterize the oxidative potential of the collected CPM. In the second proj ect, the diurnal variation of coarse particles was studied to improve the understanding of important atmospheric parameters that influence the ambient CPM concentration and composition, as well as public health impacts as a result of human exposure. The results of these two studies were integrated with published literature to provide a comprehensive investigation to establish the linkage between sources, composition, and the toxicity of coarse part icles. The objectives of the present investigation work are: I) To understand the cause of the temp oral and spatial variation ofCPM concentrations and chemical composition, and identify their sources in the atmosphere; 2) To identify the linkage between toxicity levels and source-specific chemical constituents in CPM with distinctive origins, and 3) To determine the most eff ective strategy to regulate CPM and provide valuable scientific information for environmental policy decision making. The results of this investigation will ultimately help decide if a PM10.2.5 standard is more beneficial than a PM10 standard to regulate coarse part icles. 8 1.4. Thesis Overview This thesis presents my doctoral research work under the supervision of Professor Constanti nos Sioutas with the goal to provide a scie ntific basis for effe ctive regulation of coarse particles. The thes is includes the following chapters: Chapter I provides an overview of urban particulate matter, and the rationale of this in vestigation. Chapter 2 presents the ambient CPM chemical composition at I 0 distinct locations in the Los Angeles Basin. The sources and form ation mechanisms of coarse particles are discussed. The contributi ons of different sources, as well as their temp oral and spatial variations are also examined. Chapter 3 focuses on the seasonal and spatial variations of CPM-bound individual organic compounds, and their relation to sourc es. Chapter 4 describes the historical data ( 1986-2009) of CPM mass concentrations and chemical compositions in 3 different areas in the Los Angeles Basin, and discusses their relations to sources and air quality regulations . Chapter 5 presents the diurnal chemical profile of CPM, which gives insights on the ambient conditions that affect the fo rmation of coarse part icles. Chapter 6 explores the relationships between the chemical composition and the toxicological profiles of CPM. The results will help identify the sources that are responsible for driving the toxic activity of coarse part icles. Chapter 7 concludes the findings of the present investigation and outlines the possible strategies that would help establish cost- eff ective air quality standards to protect the public health 9 fr om CPM exposure. It also identifies limitations of the current investigation and provides suggestions for future research on this sub ject. 10 Chapter 2 Spatial and Tern poral Variation of Chemical Com position and Mass Closure of Ambient Coarse Particulate Matter (PM1"2 . 5) in the Los Angeles Area 2.1. Abstract To study the seasonal and spatial charact eristics, as well as chemical composition of coarse particulate matter, 10 sampling sites were set up in different areas of the Los Angeles Basin. Ambient CPM was collected for a year-long in 24-hour periods once per week during weekdays, and was analyzed for elemental and organic carbon (EC-OC), water soluble inorganic ions, and total metals and elements. Five categories were used to reconstruct PM mass: 1) crustal materials and other trace elements 2) organic matter 3) elemental carbon 4) sea salt and 5) secondary ions. Overall, crustal materials and other trace elements were the most abundant category, accounting for an average of 47.5 ± 12% of the total reconstructed mass. Secondary ions (sulf ate, nitrate and ammonium) and organic matter also contributed significantly at mass fractions of around 22.6 % and 19.7%, respec tively. Elemental carbon was a less significant component, accounting for less than 2% of total mass across sites. Sea salt particles were more prevalent in spring and summer (12. 7%) due to the strong prevailing onshore southwesterly wind in that period. Mass fractions of organic matter, as well as crustal materials and other trace elements were higher in fall and winter, indicating that their contributions were not affected by the lower wind speed and change in wind direction during that period. PM concentrati ons of sea salt particles decreased fr om coast to inland along the tra jectory of LA Basin, while crustal materials and other trace elements became dominant at inland sites . On the other hand, organic carbon well-correlated with tracers of soil dust (R � 0. 74 and 0. 72 for Ti and Fe respectively), suggesting that humic substances might be the ma jor constituent of organic matter in coarse mode aerosols in the Los Angeles Basin. 11 2.2. Introduction Due to concerns of health effects of exp osure to coarse part icles, there has been considerable discussion about explicitly regulating CPM. Currently, coarse particles are regulated using PM 10 standard. PM 10 is a mixture of coarse and fme airborne particulate matter, both of which have diver se predominant sources and removal mechan isms; thereby, chemical composition is signif icantly diff erent. Origins of PM 2_5 include fossil fuel combustion and photo oxidation of gas precurs ors, while CPM arises predominantly from mechanical disruption and attrition proces ses. Source apportionment studies showed that the composition of coarse particles varied widely depending on the study areas; ma jor components include re-suspended soil, road and street dust, fugitive dust, sea salts and biological materials (Hwang et a!., 2008; Paode et a!., 1999). Unlike PM10 and PM 2_5, for which abundant data of continuous and time-integrated mass is generated by local and state agencies, a nationwide PM10.2_5 monitoring network is not currently in place. Moreover, limited peer-reviewed studies have examined comprehensively the chemical composition of CPM in urban and rural areas of the U.S., thereby limiting the understanding of the link between their sources, their phys ico-chemical characteristics and their toxicological propert ies. In this study, 10 sampling sites were set up in different areas in the Los Angeles Basin to collect time-integrated ambient CPM for an entire year-long period. The characterization of the CPM physical propert ies, including the mass concentrations and its spatial and temp oral variati on, has been presented in our previous publications (Moore et a!., 20 10; Pakbin et a!., 2010). This chapter focuses on the much-needed information on CPM chemical composition and their sources of formation, as well as their spatial and seasonal characteristics in urban and rural areas of the 12 Los Angeles Basin. This comprehensive dataset will be a key tool to policy makers in providing vital information for conducting epidemiological health studies, for setting up effective compliance monitoring in Southern California, and ultimately for designing effective CPM control strategie s. 2.3. Methodology 2.3.1. Site description and sampling time Ten sampling sites were selected and set up in the Los Angeles Basin to fully characterize the range of conditions encountered in Southern California. The location of each site is shown in Fignre 2.1. Selection criteria and site characteristics were described in detail in a previous publication of this study (Pakbin et a!., 2010). Sites were categorized according to their geographical locations. HUD is located in a residential and commercial neighborhood about 2 krn inland of the Ports of Los Angeles and Long Beach. It is 100 m and 1.2 krn away from the Terminal Island Freeway and I-710, respec tively, representing a pollutant source region. LAN is situated in a desert-dominated rural region in the City of Lancaster in the north of the Los Angeles County. It is a typical desert site over 2 krn west of fr eeway CA-14. The 8 remaining sites are mostly on the tra jectory of Los Angeles Basin, which can be grouped into West LA (GRD and LDS), Central LA (CCL and USC), East LA(HMS and FRE) and Rive rside (GRA and VBR) as moving from coast to inland. The 6 sites in the Los Angeles area (GRD, LDS, CCL, USC, HMS and FRE) are urban sites with most of them located near freeway (< I krn) except for CCL and GRD. The two inland sites in River side County (GRA and VBR) are considered as semi-rural receptor sites . 13 The one-year sampling campaign was conducted from April 2008 to March 2009. A 24-hour time-integrated sample was collected at each site once per week from 12:00 AM Pacific Standard Time (PST) to 12:00 PM PST. All substrates were collected on weekdays and removed from samplers within 24 hours after collection. Data recovery was over 88% overall. A total of 470 daily samples were submitted for chemical analysis. Monthly concen trations were calculated as the arithmetic averages of daily concentrations for all the chemical species. Ventura 0 5 10 Kilometers L____j__J Los Angeles l - Los � usc� • FRE • • • HMS •GRD CCL Figure 2-1: Map of the 10 sampling sites. 2.3.2. Sampling equipment and setup San Bernardino Riverside o..a ,lge 1 At each site, dual Personal Cascade Impactor Samplers (PCIS) were employed with PM 10 inlets (Pakbin et al., 2010 ) to collect size-segregated particles in the size range of 10-2.5 J...L m, 2.5-0.25 J...L m, and less than 0.25 J...Lm at 9 liters per minute (LPM). 25mm Zefluor (3 J..!ill pore, Pall Life Sciences, Ann Arbor MI) and quartz fiber (Whatman Inc., Florham Park, NJ) filters were used in the coarse stage of each PCIS. Zefluor filters were weighed before and after sampling for 14 determ ination of collected mass using a Mettler Microbalance (Mettler- Toledo, Columbu s, OH; weight uncert ainty ± I fig) after 24-hr equilibration under cont rolled temperature (21 cC ± 2cC) and relative humidity (30% ± 5%). A Coarse Particle Concentrator (Misra et a!., 2001) --composed of a virtual impactor upstream of a filter holder-was set up in parallel to collect particles between 2.5-10 f!m using the PM 10 inlet as mentioned previously. The virtual impactor was operated at 50 LPM with a minor flow rate of 2 LPM. Part icles were enriched by 25 times and collected on Teflon substrates (47mm, TefloTM, 2.0 f!m pore size, Pall Cord, East Hills, NY). Substrates were also weighed before and after sampling to measure the collected mass. Comparisons between CPM mass concentrations measured by the PCIS (coarse fraction) and Coarse Particle Concentrator were described in Pakbin et a!. (2010). In bri ef, they agree within 20% or less except in the site of Lancaster, which experienced particle bouncing at the PCIS (Pakbin et a!., 2010). 2.3.3. Chemical analysis Water extracts of the quartz filters collected from PCIS were analyzed for water soluble inorganic ions using ion chromatography (Lough et a!., 2005). Elemental carbon and organic carbon were determ ined by the NIOSH Thermal Desorption I Optical Trans mission method (Birch and Cary, 1996) using the same quartz filt ers. The PCIS Zefluor filters were analyzed by the Inductively Coupled Plasma-Mass Spectroscopy (ICP-MS) to determ ine the concentrations of the total metals and elements. Lough et a!. (2005) described in detail the procedures for the process of analysis. In brief, filter membranes were extracted using a mixture of 1.5 rnL of 16 N HN0 3 , 0.2 rnL of28 N HF, and 0.5 rnL of 12 N HCI, and subseq uently digested in closed Teflon digestion vessels. Digestates were analyzed for total metals and elements using ICP-MS (PQ 15 Excell, ThermoElementa l). Due to the issue of particle bouncing in the PCIS at the Lancaster site, substrates collected fr om the Coarse Particles Concentrator were used instead of those from the PCIS for the total ICP-MS analys is at this site. CPM samples collected using the Coarse Particle Concentrator contain a small fraction ( 4%) of fine PM that are inhe rently included in the minor flow of the virtual impactor. With the comparable levels of trace elements and metals in fme and coarse PM fractions previously reported in this basin (Krudysz et a!., 2008; Sardar et a!., 2005; Singh et a!., 2002), it is unlikely that the concentrations of coarse-pa rticulate metals and elements would be significantly affected by the inclusion of this small fraction of fine part icles, so no corrections were made to account for this very small fraction of fine PM included in the coarse mode. A detailed description of the characteristics of individual metals and elements were presented in depth in a separate paper due to their importance in source apportio rnn ent and epidemiological studies, while this chapter focuses on the chemical composition, as well as the spatial and temporal properties of coarse particles. For all analytical measurements, the samples were analyzed along with field blanks, laboratory blanks, filter spikes and external check standa rds. On average, one blank filter was analyzed for every 10 samples. Sample spike recoveries were within the acceptance range of 85-ll5 %. Uncertainties of the chemical measurements were propagated using the square root of the sum of squares method based on the analytical uncertainties and uncertainties from blank correct ions. 2.4. Results and Discussion 2.4.1. Meteo rology Table 2. 1 shows selected meteor ological parameters for four seasons at sites clustered in five distinct geographical regions of the basin. Overall, mean temperature and relative humidity 16 exhibited little seas onality at urban Los Angeles sites, with more inte nse variability observed inland. With the exception of Lancaster, temperatures in fall were similar to those of summer, due to the extended warmer months of September and October, which is very typ ical in this region. Total precipitation was virtually insignificant in spring and summer while highest in winter. In Long Beach and along the coast, relative humidity levels were highest (61.6- 76.8% and 58.3 - 81.0%, respec tively) and peaked in summer. Further inland, at the Rive rside and Lancaster sites, season-to-season fluctuations in temperature and relative humidity became more pronounc ed. Particularly, at the Lancaster site, the mean temperature spanned a broad range of 8. 7- 27.9°C over the seasons. Relative humidity was also lower than all other sites (26.9 - 55.9 %) and dropped to as low as 26.9% in summer. These rather extreme oscillations highlight the "desert-like" conditions at this site. Table 2-1: Selected meteorological parameters by location and season. Temperature (0C) Relati\€ humidity (%) Precipitation (mm) Wind (m/s) A\€rage A\€rage Total Speed (calm%) Direction Spring 16.3 65.2 12.0 1.6(4.1) sw Summer 21.3 76.8 0.7 1.9 (3.3) sw Long Beach Fall 19.9 67.8 1.2 (5.6) w 54.8 Winter 13.8 61.6 168.9 0.7(2.9) NW Spring 15.0 69.0 10.9 1.7 (9.8) w Summer 19.4 81.0 1.5 0.7(26.2) w West LA Fall 19.0 65.3 98.0 0.2(50.2) NE Winter 13.9 58.3 215.8 0.2(26.2) NW Spring 17.2 59.0 9.7 2.1( 0.6) sw Summer 22.8 66.7 0.0 3.6(1.2) sw Central and East LA Fall 20.9 55.5 184.9 1.6 (4.6) sw Winter 14.1 55.6 182.6 1.5 (0.6) NE Spring 17.4 63.4 1.8 2.6(5) NW Summer 25.6 65.0 5.6 3.6(2.7) w Ri\€rside Fall 22.0 56.8 1.5 1.6 (6.5) NW Winter 13.8 61.8 78.5 1.3 (3.9) N Spring 15.2 40.5 NA 4.1(13.5) w Summer 27.9 26.9 NA 4.3(1 1.6 ) w Lancaster Fall 18.8 36.5 NA 1.5 (30.5) w Winter 8.7 55.9 NA 1.5 (33.6) w 17 Vector average wind speeds and directions are also shown in Table 2.1. As expected, wind direction across the sites was consistent with the prevailing air traj ectory crossing the Los Angeles Basin fr om coast to inland (Eignren-Femandez et a!., 2008). In the West LA, Central LA and East LA regions, winds originated from the sou th west and west for most of the year, in accordance with the typ ical onshore flow patterns of the basin. As the air parcel proceeded inland to Rive rside, the wind was deflected to northw esterly or westerly. However, this pattern was not as persist ent in winter and I or fall as wind predominantly had a northerly component. In general, wind speeds were stronger away fr om the coast, and higher wind speeds were observed in spring and summer. 2.4.2. Data overview Statistical summaries of annual mass concentrations of selected chemical species and gravimetric CPM are listed in Table 2.2, including the average concentrations and assoc iated standard errors of gravimetric mass, prevalent inorganic ions, carbonaceous compounds, and the most significant elements and metals . Across the basin, CPM gravimetric concentrations were highest at HUD due to its proximity to industrial and vehicular CPM sources. In contrast, the CPM concentrations at the "desert-like" site LAN were the lowest because of its remote location fr om urban sources . Although not shown in the table, lower overall PM mass concentrations were observed in winter ( 6.93 ± 1.3 flg/m3) compared to spring, summer and fall ( 12.3 ± 2.3 flg/m3 ). This is consistent with the results of Sardar et a!. (2005), which determined that wintertime conditions of low wind speeds and higher soil moisture lead to less re-suspension of road and soil dust, the main contributors to CPM. 18 Nitrate was the most predominant inorganic ion fo und at the 10 sampling sites, with concentrations of 1.60 ± 0.84 flg/m3 These levels were relatively lower than those reported earlier in this basin (Sardar et al., 2005) averaging to 2. 78 ± 1.7 flg/m3 OC was overall a ma jor CPM constituent across the basin, averaging 1.01 ± 0.56 flg/m3, with more pronounced concentrations at the Riverside sites. Sardar et al. (2005) also reported relatively higher levels of coarse particulate OC (1.46 ± 0.62 flg/m3) compared to this study. Among elements, those that are typ ical of crustal materials and soil, such as Fe, Ca, AI, Mg and Ti (Lough et al., 2005), were prevalent at all sites. Fe was the most dominant crustal element (395.5 ± 180 ng/m3) across the basin with the exception of the Riverside and Lancaster sites, where concentrations of Fe and AI were similar. On the contrary to minerals, elements that are indicative of traffic origin in the form of re-suspended road dust, tire and brake wear, such as Ba, Zn and Cu (Lough et al., 2005), were less abundant. Their concentrations were generally lower than 50 ng/m3, indicating the lower contribution of traffic-related emissions to coarse particles. EC was generally low across all sites (84.6 ± 94.3 ng/m3 ), with the highest concentrations at HUD (190.4 ± 141 ng/m3 ), consistent with its mixed residential/ commercial nature and proximity to traffic. Table 2-2: Statistical summary of annual concentrations (ng/m3) of selected chemical species and gravimetric CPM mass concentrat ions. Long Beach West Los Angeles Central Los Angeles East Los Angeles RIIRrSide Lancaster �cc GCC cc� CCC c'c �M� me vee GCA CM NH4+ 546±157 43 6±14 4 48 9±15 2 58 7±17 5 472±188 65±18 4 96 7±23 2 88 5±23 4 132±41 2 17±6 65 " 631±144 690±145 699±132 525±146 554±161 429±132 306±97 5 235±67 6 179±68 201± 79 1 sol- 620±55 3 458±95 8 623±62 1 571±64 2 615±66 5 483±92 3 522±61 4 460±54 7 383±89 9 205±17 3 NOi 1510 ±158 1430±172 1670±179 1860±202 1860±240 201 0±257 1920±233 1720±239 1600±332 452o47 2 Mg 149±9 07 125±13 9 122±15 5 118±946 153±15 7 121±12 1 113±13 9 166±18 9 155±21 4 85 3±21 2 AC 357±39 6 160±20 1 248±39 8 192±23 4 293±30 2 234±174 237±42 2 433±43 8 366±46 335±83 6 Co 340±33 2 173±24 1 183±15 2 243±39 8 328±29 4 250±16 3 246±37 2 432±42 5 389±48 6 213±51 " 28 9±3 14 16 5±248 20 9±3 23 5±2 95 28 7±3 08 23±1 32 27 9±3 3 34 6±3 33 32 6±4 1 30 1±7 16 Ce 462±61 2 286±50 3 367±46 4 397±63 3 487±58 4 373±26 1 410±38 8 461±42 1 425±49 4 307±62 9 Co 21 3±3 1 16 9±324 21 2±3 07 28±4 93 28 9±3 78 22±1 78 25±2 04 15 2±1 55 16 1±1 6 8 7±1 49 " 14 1±2 13 5 88±1 24 5 62±0 99 8 3±1 4 895±1 19 80 1±08 11 2±2 86 5 96±0 58 9 62±2 22 5 17±0 73 c" 128±1 85 18 8±259 20 9±2 34 241±302 22 9±2 39 15±1 47 21 6±2 58 105±1 33 10 9±1 41 21 5±8 82 oc 1190±214 767±121 809±125 1 030±1 79 942±171 1070±121 1100±103 1380±214 1250±146 538±121 ec 190 4±43 6 63 2±19 2 54 7±20 3 871±272 81 4±27 7 88 8±32 2 80 9±21 9 84±23 2 84 2±20 6 349±101 Gra�Ametnc CPM 13350±1059 9990 ±1101 9740±857 2 1 0080±964 9 11490±1447 10800±964 7 10010±1210 13180±1641 1 0550±1 862 9373±1546 ' Results reported as mean±st andard error (nglm 3) 19 2.4.3. CPM mass reconstruction methodology For the purp ose of chemical mass reconstruction, chemical components were grouped into five categori es: crustal materials and other trace elements (CM + TE), organic matter (OM), elemental carbon (EC), sea salt (SS), and secondary ions (SI). CM represents the sum of typical crustal materials, including AI, K, Fe, Ca, Mg, Ti and Si. Each of these species was multiplied by the appropriate factor to account for its common oxides based on the fo llowing equation (Chow et al., 1994; Hueglin et al., 2005; Marcazzan et al., 200 1): CM � 1.89AI + 1.21 K + 1.43 Fe + 1.4 Ca + 1.66 Mg + 1.7 Ti + 2.14Si (1) Elemental Si was estimated by multiplying AI using a factor of 3.41 (Hueglin et al., 2005) since it was not analyzed by the ICP-MS method used in this study. The Ca and Mg oxides were calculated using the non-sea salt (nss) portion of Ca and Mg. Since oxygen was not measured in the CPM samples, the mass of Ca and Mg oxides were estimated by multiplying Ca and Mg by 1.4 and 1.66, respe ctively, to account for the oxygen assoc iated with these species. It should be noted that this approach would lead to an overestimation of the mass associated with Ca and Mg if they were ions associated with nitrate, sulfate or other directly measured ions, for which a multiplier is not needed, since these components of the Ca and Mg were measured directly. A sens itivity analysis was conducted to investigate the impact of this overestimation on the overall PM mass closure. The calculated maximum possible overestimation for Ca and Mg is 48.8 ± 7.6 ng/m3 and 19.7 ± 2. 1 ng/m3 (mean ± standard error), respec tively, which is < 1% of the overall reconstructed mass. Therefore, their impact on the overall mass closure is minimal. To test the effectiven ess of summing metal oxides in estimating mineral content, Andrews et al. (2000) applied this approach to various measured soil compositions across the United States and fo und 20 that 50-90% of the measured PM sample mass could be accounted for, depending on the soil type. Other trace elements (TE) include elements such as nss-Na, Cu, Zn, and Ba. OM was obtained by multiplying the measured concentration of organic carbon ( OC) by a factor of I. 8, which is based on an average of the recommended ratios of 1.6 ± 0.2 for urban aerosols and 2. 1 ± 0.2 for aged or non-urban aerosols (Turpin and Lim, 2001). It should be noted that the average molecular weight per carbon weight ratio may vary with source character ization, site location, and between seasons. In our analys is, the same factor has been applied across the 10 sites and over four seasons, which may introduce some uncertainties in the overall estimations of OM to total mass. The SS contribution represents particles in the form of fr esh sea salt. It is computed as the sum of measured chloride ion concentration plus the sea salt fr action of concentrations of Na +, Mg 2 +, K+, Ca 2 +, SO/ based on the composition of seawater and ignoring atmospheric transformations (Seinfeld and Pandis, 2006): SS � cr + ssNa + + ssMg 2 + + ssK+ + ssCa 2 + + ssSo /· (2) where ssNa + � 0.556 Cl", ssMg 2 + � 0. 12 ssNa +, ssK+ � 0.036 ssNa +, ssCa 2 + � 0.038 ssNa +, and ssSo /· � 0.252 ssNa+ (Terzi et a!., 2010). The EC contribution was reported as measured by thermal desorption. The SI contribution was calculated as the sum of nss-So/·, NH/, and NO;, where nss-So/· is total measured So/· minus the sea salt fr action of So/·. Fignre 2.2 shows a linear regre ssion (least squares) of the daily reconstructed and gravimetric mass concentrations for all sites and all mont hs. Thirteen out of 470 data points were omitted fr om this analys is, of which four data points were detected as outliers (standard deviation > 3) and nine data points were excluded due to incomplete chemical data. The gravimetric and reconstructed mass concentrations show a generally strong correlation, with a coefficient of 21 determination (R 2 ) of 0.69, indicating overall good agreement between the reconstructed mass and the gravimetric mass. Further analys is by season, as demonstrated in Figure 2.3 (a-d), reveals a seasonal trend in the correlation of the reconstructed and gravimetric mass concentrat ions. The season-based R 2 value ranges fr om 0.62 to 0.81. The overall ratio of reconstructed mass to gravimetric mass concentration is 0.89 ± 0.02 (mean ± standard error) with a seasonal variation of 0.95 ± 0.02, 0.75 ± 0.03, 0.84 ± 0.03, 1.02 ± 0.04 for spring, summer, fall and winter, respec tively. These ratios are consistent with other published literature, with average ratios varying from 0. 73 - 0.96 (Hueglin et a!., 2005; Sillanpaa et a!., 2006; Terzi et a!., 2010). Higher fr actions of unidentified mass were found in summer and early fall, which is consistent with other studies that showed more unidentified PM mass in summer than winter (Ho eta!., 2005; Hueglin et a!., 2005). The higher unidentified fr action could be att ributed to the uncertainty in the OC multiplication factor used as discussed earlier coupled with some uncertainties in the conve rsion and estimation of elemental oxides. The discrepancy in the conve rsion factor used to estimate organic matter (OM) from organic carbon (OC) depends greatly on source characterization of organic component, which is generally not well known for coarse particles and may have considerable seasonal variation. For example, Zhao and Gao (2008b) demonstrated the presence of dicarboxylic acid, particularly oxalic acid, in coarse size fr action, and suggested that the fo rmation of coarse particulate oxalate by possi ble photochemical reactions was favored at higher temperature. The conve rsion fa ctor used in our study may also under-represent the contribution of organic matter fr om soils, which is likely to contain sugars and amino acids that have higher molecular weight per carbon weight ratio (Turpin and Lim, 2001). 22 " OE y = 0 66 (±0 02) X+ 1 76 (±0 25) � ,; R 2 = 06 9, N =457 • • 0 I " • ,; � � u w 0 t 1 ; � w " Grav1metnc mass concentration (1Jg/m 3 ) Figure 2-2: Linear regres sion of overall reconstructed vs. gravimetric mass concentrat ions. Values in parent heses represent standard errors of the slope and intercept. "i 2 25 � 20 ' 3 15 � 1 " , ,, l ,; � " ' 3 ,; � 1 � w � ' ; " y = 0 83(±0 04) X+ 0 99(±0 48) R 2 =07 6, N= 120 Grav1metnc mass concentration (�g/m 3 ) • • .. • y = 0 65(±0 + 1 74(±0 63) R 2 =0 N= 124 Grav1metnc mass concentration (�g/m 3 ) ''""c-----------------------------y----, "i 2 25 � 20 ' 3 15 � 1 " 'E � ,; � " � ,; � 1 � w ! ; " '" ' • y=0 61(±005 )x+1 38(±062) R 2 =06 2, N= 109 Grav1metnc mass concentration (�g/m 3 ) • • y = 0 76(±0 04) X+ 1 33(±0 32) p 2 =0 81, N = 104 Grav1metnc mass concentration (�g/m 3 ) Figure 2-3: Linear regression of reconstructed vs. gravimetric mass concentrations in: (a) spring; (b) summer; (c) fall and (d) winter. Values in parentheses represent standard errors of the slope and intercept. 23 2.4.4. CPM mass chemical composition Figure 2.4 (a-d) illustrates the chemical mass closure segregated by seasons (spring, summer, fall and winter) using the 5 main CPM categories, as mentioned previously. Sites were arranged according to their distance fr om coast (i.e. West LA, Central LA, East LA and Rive rside) and their unique characteristics (i.e. the HUD site being dominated by traffic emissions, while LAN being a "desert" -like site). Annual mass ratios of the 5 categories against the total reconstructed mass, computed as the averages of all monthly mass concentra tions, are shown in Table 2.3 to demonstrate the overall spatial variation in CPM mass composition. "E 0, 3 " 20T<,�.,---r==========�--------------� - CM+ TE 15 =o M - EC =ss -S i -o- Gravimetric Mass g 10 � � 8 ME 0, 3 " 0 g l" � 0 u 20 (C) 15 10 HUO 0 HUD GRD LOS CCL USC HMS FRE VBR GRA Site GRD LDS CCL USC HMS FRE VER GRA Site LAN D LAN 15 M � 3 " g 10 "' � � 0 u ME 0, 3 " 0 g l" � 0 u 20 15 10 HUD (d) D HUD GRD LDS CCL USC HMS FRE VBR GRA Site GRD LDS CCL USC HMS FRE VBR GRA Site LAN Figure 2-4: Chemical composition and gravimetric mass concentration by location in: (a) spring; (b) summer; (c) fall and (d) winter 24 Crustal materials and other trace elements. Crustal materials and other trace elements were the major contributors to CPM mass in all sites with an average percentage of 41.2, 44.0, 42. 7, 54.7% of the total reconst ructed mass in West LA, Central LA, East LA and Rivers ide, respec tively, as shown in Table 2.3. In general, mass fr actions of crustal materials and other trace elements at the two River side sites (VBR and GRA) and the Lancaster site (LAN) were higher than those in the urban Los Angeles sites for all seasons except winter. The elevated concentrations at the inland sites can be associated with the relatively higher wind speeds and lower relative humidity in these areas (particularly in spring and summer, as shown in Table 2.1), and their "rural" nature compared to the urban Los Angeles sites; therefore, windblown dust contributed more to the coarse particle mass concentration. The percentage of the Long Beach site (HUD), located near the I-710 fr eeway, was high at 49.4%. Inc reased levels at HUD may be attributed to its location in a pollutant "source" region of the basin and the enhancement of re-s us pension of traffic emissions in the form of road dust. As discussed earlier, HUD is located near fr eeways with a large volume and high fr action of Heavy Duty Diesel Vehicles (HDDV s) due to its proximity to the Ports of Los Angeles and Long Beach (Pakbin et a!., 2010). HDDVs are known to induce higher particle re-suspension (Charron and Har rison, 2005) with brake-wear emissions than light duty vehicles (Garget a!., 2000) thus contributing to higher roadway CPM emissions. The contribution of crustal materials and other trace elements was highest at 60.5% in Lancaster. Lancaster is a rural site with completely different chemical profile compared with other sites in the basin, and metals and elements were overall the most dominant componen ts. 25 Table 2-3: Annual mass fr actions (%) (mean ± standard deviation) of CPM components. Long Beach West LA Central LA East LA Riverside Lancaster CM+TE 49.4±6.7% 41.2 ±9.5% 44.0±10% 42.7±7.9% 54.7±8.0% 60.5±21 .6% OM 18.6 ±7.3% 18.0±8.3% 19.2±8.7% 22.1 ±7.0% 21.7±6.2% 16.0±1 1. 2% EC 1.7 5±1.3% 0.88±1 .1% 0.94 ±1.0 % 1.16 ±1 .3 % 0.99±0.9% 0.89 ±1.0 % ss 10.8 ±8.8% 15 .1± 8.6% 10.2±8.3% 6.94±6.6% 3.37±3.6% 10.2±17.3% Sl 19.4 +5.3% 24.9+6.6% 25.6+9. 1% 27.2+8.1% 19 .2+7.7% 12 .4+7. 1% With a few exceptions in spring and summer, when sea salt concentration was higher and inorganic aerosol fo rmation was more prevalent, crustal materials and other trace elements were the most abundant species in coarse mode aero sols. In fall and winter, metals and elements were always the major contributors across all sites with the exception of the Lancaster site. In general, CPM mass concentrations were lower in cooler months due to stable atmospheric conditions, coupled with lower average wind speeds and higher average precipitation as shown in Table 2. 1. Despite the lower CPM mass concen trations, the percentage contributions of crustal materials in winter (43 .5 ± 9. 8%) were similar to or higher than tho se of spring (38.3 ± 13%), summer (40.7 ± 16%) and fall ( 44. 1 ± 11 %), indicating the significant contribution of the soil dust materials to CPM in this basin. In order to assess the relative contributions of anth ropogenic vs. crustal sources of trace elements bound to CPM, crustal enrichment factors (CEFs) were calculated by divid ing the selected element abundance in the PM sample by their average abundance in the upper continental crust (UCC) obtained fr om Taylor and McLennan (1985), after normalization to AI as the reference element. The CEFs of the 10 sampling sites are presented in Fignre 2.5, with sites categorized geographically into Los Angeles (GRD, LDS, CCL, UCS, HMS and FRE), Long Beach (HUD), Rivers ide (VBR and GRA) and Lancaster (LAN). Typically, CEF > 10 are indicative of PM sources different from crustal material, notably anth ropogenic sources . The 26 CEFs were generally higher at urban sites of Los Angeles and Long Beach, with lowest CEFs observed in Lancaster. Elements with the highest CEFs were Sb, Sn, Mo, S, Cu, Ph Zn, and Ba, which are indicative of traffic-related emissions (Lough et al., 2005). The higher CEFs observed at urban sites provide further evidence that the ma jority of these transition metals are released in the atmosphere through abrasive vehicular emissions, particularly the wear of brake and tire lining (Lin et al., 2005). 10000 r------------r============;] !% 1000 2 ... 0 & c s .c -� >tl 0! � u 100 10 Los Angeles ...... - 0--· Loog Beach -- --?- -- Riv6"side ---� ---· Lancaster Sb Sn Mo S Cu Fb Zn Ba Na Cr P Nt Sr La Fe V Ce Mg Ca Mn K Tt Figure 2-5: Crustal enrichment factors of individual elements at Los Angeles, Long Beach, Riverside and Lancaster. The refe rence element is Al. In this study, Fe and A1 were the two most abundant minerals, together with Si (which as we noted earlier was estimated fr om the A1 concentrati ons), their sum accounted for an average of 62.0% of total measured crustal materials and other trace elements, and 29.5% of total reconstructed mass. This is consistent with other studies which show that crustal material is a major source of coarse mode aerosol in both urban and sub-urban environments (Hueglin et al., 2005; Sillanpaa et al., 2006). On the other hand, elements related to anth ropogenic sources, such as Cu and Sb fr om brake wear (Lin et al., 2005), had concentrations that were low compared with 27 the mineral dust. Overall, natural crustal elements were the major component of this category, while anthropogenic sources contributed to a much lesser extent. Carbonaceous compounds. Organic matter was a substantial constituent of CPM, with an average contribution of 19.7 ± 8. 1% to the total reconstructed mass. EC concentrations were low in general, contributing to approximately 1.1% of total reconstructed mass, consistent with the literature that EC mostly exists in the fine PM mode (Huang and Yu, 2008). Seasonal average of OM fr action was generally higher in fall (23.0 ± 6.8 %) and winter (24.5 ± 8.3%) compared with spring (16.5 ± 6.6%) and summer (14.6 ± 6.0%). Although PM mass concentrations were lower in winter, OM fr action increased in all sites. It suggested that OM in the coarse mode was not much affected by local meteorology and it probably had a stable source strength year-long. On the other hand, the concentrations of OM were generally lowest in summer in the basin with the exception of the two River side sites. The two inland sites (VBR and GRA) had higher OM concentrations in summer and fall, when the corresponding gravimetric mass concentrations were also higher. Despite the lower OM concentrations in the source sites in summer, OM concentrations remained high in the two inland sites. With their proximity to the Chino dairy farm area which is located along the trajectories fr om coast to inland (Eiguren-Fernandez et a!., 2008), it is possible that biological materials, including humic substances, have contributed to the high organic content. This is consistent with recent literature which showed that organic materials with biological origins, such as bacteria, virus es, fungus, spores, pollens, algae and plant debris, were a significant component of PM10 in urban atmospheres (Bauer et a!., 2008; Winiwarter et a!., 2009). The Lancaster site (LAN), on the other hand, was low in both mass fr actions and concentrations of OM due to its "desert-like" site 28 characte ristics . The Long Beach site (HUD) had the overall highest EC fr action (on average 1. 75% ) likely due to its location near fr eeways with higher number and proportion of HDDV s, which generally emit higher EC compared with gasoline vehicles (Biswas et a!., 2009). Table 2-4: Regression anal ysis of organic carbon (OC) vs. 4 main coarse PM categories and selected species in: (a) spring and summer and (b) fall and winter. (a) y�mx+c R (b) y�mx+c R OC-CM+TE 0.15x+l94.9 0.56 OC-CM+TE 0.2x+278.9 0.72 OC-EC -1.73x+914.9 -0.21 OC-EC 2.97x+755.7 0.43 OC-SS -0.07x+927.4 -0.15 OC-SS -0.3x+ 12923 -0.19 OC-SI 0.09x+627.9 0.22 OC-SI 0.29x+643 .7 0.46 OC-Ti 31.26x+ 127.6 0.68 OC-Ti 39. 16x+ 114.9 0.76 OC-Fe 2.27x+ 139.7 0.59 OC-Fe 2.39x+48. 18 0.74 OC-Cu -18.43x+l094 -0.27 OC-Cu 11.86x+9 12.9 0. 17 While OC particles in fine and ultraf me modes mostly originate fr om primary sources such as fo ssil fuel combustion (Arhami et a!., 2010; Minguillon et a!., 2008) and secondary sources by photochemical reactions in the atmosphere (Verma et a!., 2009b ), the sources of OC in CPM are not well underst ood. Previous studies have fo und good correlations between OC and EC in fine and ultrafine modes PM in the Los Angeles Basin, suggesting their shared common source fr om vehicle emissions (Geller et a!., 2002; Sardar et a!., 2005). However, the overall asso ciation between OC and EC is low (R � 0.36) in this study, indicating that the sources ofOC in CPM might be different fr om-or in addition to those ofEC. As previously mentioned, biological materials constitute a potential source of coarse mode OC. To investigate the sources of organ ics, linear regres sion anal ysis was performed against OC and other CPM categories and specific 29 tracers of some of these CPM group s. Table 2.4 (a-b) shows the regres sion equation and the corresponding correlation coefficient in different seasons. As mentioned above, low ass ociations are fo und between EC and OC. Cu, a tracer of brake wear (Lin et al., 2005), also displayed low correlations with OC. On the contrary, OC is well-correlated with CM and TE, and soil dust tracers ofT i and Fe in all seasons (R = 0.56 - 0.7 6), consistent with results reported in other studies (Koulouri et al., 2008). Table 2-5: Correlation coefficient (R) between organic carbon (OC) and soil dust tracers of Fe and Ti at the 10 sampling sites. HUD GRD LDS CCL usc PRE HMS GRA VBR LAN Fe 0.80 0.69 0.62 0.72 0.87 0.07 0.61 0.83 0.77 0.60 Ti 0.79 0.74 0.52 0.68 0.78 0.12 0.77 0.81 0.81 0.65 Site-specific correlations are shown in Table 2.5 to examine the spatial variability of these high ass ociations between OC and soil dust tracers. With the exception of one site (PRE), which is located <50 m fr om fr eeway 1-10, soil dust tracers are well-correlated with organic carbon (R = 0.52 - 0.87). These correlations indicate that OC and crustal materials either share common sources or that OC may be adsorbed or absorbed onto soil dust PM and collected simultaneously during sampling. Humic substances-a ma jor organic component of soil, might be re-suspended with soil-derived dust particles, which could result in the observed high correlatio ns. Overall, the low correlations between OC and traff ic-related emissions, coupled with generally good correlations between OC and mineral dust tracers suggest that humic substances might be the major component in coarse particulate OC. Since the composition of OC in the coarse fr action might have important public health impacts, further investigation of coarse PM-bound OC composition is essential to evaluate its role in CPM-induced adverse health eff ects . Sea salt. Chloride and sodium ions, and to a lesser extent magnesium and sulf ate ions, comprise 30 the ma jority of sea salt components . Overall, sea salt particles contributed to an average of 9.1 ± 9% across all sites and mont hs, indicating its significance in coarse mode particles in this basin. The percentage contributions during spring and summer ( 12.7 ± 9. 7%) were higher than those in fall and winter (5.9 ± 7.7%). The higher standard deviations were driven mostly by spatial variatio ns, but the overall trend of higher sea salt content in spring and summer was prevalent at every site with the exception of Lancaster. Concentrations of sea salt particles were generally higher in spring and summer due to the increa sed strength of southwesterly winds in the afternoon that transport air masses from the source areas along the coast to the inland receptor areas . Lower concentrations were observed in fall and winter because of the lower wind speed and the change of wind direction at certain sites (Table 2. 1.) Sea salt particles originate fr om the Pacific Ocean located in the west of the Basin. As a result, the mass fr actions of sea salt decreased fr om west to east in all seasons, with an average fr action of 15.1, 10.2, 6.9, 3.4% in West LA, Central LA, East LA and Ri verside respectively as shown in Table 2.3. GRD and HUD had highest concentrations due to their proximity to the ocean. The two inland sites in Ri verside (about 80 krn inland of downtown Los Angeles and about 100 krn inland of the Pacific coast) on the contrary, had lower concentrati ons, ofless than half of the concentrations of the coastal sites. Since the atmospheric lifetime of CPM typically varies fr om hours to days (Ruzer and Harley, 2004), depending on its aerodynamic diameter and wind speed, the loss of these species along the prevailing air parcel advection traj ectory fr om the west to the east of the basin is consistent with expectations. Sea salt concentrations were higher in winter in the Lancaster site. The high concentrations might be attr ibuted to specific local sources of chloride, the identification of which requires furth er in vestigation. 31 Secondar y ions. Overall, secondary ions contributed on average 22.6 ± 8.7% of the total reconstructed mass, with nitrate being the major component. Sulfate particles of sea salt origins contributed to approximately 12.3% of total measured sulfate in this study, which is not included in this category. Mass fr actions of secondary ions were higher in spring and summer (25.4 ± 8. 7%) than in fall and winter (19.7 ± 7.9%). Spatially, sites that were further inland had lower mass fr actions (19.2 ± 7.7% and 12.4 ± 7. 1% for Ri verside and Lancaster respec tively) than sites in the West LA (24.9% ± 6.6%), Central LA(25.6 ± 9. 1%) and East LA(27.2 ± 8. 1%) regions as shown in Table 2.3. 3000 2500 '"'i 2000 c g 1500 � 8 § 1000 0 500 (a) 1- AmrnonJJm 1 c:::::J Non-sea salt surate -N trale J J r J r I J I ' HUD GRD LDS CCL USC H� S FRE VBR GR A Site rl LAN �� -------------------------- -, ( c ) 500 r r r r r r HUD GRD LOS CCL USC HMS FRE V8R GRA LAN Stte 3000 (b) 2500 500 I J J J J J 1 1 J � HUD GRD LOS CCL USC H""S FRE VBR ORA LAN Site 3000 "<dec-) -------------------------- ----, 2500 (')i 2000 c 11500 § 1000 0 500 HUD GRD LOS CCL USC HMS FRE VBR ORA Site LAN Figure 2-6: Ammonium, non-sea salt sulfate and nitrate concentrations by location in: (a) spring; (b) summer; (c) fall; and (d) winter. 32 Figure 2.6 (a-d) shows the seasonal concentrations of ammonium, nss-s ulfate and nitrate ions. Nitrate was the most abundant inorganic ion of CPM, accounting for about 74% of the category of secondary ions. High levels of coarse particulate nitrate, fo rmed by the reactions of nitric acid and its precursors with soil dust and sea salt aerosols, were observed in some other studies in urban areas (Noble and Prather, 1996; Zhuang et a!., 1999). In fall and winter, mass fr actions of nitrate were relatively lower (15.0 ± 6.2%) than those of spring and summer ( 19.7 ± 7. 1% ). Sulf ate concentrations, on the other hand, were considerably lower than those of nitrate, probably because most of the measured sulfate in CPM reflects a "tail" of the upper size range of (NH.) 2 S04, which mostly exists in the fine PM mode (Lun et a!., 2003; Sein feld and Pandis, 2006). The presence of gypsum, which is slightly-to-moderately soluble in water, might also be a possible source of coarse particulate sulfate. The overall low concentration of sulfate ion, coupled with the relatively comparable sulfate concentration at the inland sites, suggests that the contribution of soluble gypsum to overall PM mass is not dominant in this basin. Little spatial variation was observed on sulfate across this basin (excluding Lancaster), with the exception of winter, when sulfate concentrations were generally lower, and siguificant spatial variation was observed. Ammonium concentrations were very low, contributing to an average of less than 1% to total reconstructed mass, consistent with many studies that showed ammonium particles mostly exist in the fine mode (Karageorgos and Rapsomanikis, 2007; Sillanpaa et a!., 2006). NH.N0 3 and (NH.) 2 S04 are formed by the reactions of ammonia with acidic gases and are the two major fo rms of PM ammonium in the urban atmosphere; both ammonium salts exist predominately in the fine mode, with a minor fr action extended to the size range of coarse particles (Lun et a!., 2003; Yo shizumi and Hoshi, 1985). Thus, the low ammonium concentrations observed here might 33 reflect the tail of the upper size range ofPM 2 .5-bound ammonium salts. Ammonium concentrations were particularly low in winter, probably because acidic gases- form ing primarily fr om photochemically initiated reactions-were less abundant in that time period. Higher ammonium concentrations were generally observed in the Ri verside sites (VBR and GRA) possibly due to the higher ammonia concentrations in that area, generated by the nearby dairy farm (Geller et a!., 2004; Hughes et a!., 1999) . 2.4.5. Coefficients of divergence (COD) calculations for chemical species concentrations To study the int ra-urban variability of different chemical species, seasonal CODs were calculated across all sites (with the exception of LAN) for the monthly concentra tions . COD value is a measure of the heterogeneity between sites with a range from 0 to I. A low COD value (< 0.2) indicates a high level of homogeneity in concentrations between sites, while CODs larger than 0.2 are considered heterogeneous (Wilson et a!. 2005). It is important to note that the analytical uncertainty might also increa se the COD values, particularly for the species with lower concen trations. CM+TE OM EC SS Sl N03- nss-804-- NH4+ CM+TE OM EC SS Sl N03- nss-804-- NH4+ Figure 2-7: Coefficients of divergence (COD) for all sites (excluding Lancaster) in: (a) spring and summer and (b) fall and winter. 34 Figure 2.7 (a-b) shows the COD for the 5 main categories and the ionic species in spring and summer, and fall and winter. Box plots were used to show the minimum, lower quartile ( Q 1 ), median, upper quartile (Q3) and maximum values observed. Overall, the median CODs range fr om 0.21 to 0.52 in spring and summer, and 0.18 to 0.41 in fall and winter. In spring and summer, nss-su lfate and nitrate ions, as well as the categories of CM + TE and OM experienced modest heterogeneity, as shown by the relatively lower median COD values. On the other hand, the highest median COD value was found for EC, probably because its concentrations varied depending on proximity to fr eeways or heavy-duty vehicles. Given the low EC concentrati ons, analytical measurement uncertainty might have also contributed to its high COD value. In fall and winter, median COD value was lowest for secondary ions. In particular, nitrate ion has both a low median COD value and a small inter-quartile range, consistent with the results in spring and summer, highlighting its low spatial and temporal variation in this basin. On the other hand, sea salt particles experienced higher heterogeneity and had a large inter-quartile COD range, consistent with their high spatial variation as discussed previously. The concentrations of nss-su lfate also varied sig uificantly across sites in winter, as shown in Figure 2.6d. The variations are illustrated in the higher range of COD of nss-su lf ate. Overall, the categories ofEC and SS had relatively higher COD values than the rest of the measured species or groups of CPM because their concentrations depend heavily on the proximity to their sourc es. In contrast, CM + TE, OM and nitrate had relatively lower COD values, consistent with our discussion in previous sections indicating that these species are dominant components at every site and have relatively stable source strengths year-long. 35 2.5. Summary and Conclusions Crustal materials and other trace elements comprise overall the major fr action of CPM mass. The annual mass fr actions vary from 41.2% - 60.5% across the 10 sampling sites, with higher fr actions inland. Secondary ions (nitrate, nss-su lfate and ammonium) also contributed sig nificantly to the CPM mass, with an average ratio of 22.6 %. In particular, particulate nitrate accounted for about 17.2% of the total reconstructed mass. Carbonaceous compounds comprised a substantial portion of coarse particles. Organic matter contributed to an average of 19. 7%, while elemental carbon was a much less significant component at 1.1 %. Sea salt particles accounted for an average mass fr action of 9. 14%, with generally higher fr actions in spring and summer, due to the strong prevailing onshore winds in that period. The origin of organic carbon was investigated using linear regression analys is. Organic carbon was found to be well correlated with crustal materials and other trace elements. Higher correlation coefficients were fo und with soil dust tracers (R � 0.74 and 0.72 fo rT i and Fe, respectively), suggesting that OC and mineral dust either share common origins or that OC may be adsorbed or absorbed onto soil dust particles and collected simultaneously during PM sampling. 2. 6. Acknowledgements The work presented in this paper was funded by the Science to Achieve Results program of the United States En vironmental Protection Agency (EPA-G2006-ST AR-Ql). The authors would like to thank the staff at the Wisconsin State Laboratory of Hygiene and Mike Olson of University of Wisconsin, Madison for the assistance with chemical analys es. 36 Chapter 3 Seasonal and Spatial Variations of lndividual Organic Compounds of Coarse Particula te Matter in the Los Angeles Basin 3.1. Abstract To study the organic composition of ambient coarse particulate matter (CPM; 2.5-10 f!m), CPM were collected one day a week fr om April2008 to March 2009 at 10 sampling sites in the Los Angeles Basin. Samples were compiled into summer (June to September 2008) and winter (November 2008 to February 2009) composites, and were analyzed for individual organic constituents using gas chromatography-mass spec trometry. n-alkanoic acids and medium molecular weight (MW) n-alkanes (C25 to C35)- the ma jor constituents in the coarse size fr action- showed good associa tions with crustal materials. Polycyclic aromatic hydr ocarbons (PAHs) and hopanes (both in low concentra tions), as well as high MW n-alkanes (C37 and C38), were associated with traffi c-related emis sions. In the summer, when prevailing onshore winds were strong, the downwind/rural sites had higher concentrations of PAH s, n-alkanes and n-alkanoic acids. An opposite trend was observed at the urban sites, where the levels ofPAHs, n-alkanes and n-alkanoic acids were higher in the winter, when the wind speeds were low. In general, the contribution of organic compounds to CPM was higher in the winter, due to a reduction in the fr action of other CPM components and/ or the inc rease in source strengths of organic compou nds. The latter is consistent with the traff ic-induced re-s us pension of mineral and road dust, as previously observed in this basin. Overall, our results suggest that emissions fr om natural sources (soil and associated biota) constitute the majority of the organic content in coarse particles, with a more pronounced influence in the semi-rural/rural areas in Ri verside/Lancaster compared with urban Los Angeles in the summer. 37 3.2. Introduction Evidence has linked exposure to coarse particulate matter (CPM) to adverse health out comes, including hospital admissions and mortality (Brunekreef and Forsberg, 2005; Qiu et a!., 20 12). Investigators suggest that these ass ociations may be driven by specific chemical components (Castille jos et a!., 2000). Mineral dust, sea salt, secondary ions and organic matter are the dominant components of CPM. Organic matter (OM) contributed to �10-20% of the CPM gravimetric mass in most studied areas (Cheung et a!., 20 llb; Ho et a!., 2005). Despite its significant contribution, the organic composition of coarse particles has remained largely uncharacte rized. Organic content in the fine PM mode mostly originates fr om primary emissions, such as fo ssil fuel combustion (Fine et a!., 2004a), and secondary form ation by photochemical reactions (Do cherty et a!., 2008). Good correlations between organic carbon (OC) and elemental carbon (EC) were observed in studies conducted in the Los Angeles Basin measuring fine and ultrafine PM (Geller et a!., 2002; Sardar et a!., 2005), suggesting vehicular emission as a major source of OC. Yet, particles originating fr om combustion and secondary fo rmation mostly reside in the fme mode, with only a minor fr action intercalating into the coarse size fr action (Gertler et a!., 2000), thereby making their contribution to the OC measured in CPM rather unlikely. On the other hand, biological material, such as bacteria, viruses, spores, pollens, and plant debris, was shown to be a significant component ofPM10 in urban atmospheres (Bauer et a!., 2008; Samy et a!., 2011). Using scanning electron microscopy, Falkovich et a!. (2004) also demonstrated the adsorption of organics species onto mineral dust particles. Noneth eless, our understan ding on the sources and fo rmation mechanisms of CPM-bound organic species is still quite limited. 38 To study the phys ico-chemical properties of CPM, a comprehensive study was conducted at 10 sampling sites in the Los Angeles Basin fr om April2008 to March 2009. The chemical mass closure was reported in a previous publication of this study (Cheung et a!., 20 llb), and an overall low associ ation between OC and EC was reported (R �0.36), suggesting that the sources ofOC in the coarse mode might be different fr om-or in addition to tho se ofEC. This paper fo cuses on characterizing individual organic compounds in CPM. The examination of the spatial and seasonal variations of these constituents further elucidates their sources and form ation mechanism. This information is vital in understanding the role of organic compounds in CPM-induced toxicity, and may help the regulatory community to desigu more effective strategies to control the emissions and ambient levels of these particles. 3.3. Methodology Details of the sampling location, equipment and methods are described in the previous chapter (Chapter 2). 3.3.1. Chemical analysis The quartz substrates were analyzed for organic carbon ( OC), elemental carbon (EC), and water-soluble inorganic ions, while the Zefluor substrates were used for the analys is of total elemen ts. The quantification of these species allowed the chemical mass reconstruction of the sampled coarse particles, the results of which were reported and discussed in detail in a previous publication of this study (Cheung et a!., 20l lb). The quantification of particulate organic compounds -i ncluding n-alkanes, organic acids, hopanes, steranes and PAHs - constitutes the core of this article. These organic molecular markers were selected because they have been traditionally used for source characterization. To measure the level of individual organic species, 39 quartz substrates were composited for two seasons: (1) summer, fr om June to September 2008 and (2) winter, from November 2008 to February 2009. The overall average CPM mass concentrations were 13 (±4.6) and 7.6 (±5.0) flg/m3 for the summer and winter, resp ectively. Samples were extracted in a 50/50 dichloromethane/acetone solution using Soxhlets, fo llowed by rotary evaporation and volume reduction under high-purity nitrogen, until the final volume of 100 f!l was reached for each sample. The samples, along with field blanks, laboratory blanks, filter spikes, and check standards, were subseq uently analyzed for organic composition using gas chromatography-mass spec trometry (GCMS) (Sheesley et al., 2004). Total uncertainties were propagated using the square root of the sum of squares method based on the analytical uncertainties and uncertainties fr om field blanks, which account for the uncertainties arising fr om field handling, storage, shipping and analytical instruments . Detection limit is def med as two times the total uncertainty in this study. Concentrations were reported after field blank subtraction to account for the background levels of the measured species. 3.4. Results and Discussion 3.4.1. Overview The meteorological conditions during the sampling campaign are presented in Chapter 2. In brief, the subtropical Mediterranean climate of the Los Angeles Basin was highlighted by a warmer summer ( avg. temperature�23±3 .1 'C) and a cooler winter ( avg. temperature� 14±2.3 'C). The diurnal variation of temperature could also be substantial in this basin, and the difference between the highest and lowest temperature in a day exceeded lO'C in the surnrner and winter. Westerly onshore winds were prevalent in the surnrner (ca. 0.75-3.8 rn/s), resulting in the long-range transport of atmospheric poll utants fr om coast to inland. Wind speeds were lower in 40 the winter (ca. 0.17-1.4 m/s ), with a more diverse wind direction. Seasonal fluctuations of meteorological parameters at the inland sites (Ri verside and Lancaster) were more pronounced, as highlighted by the more extreme oscillations of temperature. Table 3.1 shows the concentrations (ng/m3) of organic compound class at the 10 sampling sites in the summer and winter. PAHs are segregated into 4 categories according to their molecular weights : MW�202 g/mole [fluoranthene, acephenanthrylene, pyre ne], MW�228 g/mole [benz( a )anthracene, chrysene ], MW�252 g/mole [Benzo(b )fluoranthene, benzo (k)-fluoranthene, benzo(j )fluoranthene, benzo(e)pyrene and benzo(a) pyrene, perylene] and MW2>276 g/mole [Inden o(l ,2,3-cd)pyrene, benzo(ghi)perylene, dibenz(ah)anthracene, picene, coronene, and dibenzo(ae)pyre ne]. The overall concentrations ofPAHs were low, with mostly undetectable levels. Hopanes and stera nes, both tracers of vehicular emissions, were also present in low concentrations ( < 1 ng/ m'). The group of n-alkanes was calculated as the sum of n-alkanes analyzed in this study, ranging fr om Cll to C38. The levels of Cll to Cl9 were below detection limits for all sites /seasons, likely due to the semi -volatile nature of these light alkanes (Schne lle-Kreis et al., 2005).T otal n-alkanes levels were significant at all samplings sites (� 3.5-15 ng/ m3). The group of n-alkanoic acids (sum of Cl5 to C30) is the most prominent compound class in this study, with concentrations ranging from 9.3-64.2 ng/m3 Levels under the detection limit were treated as zero in these sums. The level of detection (LOD) of the sum of a group of compounds was calculated as 2 times the square root of the sum of squares of the uncertainties for each species in the limit as the concentration of the species approaches zero. The spatial and seasonal variations of these compound classes are discus sed in detail in the fo llowing sections. 41 Table 3-1: Summary table of organic component concentration (ng/m3) at the 10 sampling sites in: (a) summer and (b) winter. BDL denotes below detection limit. I 1,1 HUO GRO LOS CCL 'L PAHs, MN-202 BOL BOL BOL BOL 2L PAHs, MN=228 0 081 BOL BOL BOL 3 L PAHs, MN=252 0 095 BOL BOL BOL '1: PAHs, MN�276 BOL BOL BOL BOL 5L Hopanes 0 10 BOL 0 041 0 11 "L Steranes BOL BOL BOL BOL L n-Aikanes 35 48 53 40 L n-Aikano1c ac1ds 26 9 20 3 20 4 19 1 b HUO GRO LOS CCL 'L PAHs, MW=202 0 092 BDL BDL BDL 2L PAHs, MW=228 0 12 BDL BDL BDL 3 L PAHs, MW=252 0 16 BDL BDL 0 14 '1: PAHs, MV\12276 0 069 BDL BDL BDL 5L Hopanes 0 81 0 17 0 18 0 28 "L Steranes 0 28 BDL BDL 0 045 L n-Aikanes 14 6 8 8 8 0 8 8 T n-Aikano1c ac1ds 31 1 21 2 26 4 53 4 1 The level of detection for L PAHs, MW=202 1s 0 055 ng/m 3 2 The level of detection for L PAHs, MW=228 1s 0 038 ng/m 3 3 The level of detection for L PAHs, MW=252 1s 0 051 ng/m 3 4 The level of detection for L PAHs, M\f\12276 1s 0 061 ng/m 3 5 The level of detection for L Hopanes 1s 0 031 ng/m 3 6 The level of detection for L Steranes 1s 0 023 ng/m 3 3.4.2. PARs usc HMS FRE VBR GRA LAN BOL BOL BOL BOL BOL BOL BOL BOL 0 11 0 083 BOL BOL BOL BOL 0 20 0 10 0 084 BOL BOL BOL BOL BOL BOL BOL 0 18 0 12 0 067 0 065 0 12 0 22 BOL BOL BOL BOL BOL 0 054 8 1 49 79 10 8 10 9 87 35 0 24 9 26 1 64 2 53 3 26 6 usc HMS FRE VBR GRA LAN BDL BDL BDL BDL BDL BDL BDL BDL 0 086 0 041 BDL BDL BDL BDL 0 11 BDL BDL BDL BDL BDL BDL BDL BDL BDL 0 28 0 24 0 087 0 18 0 18 0 088 BDL 0 043 BDL BDL BDL BDL 7 4 10 3 7 0 10 0 6 8 4 6 32 8 26 3 24 3 40 4 23 3 8 3 Mounting evidence has documented adverse health effects caused by PARs (Schwarze et al., 2007; Sram et al., 1999). Particulate phase PARs are mostly sub-micron particles originating from combustion emissions and gas-to-particle partitioning (Miguel et al., 2004). 1n the coarse size fr action, PARs are likely to originate either fr om the debris of tire abrasion (Rogge et al., 1993a), and/or the condensation of semi-volatile species on CPM surfac es. Figure 3.1 shows the levels of PA Rs, as well as the LOD, among the 10 sampling sites, segregated by MW. In the summer, cbrysene, which could be fo und in both industrial and vehicular emissions (Khalili et al., 1995), contributed to the sum of PAR with MW�228 g/m ole, with the highest concentrations at FRE (0. 11 ng/m3 ), where it is �50 m north of Freeway I-10. 1n Riverside (a receptor area particularly during the summer when onshore winds were strong), PARs with MW�228 g/m ole and MW�252 42 g/mole had generally higher levels in the summer than in the winter, a fmding consistent with an earlier study conducted in Claremont, CA, a downwind receptor location, where elevated levels of PAHs were observed in the coarse mode in the spring and summer (Miguel et al., 2004). In the winter, fluoranthene and pyrene (both of which are semi-volatile PAHs that partition between gas and particle phase) contributed to the sum oflow MW PAHs (MW=202 g/mole) in the Long Beach site (HUD). Among the 10 sampling sites, HUD and FRE, both in close proximity to traffic (within 100 m from nearby highway/freeway), had the highest overall PAH levels. c:::::J Coo eerira tiM , Summ« c::::J Concentration, Wilter ···· ·· LOD 0 Noonaized to OC. SUmmer 6 Norrnaized to OC, VVinter 0.14 (a) • Normalzed to PM, SUmmer ... Nonnaized to PM. Wnter 0.16 0.12 0.14 u 0 � .E 0.10 0.12 ::;: Q. 0 "' .s 0.08 � .. 0.10 "' B 0.08 .[ � 0.06 � c .Q 0.06 .. tr 0 0.04 u . . . ... . ... -o 0.04 .� Iii E 0.02 0.02 0 z 0.00 0.00 HUD GRD LOS CCL USC HMS FRE VBR GRA LAN 0.25 0.25 (b) 0 u 0 0.20 0.20 ::;: .r Q. E 0 0, "' .s 0.15 0.15 B c "' 0 ., §. !" 0 c 0 .1 0 0.10 � g tr 0 -o u ., .'i 0.05 0.05 Iii E 0 z 0.00 0.00 HUD GRD LOS CCL USC HMS FRE VBR GRA LAN Figure 3-1: Concentration (ng/m3) ofPAHs with (a) MW = 228 g/mole and (b) MW = 252 g/mo le. The level of detection (LOD) of the compound class is shown as a reference line. 43 3.4.3. Hopanes and steranes Hopanes and steranes are indicators of fo ssil fuel combustion (Simoneit, 1999). In particular, they are known to be emitted from gasoline and diesel vehicles primarily through the use of engine lubricating oil, and are widely used as the organic markers of vehicular emis sions in urban environments (Fine et a!., 2004a). Wang et a!. (2009) showed that the maj ority of hopanes resided in the fme PM mode, and lower hopane levels were observed in the coarse size range. Thus, the low hopane level (avg. �0.18 ±0. 17 ng/m3) observed in this study is consistent with the published literature. In the summer, the average hopane concentration was 0. 10±0.052 ng/m3 across the 10 sampling sites . The highest hopane level was fo und in Lancaster (Figure 3.2 ), which was unexpected due to its remote location. It is possible that road dust is erniched by anth ropogenic particles over time in that area, particularly in the summer when atmospheric transport is strong. In the winter, hopane levels were higher, with an average of0.25±0.21 ng/m3 The high standard deviation was driven by the peak at the Long Beach site (HUD), which coincides with the high level of EC in the same period. This site is located � 2 krn fr om the Port of Los Angeles and the Port of Long Beach, and is influenced by port-related em issions. Therefore, the high EC and hopane levels in the winter may be attributable to emissions fr om heavy-duty diesel vehicles and, to a lesser extent, marine vessels. These emissions could accumulate around the harbor area when atmospheric dilution was low. Furthermore, with the exception of Lancaster, the levels of hopanes were all elevated in the winter. The normalized ratios (to PM) were also much higher in the winter (ca. 3 times), suggesting stronger source strengths of vehicular emissions during that period. Considering the relative consistency of traffic volume and composition in the Los Angeles 44 Basin throughout the year, the elevated hopane levels are likely due to the lower atmospheric mixing and I or the increased re-suspension by traffic during cooler periods (Cheung et al., 2012). Concentrati on, Summer c:::::::J Concentrati on. W1nter ........ LOD 10 0 Normalized to OC, Summer 06 /',. Normalized to OC. 'Mnter • Normalized to PM, Summer 0 u /',. ... Normalized to PM. 'Mnter 0.5 0 0.8 - :::;; 1 04 c.. .... 0 C) C) -=.. 06 .B c: C) 0 ·� /',. 03 s c 0 /',. 0 "' 0.4 0 "' iii 0 "' /',. 0:: c: /',. 0.2 0 "'CJ u "' /',. Q) .!::! 02 "'iii 01 E 0 z 00 00 HUD GRD LOS CCL USC HMS FRE VBR GRA LAN Figure 3-2: Concentration (ng/m3) of sum of hopanes . The level of detection (LOD) of the compound class is shown as a refe rence line. Table 3.2 shows the correlation coefficients among individual organic component, compound class, OC, EC, selected elements and meteorological paramet ers. The method of quantification of OC, EC and total elements, as well as their results were described in two previous publications produced from this study (Cheung et al., 2011b; Pakbin et al., 2011), and they are not discussed in detail here. Overall, the high correlation (R=0.91, p<0.001) between hopanes and EC, a tracer of vehicular emission, confirms traffic as the dominant source of hopanes in CPM. Although the levels of steranes were much lower, with most of the data points below detectable levels (Table 3. 1), their results were consistent with those observed for hopanes, with more detectable concentrations and a significant peak at the HUD site in the winter. 45 Table 3-2: Correlation coefficients (R) among individual organic component, compound class, organic carbon (OC), elemental carbon (EC), selected elements and meteorological paramet ers. Species with less than half data points with detected levels are excluded in this analys is. 'i_A ik<l1oic He�Xatriac Octadec a Triacoota Acid, C15· Tricos..-.e Nonacosa !jane, nOC acid, nOC acid, Ann Lo!<>;jo: Rel<i:i'l'e T"""'� wm LA II<a1es C30 , C23 ne, C29 C37 C18 C>J acid oc EC AI - n cu Zn Ba l"urK:lt y �eed �1--q:lanes 066 0.16 0.42 038 058 -0_07 -0.14 -0.13 O>J 0.45 091 0.30 0.40 0.24 073 0 . 52 -0.29 -0 )J -0.25 LAkanes 054 0.71 0.84 0.67 0.42 0.45 0 .47 O>J 0.73 0.63 0.59 0.72 0.14 040 0.48 -0.35 -0.04 -0.07 IAkardc Acid, C15-C30 038 0.78 022 089 075 0.71 048 0.74 011 0.68 071 -0.11 019 032 -0.05 038 03< Tricosane, C23 0.64 0.45 0.39 0.31 0.47 0.25 0.64 0.41 0.43 0.56 0.23 0.45 047 -0.36 -0 .01 O.oJ �cosaoe> C29 O>J 0.75 078 000 023 0.78 033 083 089 0.01 025 0 . 27 -0.38 028 0.23 Hertatriacd:ane, C37 000 000 -0.03 057 056 065 -0.06 013 0.18 038 068 -0.13 -054 -0.36 OctadecaOOc acid, C18 0.89 0.81 0.12 0.59 -0.15 0.73 0.66 -0.37 -0.11 -0.10 O.oJ 0.52 0.46 TriacOO:arx)[: ocid, C30 0.87 -0.03 0.59 -0.12 0.69 062 -0.44 -0.13 -014 0.03 041 036 Pinonic acid -0.08 0.56 -0 .18 0.73 0.77 -0.11 -0.14 0 . 04 -0.26 0.53 0.49 Levogtx:osan 056 0.42 -0.18 001 0.38 O:M 0.67 -0.19 -0 .49 -0_39 oc 0.55 0.47 0.62 -0.03 0.48 0.63 -0.07 -0.03 -0.04 EC 0.11 024 0.15 0 74 060 -0.18 -0.53 -0.51 AI 0 .92 -0.09 0.17 -0.06 -0.26 0.59 0.45 Ti 0.16 0 27 0.26 -0.45 048 0.43 cu 0.10 054 -0_66 -0 09 -0.09 Zn 0.50 -0.07 -0.18 -0.18 sa -0.27 -032 -0.2'5 Rel<i:Weh i.Xlidily 0.06 -0.17 Terrperattxe 0.82 3.4.4. n-Alkanes The group of total n-alkanes is a predominant compound class in CPM, with an average of 6.9±2.8 ng/m3 and 8.7±2.7 ng/m3 in the summer and winter, respec tively. At the three inland sites (VBR, GRAand LAN), total n-alkanes concentrations were higher in the summer (Figure 3.3). This is probably due to the "receptor" location of these areas, where po llutants are transported fr om the upwind locations. These inland sites are also more rural, with more vegetation and open fields that could be biogenic sources of these species. In contrast, higher n-all-cane concentrations were observed in the winter in the urban areas, namely Long Beach (IDJD), West L.A. (GRD and LDS), and to a lesser extent Central and East L.A. (CCL and HMS). These elevated levels could be due to the lower atmospheric transport in the winter that limited the dilution of total n-alkanes, and/or the increase in the source strengths of these species in the winter. The carbon preference index (CPI), calculated using the wide range ofn -alkanes (LC13-C35/LC12 -C34), was used to estimate the sources of these n-alkanes in coarse mode aer osols. Emissions fr om fo ssil fuel exhibit a CPI near unity, whereas a CPI greater than 2 indicates a dominance of biogenic sources 46 (Daher et a!., 2011). The average CPI in this study was 2.1±0.75, suggesting an overall prevalence of biogenic origins for these n-alka nes. To further investigate the origins of n-alkane s, the relative mass concentration (percent of sum of n-alkane s) is calculated for Cl9-C38 analyzed in this study, as presented in Figure 3.4. In general, peaks were observed for C29 and C31, to a lesser extent C23, C37 and C38 in both summer and winter. These distributions are compared to the source profiles of PM garden soil, road dust, brake wear, leaf abrasion products and total suspended tire wear, developed by Rogge et a!. in the 1990s (Rogge et a!., 1993a, b). The most dominant peaks of C29 and C3 1 in CPM are similar to the profile of soil, leaf abrasion (both dead and green) and road dust, suggesting that these peaks originate mainly fr om mineral and road dust. This is confirmed by the high correlation between C29 and AI, a soil dust tracer (R �0.83, p<O. OO l). Furthermore, the two Rivers ide sites (VBR and GRA) experienced higher levels of C3 1 than C29, which is a signature of soil and leaf abrasion (as opposed to road dust, which has a higher peak at C29 than C3 1 ), highlighting the suburban nature of the River side area. On the other hand, peaks in C37 and C38 are consistent with the profile of tire wear, which dominates the higher molecular weight n-alkan es. Although vegetative detritus is also a source of heavy alkanes (Zygadlo et a!., 1994), the moderate correlations ofC37 with EC (R�0 .65, p�0.004) and Ba (R�0.68, p�0.002) and its insignif icant associ ations with AI (R�-0.06, p�0.81) and Ti (R�O .l3, p�0.60) further confirm tire wear and road dust as the dominant sources ofC37 and C38. 47 l 5 1::: 0 � "E .. u 1::: 0 u - Concentrat ion, Sumrrer [:::::J Co nee ntrati o n \!Vinte r 16 - .. .... LO D 14 12 10 0 Normalized to OC, Summer C:::.. Normalized to OC, Winter • Normalized to PM, Summer .& Normalized to PM, Winter 0 6 HUO GR D LOS CCL USC HMS FRE VBR GRA LAN 25 G' 0 20 - :2 a.. '0 15 Cl .B Cl .s 0 10 � 0::: "0 .. .!>! 5 iii E 0 z Figure 3-3: Concentration (ng/m3) of sum of n-alka nes. The level of detection (LOD) of the compound class is shown as a refe rence line. The spatial and seasonal patterns of C37 and C38 were diff erent fr om those of n-alkanes (Figure 3.3), C29 and C3 1. In contrast to the significant spatial and seasonal patterns among the 10 sampling sites for n-alkanes, C29 and C3 1 (higher concentrations at the inland sites in the summer and higher concentrations at the urban sites in the winter), most sampling sites displayed less variation for C37 and C38, as highlighted by the low standard deviations (<15%) of both species acro ss the 10 sampling sites in both seasons. This suggests that source strengths of these two species were relatively stable throughout the basin and during different seasons, consistent with the relatively low variability of traffic volume and composition- the source of tire wear emissions in this basin. Both C37 and C38 experienced slig htly elevated levels in the winter (avg.=0.78±0.059 ng/m3 and 0.91±0.058 ng/m3 respectively) compared to summer (avg.=0.65±0.082 ng/m3 and 0.77±0.098 ng/m3 respectivel y). As discussed earlier, traffi c-induced re-suspension of coarse particles is more dominant in the winter, which might explain the higher levels of these tire wear particles in the cooler period. Note the significant levels of C23, and to a lesser extent C22 and C21, in our study that do not fit the source profiles of exhaust emission, 48 cooking, tire and brake wear, road dust, garden soil or leaf abrasion (Rogge et al., 1993a, b, 1998; Rogge et al., 1991), which are mostly developed using fine particles, with the exception of tire wear from total suspended particl es. C23 , tricosane, could be a component of OC that is more dominant in CPM than in fi ne particles. Its moderate correlations with AI, Ti, Zn and Ba (R range from 0.43-0. 56) suggest that it has both natural and anthropogenic origins . 30 «> ( a) (') u 25 a, u "' 20 " ,.-- --------------------1 c:::::J HUD f-------, c:::::J GRD -L OS c:::::J CCL c:::::J usc c:::::J HMS c:::::J FRE ..!:! -V BR <( 15 -G RA .:: -LA N 0 10 E " (J) 0 5 � 0 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 n·Aikane Carbon Number �----------------------------� � u 25 a, u "' 20 " J:! <( 15 c 0 E 10 " (J) 0 5 ( b) 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 n-Aikane Carbon Number Figure 3-4: Concentration distribution (per sum of n-alkane s, C19-C38) of coarse particulate matter for n-alkanes in: (a) summer and (b) winter. 3.4.5. n-Alkanoic acids The n-alkano ic acids group is the most abundant organic compound class identifi ed in CPM at all sampling sites, with average levels of 32± 16 ng/m3 and 29± 12 ng/m3 in the summer and winter, respectively. In the more downwind/ru ral locations (Rivers ide and Lancaster), levels of 49 n-alkanoic acids were higher in the summer (Figure 3.5). Their normalized ratios (to PM) are also generally higher than other sites, suggesting the source strengths of n-alkanoic acids were stronger in the downwind areas in summertime. In the winter, when long-range transport is limited due to lower wind speeds, n-alkanoic acid levels were either similar to or higher than their summer levels at the urban sites. When normalized to OC (shown as open symbols in Figure 3.5), the ratios in the winter are either similar to or lower than those in the summer. It indicates that the relative contribution of n-alkanoic acids to OC concentration was generally lower in the winter. The CPI of n-alkanoic acids, calculated as the IC24-C30/IC23-C29, was higher in the summer (avg.=2. 3± 1.1) than winter (avg.=2.0±0.58), particularly at the two Riverside sites in the summer, where the CPI was 4.7 and 3.8 at VBR and GRA, respec tively. "' .€ "' s <:: 0 � -= <:: "' 0 <:: 0 () c::::::J Concentration, Summer c:::::J Concentration, Winter ........ LOD 0 Normalized to OC, Summer 70 t:. Normalized to OC, W1nter • Normalized to PM, Summer 60 • Normalized to PM, Wnter 50 0 0 0 40 0 30 20 0 2i 0 t:. 60 40 20 0 0 - :2 c.. .... 0 "' .2 "' 5 0 � et:: "CJ "' .� 'iii E 1: + . _ .. . . ... . '1-- .. L .. _ ... .,.. ... _ ... -"'f- .. L .. -"' . J-. L .. ..., · J-. L ... -"'!-- .. L .. .q_ .. L .. � .. L .. .I.Ij-J .. L ... "'f-' .. - ... ...., . .. , .. . _ ... "'1--' U-"' -+ ". o :!t HUD GRD LDS CCL USC HMS FRE VBR GRA LAN Figure 3-5: Concentration (ng/m3) of n-alkanoic acids, as a sum of C15-C30. The level of detection (LOD) of the compound class is shown as a reference line. Figure 3.6 shows the distribution of relative mass concentrations (percent of sum of n-alkanoic acids, C15 to C30). The dominant peaks in hexadecanoic acid and octadecanoic acid (C16 and C18) are consistent with the source profiles of tire wear, PM road dust and green leaf 50 debris (Rogge et a!., 1993a, b). Note that Cl6 and Cl8 are the most common fatty acids present in biological matter (Simoneit, 1986), and they may also arise from many other sources such as fireplace use (Dutton et a!., 2010), fo od cooking (Schauer et a!., 2002) and fo ssil fuel combustion (Cheung et a!., 2010). The high assoc iation between Cl8 and AI (R �0.73, p<O.OO l) coupled with the insignif icant associ ations with Ba (R�-0.10, p�0.69), EC (R�- 0.15, p�0.53) and levoglucosan (to be discussed later, R� O. l2, p�0.64) suggest its dominant origin as soil dust as opposed to traffic-related emissions and biomass burning. The source profiles of soil and dead leaf abrasion showed dominance within the higher MW n-alkanoic acid range (2>C22) (Rogge et a!., 1993b ). Using our data, C30, a heavier alkanoic acid that is predominant in the wax filaments of plants (Yue and Fraser, 2004), also demonstrated good ass ociations with AI and Ti (R �0.69, p<O.OOl and R�0.62, p�0.004 respectively). This finding is consistent with some studies that demonstrate the presence of biological materials in the urban/suburban aero sols ofPM10 (Jia et a!., 2010). In particular, Cahill et a!. (2010) showed that sugars (including glucose, sucrose and fr ucto se) dominated the coarse size fr action of PM in particles collected in Califo rnia's Central Valley, and attributed its sources to biological materials of plant det ritus, spores, etc. On the other hand, wind speed may affe ct the levels of some CPM constituents, such as mineral dust tracers of AI and Ti (R �0.45 and 0.43 respectively; p�0.045 and 0.057 respectively), and to a lesser extent n-alkanoic acids (R�0 .35, p� O.l3). Similar degrees of associ ations are observed when the analys is is done using the data fr om the summer only. The moderately low ass ociations are likely due to the supp ressed effects of meteorology in 24-hour samples, as well as their differential contributions to urban and rural areas (Cheung et a!., 2011). In particular, wind speed could be an important parameter of CPM levels in rural areas, where local emis sions and 51 re-suspension of dust by anthropogenic activities are minor, as evident by the stronger associations between wind speed and Al (R=0.86; p=0.026) when the correlation analy sis is restricted to the three inland sites (GRA, VBR and LAN). The higher wind speed in the summer increased both local re-suspension and possibly long-range transport of upwind air pollutants to these sites, contributing to the elevated conc entrations of these constituents in the summer. Overall, the good associations between n-alkanoic acids and soil dust tracers (R=0.68, p=O.OO l for Al; R=0 .71 , p<O.OO l forTi) suggest that re-suspended dust contains a mixture of crustal materials and vegetative debris, and that mineral dust is enriched with biological matter in coarse particles sampled in Southern California. "' "' u 60 (a) d, u -o " 50 ·o -HUD -GRD -LOS <( 40 0 · c; -CCL -usc c J:l 30 <( c 0 20 E -HMS -FRE -VBR -GRA -LA N " 10 (j) 0 ;ft 0 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 n-Aikanoic Acid Carbon Number "' "' u 60 (b) d, u -o " 50 ·o <( 40 0 · c; c J:l 30 <( c 0 20 E " 10 (j) 0 ;ft 0 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 n-Aikanoic Acid Carbon Number Figure 3-6: Concentration distribution (per sum of n-alkanoic acids, C 15-C30) of coarse particulate matter for n-alkanoic acid in: (a) summer and (b) winter. 52 3.4.6. Other carboxylic acids Figure 3.7 (a-c) shows the levels of pinonic acid, palmitoleic acid and oleic acid respec tively. Pinonic acid, formed primarily fr om the oxidation of pinenes, is considered a biogenic secondary organic aero sol (Cheng et a!., 2011). Pinonic acid had slig htly higher concentrations in the summer (avg. � 0.68±0.35 ng/m3) than in the winter (avg. �0.56±0. 12 ng/m3 ), consistent with a recent study conducted in Europe, which showed higher levels of pinonic acid in coarse mode aerosols during spring and summer than autumn and winter (Zhang et a!., 2010). In particular, high concentrations of pinonic acid were fo und in the receptor and rural sites in the summer, when temperatures are higher and the fo rmation of secondary aerosol is more dominant in inland areas (Sa rdar et a!., 2005). The high assoc iations between pinonic acid and crustal material tracers of AI and Ti (R �0.73 and 0.77, respectively; p<O.OO l for both) suggest that although pinonic acid arises from a different fo rmation mechanism with crustal materi als, it might be adsorbed onto or absorbed into soil dust and collected simultaneously during sampling. Palmitoleic acid, on the other hand, could be fo und in marine aero sols (Sirnoneit et a!., 2004) and fo od cooking (Rogge et a!., 1991), and it displayed a distinct seasonal variation. The higher summer concentrations (avg. �0.60±0. 14 ng/m3 ), which parallel the strong onshore winds that transport sea salt particles fr om the coast to inland, suggest that marine aero sols are the primary source of palmitoleic acid in CPM in this basin. Oleic acid experienced higher levels in the winter than in the summer and a peak level at CCL in the winter. This obse rvation is consistent with an earlier study conducted in the basin, which revealed winter maxima and extended summer minima of oleic acid, and the investigators attributed the discrepancy to the enhanced photo-oxidation of this species during summer months (Rogge et a!., 1993c). High level of oleic 53 acid (�1100 11g per g of sample) was previously reported in the total suspended debris of tire wear (Rogge et al. 1993a). The same group of investigators also fo und significant levels of oleic acid in road dust, vegetative detritus, and emissions fr om wood burning (Rogge et al., 1993a, b, 1998). Therefore, these could be the potential sources of oleic acid in coarse mode aerosols. 1.4 20 1.0 ,...-- ------------- --. 2. 0 (I) 1.2 � 1.0 .5. 0.8 B � "' o.e 1l c 0 8 o .• 0.2 0.0 HUO -=:I Concentration, Summ-er (b ) � Concentration, Winter G" LOO � 0 NormaliZed to OC, Summer 08 f). Normalized to OC, INinter f). 15 ::!: • Normalized to PM. Summer Q. ...... 0 � 0 .. Normalized to PM, Winter "' g .5. 0.6 c 0 10 E � 0 B � 8 o• ce c al 8 05 .� �· � 02 0 .... . . I· z 00 0.0 GRO LOS CCL USC HMS FRE VBR GRA W< HUO GRD LOS CCl USC ...... s FRE VBR GRA (C) f). HUO GR O LOS CCL USC HMS FRE VBR GRA l..MI g • i Q. 0 3 "' g E 2 0 "' � "' al ,l:l 1 § 0 z f). 1.5 � 0 0 "' g 1.0 r W< Figure 3-7: Concentration (ng/m3) of: (a) pinonic acid, (b) palmitoleic acid and (c) oleic acid. The level of detection (LOD) of each organic compound is shown as a reference line. 3.4. 7. Levoglucosan Levoglucosan, a pyrolysis product of cellulose, is widely used as a ma jor marker of biomass burning (Simoneit, 1999). It could be emitted fr om various wood combustion (Hays et al., 2011), cigarette smoke (Kleeman et al., 2008), as well as wildfires (Yue and Fraser, 2004) or management bums (Zhang et al., 2008a). Fine et al. (2004a) showed higher levels of 54 levoglucosan in the accumulation mode particles at the USC sampling site in the winter (�47 ng/m3) than in the summer (�4.2 ng/m3 ), and attributed the difference to the increased residential wood combustion activity during wintertime. In this study, the overall concentration of levoglucosan was low (avg.�3.3 ng/m3 ), consistent with its combustion origins, which yield only a minor fr action of particles in the coarse mode (Herckes et a!., 2006). Figure 3.8 shows the concentrations oflevoglucosan across the 10 sampling sites in the two seasons. Levoglucosan concentration was siguificantly higher in the winter ( 4.9±3.0 ng/m3) than the summer (1.8±0.41 ng/m3 ), paralleled with the seasonal variation of water-soluble potassium another tracer of wood smoke (Chow et a!., 2007) - in this study, and is consistent with the higher wood burning activity in the cooler period. The higher normalized ratios in the winter at most sites indicate that the higher contribution of levoglucosan to both OC/PM in cooler mont hs. The moderate correlations between levo glucosan and CPM of anth ropogenic origins including tracers of tire and brake wear (R�0.67 and 0.38 for Ba and Cu respectively), as well as vehicular emissions (R �0.42 for EC) are unexpected due to their distinct origins and fo rmation mechanis ms. Note that high correlations between species do not imply common origins due to possible confou nding and CPM re-suspension of a common primary mechanism. The high assoc iation between levo glucosan and EC could be also driven by their seasonal pattern (higher levels in the winter than the summe r). The higher levels of coarse particulate levogluc osan in the winter, driven by inc reased wood burning activiti es, might accumulate in the mineral and road dust, and be re-suspended together with Ba, Cu or EC by vehicular-induced turbulence in the cooler period. 55 - Concentration. Sun mer c::::J Concentration. Winter 14 ,---------------1 '"''" LOD 0 Normalized to OC, Summer 12 10 6 Normalized to OC, Winter • Normalized to PM, Summer 6 ... Normalized to PM, Wirter 6 6 6 6 0 0 a 6 6 0 HUD GRD LOS CCL USC HMS FRE VBR GRA LAN 10 u 0 8 - :;;; CL 0 0) 2 0) _§_ 0 4 "" "' '0 <l> !;:! 2 "' E 0 z Figure 3-8: Concentration (ng/m3) of levoglucosan. The level of detection (LOD) oflevoglucosan is shown as a refer ence line. 3. 5. Summary and Conclus ions In summary, n-alkanoic acids, to a lesser extent n-alkanes, are the predominant classes of organic compounds in CPM. Constituents of vehicular emissions, including PAH s, hopanes and steranes experienced low concentrations in the coarse mode. The profiles of n-alkanoic acids are similar to those fr om leaf abrasion, garden soil, tire wear and road dust. Coupled with the high correlations with A1 and Ti, our results suggest that n-alkanoic acids in CPM are mostly derived fr om mineral and road dust. On the other hand, n-alkanes in the coarse size fr action originate fr om both biogenic and ant hropogenic sourc es. Medium MW n-alkanes (C29, C31) highly correlated with A1 and Ti, while high MW n-alkanes (C37, C38) moderately correlated with Cu and Ba. Previous studies conducted in the Los Angeles Basin showed that vehicular emissions and secondary photo-oxidation are the ma jor sources of organic compounds in the fine and ultrafine modes (Fine et al., 2004a). Based on the analys is conducted using only samples of ambient coarse particles, soil and the associated biota represent the predominant source of organic constituents in CPM. In general, the mass fr action of organic compounds to CPM was higher during wintertime, 56 due to lower concentrations of other components (sea salt, secondary ions, etc.) and/or the increa se in source strengths of organic compounds, the latter of which is supported by traffi c-induced re-s us pension of mineral and road dust in cooler mont hs. The distinct organic composition of fine and coarse PM might have important implications on the chemical mass reconstruction of PM. Higher fractions of unidentified mass fr action are usually observed for coarse particles compared with fine and ultrafine PM in ambient enviro rnn ents (Ho et a!., 2005). Organic matter from soil and the associated biota, which could contain sugars and amino acids that have a higher molecular weight per carbon weight ratio, might be underestimated by the typ ical conve rsion factor (rang ing fr om 1.4 to 2. 1) used to estimate the average molecular weight per carbon weight for the organic aerosol (Turpin and Lim, 2001). The dominance of biological materials in CPM suggests that a biogenic-specific conve rsion factor needs to be deve loped for coarse particles to improve closure between gravimetric and chemical measurements of coarse particles. 3. 6. Acknowledgements The work presented here is fu nded by the United States Environmental Protection Agency through the Science to Achieve Results program (EPA-G 2006-STAR-Q l). We would like to thank the staff at the Wisconsin State Laboratory of Hygiene and Mike Olson of University of Wisconsin, Madison for the assist ance with chemical analys es. Although the research described in this article has been funded wholly or in part by the United States Environmental Protection Agency, it has not been sub jected to the Agency 's required peer and policy review and therefore does not necessarily reflect the views of the Agency and no official endorsement should be infer red. 57 Chapter 4 Historical Trends in the Mass and Chemical Species Conc entrations of Coarse Particula te Matter (PM) in the Los Angeles Basin and Relation to Sources and Air Quality Regulations 4.1. Abstract To assess the impact of past, current and proposed air quality regulations on coarse particulate matter ( CPM), the concentrations of CPM mass and its chemical constituents were examined in the Los Angeles Basin fr om 1986 to 2009 using PM data acquired fr om peer-reviewed journals and regulatory agency data base. PM10 mass levels decreased by approximately half from 1988 to 2009 at the three sampling sites examined- located in downtown Los Angeles, Long Beach and Rivers ide. Annual CPM mass concentrations were calculated fr om the difference between daily PM10 and PM 2 5 fr om 1999 to 2009. High CPM episodes driven by high wind speed I stagnant condition caused year-to-year fluctuations in the 99 t h I 98 1h percentile CPM levels. The reductions of average CPM levels were lower than those ofPM 2 _5 in the same period, despite their comparable concentra tions; therefore the decrease of PM10 level was mainly driven by reductions in the emission levels of PM 2 _5 (or fine) part icles. This is further conf mned by the significant decrease of Ni, Cr, V, and EC in the coarse fr action after 1995. On the other hand, the levels of several inorganic ions (sulf ate, chloride and to a lesser extent nitrate) remained comparab le. From 1995 to 2008, levels of Cu, a tracer of brake wear, either remained similar or decreased at a smaller rate compared with elements of combustion origins. This differential reduction of CPM components suggests that past and current regulations may have been more effective in reducing fugitive dust (AI, Fe and Si) and combustion emissions (Ni, Cr, V, and EC) rather than CPM fr om vehicular abrasion (Cu) and inorganic ions (NO;, SO/- and Cr) in urban areas . 58 4.2. Introduction In the last decade, a significant body of new research has documented substantive differ ences of fine and coarse particles, and there is a growing evidence of health effe cts induced by CPM (Chen et a!., 2005; Yeatts et a!., 2007). As a result, the need for a new PM10_ 2 _5 standard was proposed in a recent review (2006) of NAAQS (U.S.EPA, 2006). The proposed PM10_ 2 _5 standard was suggested to be 65-85 flg/m3, in the form of a 98 1h percentile, and it was intended to be generally equivalent to the 1987 24-hour PM10 standard (150 flg/m3, with a 99th percentile fo rm) . Due to the different sources and size distribution of PM in urban and rural areas, the proposed CPM standard would impact these areas differently. Nonet heless, the effects of the proposed standard on industrial and agricultural communities, as well as local regulatory agenci es, are largely unk nown. The study described here examines the trend of CPM mass concentration and chemical composition in the Los Angeles Basin over the past two decades, as well as the implications of the proposed PM1 0. 2 _5 standard given the nature and sources of PM components in the basin. The Los Angeles Basin is located in Southern California, and is the second most populous metropolitan area in the United Stat es. The basin is served by the nat ion's largest port complex, a large fr eight and passe nger rail infrastructure, numerous airports and an extensive network of fr eeways and highw ays. Due to its local topography and meteorological conditions, heavy reliance on vehicles and traffic activity fr om the ports, Los Angeles is ranked among the worst cities for ozone and particulate pollution (Ass ociation, Available at: http ://www.stateofth eair.org. Accessed November 30, 2011.), and has been an area of active air quality research. Studies examining characteristics of ambient aerosols across the basin have been conducted since the 59 1940s (Neiburger and Wu rtele, 1949), and a large body of historical PM data is available fr om regulatory agencies and academic institute s. Additionally, the contribution of coarse particle to PM10 is relatively high (on average 33-58 %) in Los Angeles (Gauderrnan et al., 2000; Pakbin et al., 2010) compared to several other large US metropolitan areas (U.S.EPA, 2004), thereby making it feasible to study coarse particle trends I composition with lower uncertainty. All of these provide a unique oppo rtunity to examine the historical trends of coarse particle in this basin, which will allow us to assess how the CPM fr action has been impacted by past and current air quality regulations, as well as implications of the proposed PM1 0. 2 _5 standards. 4.3. Methodology To study the CPM trend in the Los Angeles Basin, three repres entative areas (downtown Los Angeles, Long Beach and Ri verside) were selected based on the availability of historical dat asets. Figure 4. 1 shows a map of the sampling sites in the three regions . Downtown Los Angeles (sites DLA and USC) is a typ ical urban area with high commercial and traffic activity. The downtown area is characterized by dense freeway networks and a high volume of both heavy and light-duty vehicl es. DLA is a monitoring site maintained by the South Coast Air Quality Management District (SC AQMD) and is located within 900 m of fr eeways I-5 and I-llO. The other site in Los Angeles (USC) is located about 150 m east ofi-llO. Long Beach (sites HUD , NLB, LBCC and S3) is a mixed residential and commercial neighb orhood. The sampling sites in Long Beach are in close vicinity to the Port of Los Angeles and Port of Long Beach, and are highly influenced by heavy-duty diesel vehicle traffic related to harbor activity. NLB and HUD are maintained by the SCAQMD. NLB is located approximately 600 m north ofl-405 and 1.2 krn east ofl-710. HUD is about 100 m east of Terminal Island Freeway and 1.2 krn west ofl-710. LBCC is located at the 60 campus of Long Beach City College approximately 2.5 km north ofl-405. The S3 site was one of the sampling sites in a study that examined size-segregated PM in communities of the Los Angeles Harbor (Arhami et al., 2009), and it is about 900 m south of the HUD site. Riverside (sites VBR, RUB, U CR, and RIV) is a sub-urban area located 80 km inland and downwind of Los Angeles, and is generally considered to be a receptor site for pollutants generated in urban Los Angeles and advected to the area within a few hours after their emissions. VBR and RUB are SCAQMD monitoring sites located about 2.5 km and 800 m south of CA-60 respectively. The UCR site is situated at the Citrus Research Center and Agricultural Experiment Station of the University of California, Riverside, and it is upwind of surroun ding fr eeways . RIV is located at a retirement home which is 15 km southeast of downtown Riverside. 0 10 20 40 Kilometers I I II I I I I Los Angel� .oLA • usc orange san Bernardino o RIV Figure 4-1: Map of the samplings sites in downtown Los Angeles (DLA and USC), Long Beach (NLB, HUD, LBCC and S3) and Riverside (RUB, VBR, UCR and RIV). Sites operated and maintained by the SCAQMD are represented in triangles. 61 The CPM mass concentration data presented in Figures 4.2-4.4 were calculated as the differ ence between PM10 and PM 2 _5 levels (the subtraction method). Daily PM10 and PM 2 .5 measurements were obtained from the online data base of California Air Resources Board (CARB). The three sampling sites (DLA, NLB and RUB) used in this analysis were maintained by the SCAQMD, and the details of the sampling methods are described under Title 40 Part 58 of Code of Federal Regulations. In brief, daily PM 2 .5 measurements were quantifi ed gravimetrically using sequential samplers operating at 16.7 Llmin (Andersen Model Ref erence Ambient Air Sampler 2.5-300), loaded with a 2-flm Teflon substrate. PM10 measurements were determined gravimetrically using high-volume samplers (General Metal Works Model 1200) loaded with a quartz micro-fiber filter. Daily PM10 concentrations from 1988 to 2009 were analyzed. PM 2 .5 measurements were only available after 1999, and daily CPM mass concentrations were calculated fr om 1999 to 2009 using the "subtraction method" when both measurements were available for the same day. Note that the reported mass concentrations were collected using samplers with no denud ers. The distinct flow rates used in the PM10 and PM 2 .5 samplers might contribute to the differential loss of semi-volatile compounds, in additional to other sampling artifacts that could be different in the two samplers . Since the maj ority of semi-volatile compounds resides in the fm e PM fr action (Miguel et a!., 2004; Yo shizumi and Hoshi, 1985), the estimation of CPM using the subtraction method (calculated based on reported concentrations fr om the low-volume-PM 2 .5 sampler and the high-volume-PM10 sampler) might have underestimated the mass concentrations of coarse particles. Nonet heless, the sampling artifacts of semi-volatile organics are likely to be reduced by the long sampling time (24 hours) and higher mass loadings (Mader et a!., 2003; Sardar et a!., 2005). On the other hand, due to the stronger 62 removal process of CPM compared to fme PM, coarse particles could be more spatially heterogeneous (Pakbin et al., 2010), and the extrapolation of this analys is to other air basins needs to be proceeded with cautions. Daily PM10 levels were generally measured once every six days, with an average yearly count of 59 ± 3.5, 85 ± 27 and 57.8 ± 3.4 fr om 1988 to 2009 in downtown Los Angeles, Rivers ide and Long Beach resp ectively. The higher number count in Riverside was driven by the more fr equent measurements (once every three days) from 2000 to 2009. PM 2 .5 was measured more fr equently, with an average annual measurement count of 305 ± 58, 308 ± 56 and 303 ± 60 fr om 1999 to 2009 in DLA, RUB and NLB resp ectively. The higher standard deviation was driven by the lower count in year 1999 when the measurements first started. Table 4-1: Sampling location, sampling time and fr equency, sampling instrument and method, analytical method and other background information of the seven studies analyzed. Loutiea S3mpling Time CPid Catuld111 '3JQ·1;j: MMijB iifii@tiMM -:U#Ji!iJI II - �j!t.!!iif!MAI IMM Mifuffi@ijilMAI It M M.fii@tii!MAI It M MH@\Itt,tifiiAiiiM AQWDlos Angeles(DLA) AQMD Rubidoox (RI£) AQIIIID los Angeles (DLA) AQiuiD los Angeles (Dl,l,) AQMD Rubidoox (RI.SJ) AQMDlong8eaci1(NL8) longBeachCityColege(lBCC) JMJ-Oec 1986 \),cee\ft'ry G"'day,dMy samples 11 daysin Jun-Se p,6cl osysin I'«Jv-Ott in1387 Se<.!uetrtiaj sllflllle slornch episode (5W1s l.m'ller, 3in AQMD Rub4doox (P!.E) Jan1995-Feb 19% lroirect, t!i!ference het.oieen PM� Indirect dlereoce betlll! een PM� Indirect, o:li!ereoce he!IAieen Plto1� and PM,_, and PM, USCinlos �geles USC:Oct 2002-S ep2003; UCR: Mi!r-J.in 2001 '"'"' Riverside retirement home ( Rl'i) S3in loog8each .t.ug-Oct 20�,�-Feb2007 Mar-May 2007 Direct Direct USCfl losAngeles AQMD Van Buren ('118 R) AQMO loog Be�h (HUO) Apr2008-Mar2009 01ce evt'ry �,dailysa rll'ies oo lflteclnes�y USC 'Wtl;J! Imp�1cd'¥1) opermed !ll aa>taltowriSO ipn attd aminorflow d2�w illl a PM, inlet; PC IS opet*d at 9 \:lm willl aPM, nlet Therrn ald�·oo/OJJ'ical hmorrlission Concentrations of CPM mass and chemical constituents presented in Figures 4. 5-4.7 were obtained fr om studies fully or in part published in the peer-reviewed literature before 2010. 63 Although a number of studies have examined the chemical composition of ambient PM in Los Angeles, very few of them report data for coarse particles (or allow the calculation of such). Seven studies, providing chemical speciation of elemental species, were included in our analys es. The earliest study that performed chemical speciation on both PM10 and PM 2 .5 in the Los Angeles Basin was conducted in 1986. The year-long dataset was available at CARB (Solomon et a!., 1988) and part of the PM data were published in peer-review literature (Eldering et a!., 1991; Solomon et a!., 1989; Solomon et a!., 1992). Shortly after, Chow et a!. conducted an intensive study in 1987 to capture photochemical episodes in warmer periods (June to September) and stagnation episodes in cooler periods (November to December) (Chow et a!., 1994). Am ore comprehensive proj ect was carried out by the SCAQMD fr om January 1995 to February 1996 to investigate the chemical composition of ambient PM10 and PM 2 .5 in an effort to better character ize emission inventories and improve performance of modeling tools (Kim et a!., 2000a, b). From 2000 to 2009, a few additional studies were conducted in Los Angeles, Long Beach and Riverside to examine the chemical composition of size-fractionated airborne PM to advance the understanding of the sources, atmospheric processing of ambient particles and their human health impacts (Arhami et a!., 2009; Polidori et a!., 2009; Sardar et a!., 2005). In 2008, a comprehensive investigation on ambient CPM was conducted to study the chemical mass closure, as well as the spatial and temporal variation of CPM in the Los Angeles Basin (Cheung et a!., 20llb; Cheung et a!., 20 llc; Moore et a!., 2010; Pakbin et a!., 2010; Pakbin et a!., 2011). Table 4. 1 summa rizes the sampling location, sampling time and fr equency, sampling instrument and method, analytical method, and other background information of each study. CPM concentrations were obtained using the difference between reported PM10 and PM 2 .5 levels in the three earlier studies (Chow et 64 a!., 1994; Kim et a!., 2000a; Solomon et a!., 1988). The more recent studies employed samplers segregating particles by size, and CPM concentrations and chemical composition were obtained directly as measured in the coarse fr action (Arhami et a!., 2009; Cheung et a!., 20 llb; Polidori et a!., 2009; Sardar et a!., 2005). 4.4. Results and Discussion 4.4. 1. Trends in CPM mass concentrations Selected meteorological parameters at the downtown Los Angeles (DLA), Long Beach (NLB) and Rivers ide (R UB) sampling sites, respectively were acquired fr om the online dat abase of CARB; only data after 1994 were available at the three sampling sites. The annual averages of temperature and relative humidity were calculated based on hourly data. Overall, annual mean temperatures were consistent (A vg. � 17.9 'C, standard deviati on � 0.81 'C) fr om 1994 to 2009, while higher year-to-year fluctuations (A vg. � 59.3%, standard deviation � 11%) were observed for the relative humidity (RH). In general, the annual mean temperature and relative humidity was similar in Los Angeles and Long Beach. Riverside's inland location was highlighted by the higher mean temperature and standard deviation, coupled with the lower relative humidity. Wind originating fr om the west dominated most of the year, in accordance with the typ ical onshore flow patterns in the LA basin. The prevailing westerly onshore wind fr om the Pacific Ocean started in late morning and remained strong until eveni ng. At night, wind direction rever sed and the northerly I northeasterly wind prevailed. The wind speed was generally lower overnight. Overall, large-scale meteorological conditions were relatively consistent in this basin, and did not appear to influence the observed PM trends, as discussed and presented below. 65 Table 4-2: Linear regre ssion analys is of PM data (PM10 fr om 1988 to 2009, PM10.2_5 and PM2 _5 fr om 1999 to 2009) in: (a) downtown Los Angeles, (b) Long Beach and (c) Rive rside. Values in parent heses represent standard errors of the slope and intercept. a) Downtown Los Angeles Linea..r regres sion R 2 n PM 10 99th ercentile � -3.7 1(±0.39) X+ 136(±5.2) 0.82 <0.001 28th P-ercentile y = -3.2 1(±0.36) X+ 124(±4.8) 0.80 <0.001 50th percentile V = -1.19(±0.12) X+ 54.0(± 1.6) 0.83 <0.001 y = -1.33(±0. 12 X+ 57.±(± 1.6) 0.86 <0.001 PM 1 0.2 . 5 � -0.02(± 1.1 ) X+ 37. 5(±7. 1) 0.00 0.98 y = -0.42(±0.82 X+ 36.±(±5.6) 0.03 0.62 50th percentile V = -0.54 (±0.23) X+ 19.5(±1.6) 0.37 0.05 y = -0.32_ (±0.23) X+ 18.2_(± 1.5) 0.25 0. 11 PMz� Aver age y = -0.92 {±0.09 1 X+ 24.6 {±0.64 2 0.91 <0.001 b) Long Beach n PM 10 0.46 <0.001 0.47 <0.001 0.64 <0.001 0.72 <0.001 PMl0-2 . 5 0.06 0.48 0.3 1 0.08 0.43 0.03 0.3 1 0.07 PMz:i Aver age y = -0.87 {±0.09 2 X+ 22.4{ ±0.612 0.91 <0.001 c) Riverside Linear regression PM 10 V = -5.3 8(±0.85) X+ 207(± 11 )_, 0.67 y = �.72(±0. 77) X+ 182_(±10), 0.65 <0.001 Y = -1.59(±0.21) X+ 82.2(±2.8), 0.74 <0.001 y = -1.16(±0.23) X+ 84.6(±3.0), 0.75 <0.001 PMl0-2 . 5 � -2.36(± 1.1 ) X+ 93. 8(±7.7)_, 0.32 0.07 y = -1.77(±0.28) X+ 84.2(±6.7), 0.27 0. 11 Y = -0.56(±0.37) X+ 35.2 (±2.5)_, 0.21 0. 16 y = -0. 57(±0.29) X+ 35.7 (±2.0), 0.30 0.08 PMz:i Aver age y = -1.69 {±0. 12 2 X+ 33.5 {±0.842, 0.96 <0.001 66 In downtown Los Angeles, the 99th percentile for PM10 (Figure 4.2a) and CPM (Figure 4.2b) ranges fr om 56.7 - 88.6 flg/m3 and 22.7 - 60.4 flg/m3 respectively fr om 1999 to 2009. The spike in the 99 1h and 98 1h percentile in 2007 was driven by two days of high CPM concentrations ( 69 and 51 flg/m3 in April and October, respective ly). The PM10 level was in attai rnn ent of the federal 24-hour standard (150 flg/m3) while in violation of the California standard (50 flg/m3) throughout the entire study period. Since the current PM10 standard is expressed in the 99th percentile form and the proposed CPM standard is set in the 98th percentile fo rm, their comparability is exa mined. The diff erence between the 99th and 98th percentile is more apparent in the early 90s, with insiguificant difference after 1998. The average and median PM10 concentrations agree fairly well. The slope of the 99 1h percentile concentration is -3.71 with a R 2 of0.82 (Table 4.2a), indicating an average annual decrease of3.71 flg/m3 of the 99th percentile PM10 mass concentration fr om 1988 to 2009. The corresponding reduction is 1.33 flg/m3 for the average PM10 concentration. CPM levels did not exhibit the typ ically consistent decreasing trend observed in the PM10 data. The slope of the average CPM concentration is -0.39 flg/m3 (p � 0.11) fr om 1999 to 2009 (corresponding slope is -1.33 flg/m3 and -0.92 flg/m3 for PM10 and PM 2 .5 respec tively, both with a p-value < 0.001), indicating that the reduction ofPM10 is mostly driven by the decrease ofPM 2 .5 given the comparable levels off rne and coarse PM (average annual frn e-to-coarse ratio� 1.18 ± 0.21 fr om 1999 to 2009). To investigate the degree of equivalency between the current PM10 and the proposed PM1 0. 2 _5 standard, the number of exceedances was examined using the 99 1h percentile of PM10 level and the 98th percentile of the CPM level fr om 1999 to 2009. In both cases, zero violations were observed. For further comparisons of the two standard s, the ratio of the annual PM level to the corresponding standards (i.e., calculated as the annual 99 1h percentile PM10 to !50 67 flg/m3 for the PM 10 standard and the annual 98 1h percentile CPM to 70 flg/m3 for the CPM standard) was examined. Using the data fr om 1999 to 2009, the ratios for the PM 10 and prop osed PM 10.2_5 standard were 0.48 ± 0. 12 and 0.48 ± 0.07 respe ctively, suggesting that the two standards are comparable in downtown Los Angeles. '" " 1,1 --0-- 99th Percentrle (b) '" ------ 98th Percentrle "o oc ·� ---&- 5oth Percentrle � '" -A- Average " " i '" I " ec § " oc u 8 if " if " " i1 " u Year ffi � � --0-- CPM 99th Percentrle _._ CPM 98th Percentrle ---&--- CPM 5oth Percentrle --A- CPMAverage ----- PM,. Average � � � � Year • • � � � � � Figure 4-2: Annual concentrations of: (a) PM10 fr om 1988 to 2009, and (b) CPM I PM2.5 fr om 1999-2009 in downtown Los Ange les. In Long Beach, the 99 1h percentile PM10 level exp erienced a fluctuating decreasing trend with a R 2 of0.46 (Figure 4.3a). The spike in 1995 was driven by an episode of high PM10 level fr om mid-November to early December. The wind speed during that period was low with more than 70% of the time with "cahn" conditions (wind speed < 0.5 rn/s). It is possible that the high PM10 levels were due to the low dispersion of the emissions generated from the nearby harbor activity. The 99 1h percentile PM10 level ranges fr om 52.5 to 143.1 flg/m3 fr om 1988 to 2009, attaining the EPA standard but violating the California standard. The decreasing trend of the average PM10 levels exhibits a high R 2 of0.72 from 1988 to 2009 (Figure 4.3a and Table 4.2b). However, the average PM10 reduction rate is lower in the last 5 years, as evident by the lower slope of -0.14 flg/m3 (standard error � 0.25 flg/m3) from 2005 to 2009. Similarly, the average 68 CPM concentration was not sig nificantly reduced fr om 1999 to 2009, as demonstrated by the close to zero slope (m � -0.22 flg/m3, p � 0.07) in Table 4.2b, in contrast to the significant reduction in PM 2 5levels (m � -0.87 flg/m3, p < 0.001). The Long Beach site is heavily impacted by the Port's traffic activity. In addition to particle re-suspension by wind, CPM is re-entrained into the atmosphere by traffi c-induced turbulence (Pakbin et al., 20 10). Heavy-duty vehicles generally induce higher roadway re-suspension than light-duty vehicles (Charron and Harris on, 2005). Therefore, it is likely that the effort of PM reduction has been counterb alanced by the increa se in the number of heavy-duty vehicles resulted by the expans ion and increasing activity of the LA Ports, as the container volume grew fr om 1 million in 1985 to 7.8 million in 2010 at the Port of LA. In Long Beach, none of the recorded PM10 and calculated CPM levels violated the current federal PM10 and proposed CPM standar ds. The ratio of the 99th percentile PM10 and 98 1h percentile CPM to the current PM10 and proposed PM1 0. 2 .5 standard was 0.46 ± 0.08 and 0.45 ± 0.05 respec tively, again demonstrating that the two standards are comparable in Long Beach. 16o (a) � � 120 " we i 80 § 60 u := 40 i1 Year --o-- 99thPercentrle _._ 98th P ercentrle ---b- 50th P ercentrle ___.___ Average cc "t b"l ----- -=� o=cic c 'ii" , "" "' " '""' "" " "" '"""' --- ---, _._ CPM 98th Perrentrle ---b- CPM 5oth Perrentrle --A- CPM Average ----- PM,.A verage Year Figure 4-3: Annual concentrations of: (a) PM10 fr om 1988 to 2009, and (b) CPM I PM 2 .5 from 1999-2009 in Long Beach. 69 Due to its downwind receptor location and sub-urban nature, PM10 I CPM mass concentrations in Ri verside were higher than those observed in Los Angeles . The 99 1h percentile PM10 concentration (Figure 4.4a) was higher than the federal 24-hour standard of 150 flg/m3 in 1988-91, 1993-95, and 2003, and exp erienced a moderately fluctuating downward trend fr om 1988 to 2009 (R 2 � 0.67). This is probably due to the sub-urban nature of the area, where high level ofPM10 is mostly driven by particle re-suspension in episodes of high wind speed (Cheung et a!., 20llb; Pakbin et a!., 2010). As shown in Table 4.2c, the average PM10 level experienced a more consist ent reduction with a R 2 of 0. 75. From 1999 to 2009, the reduction of average PM10 was largely driven by the decrease in fine PM concentrations (m � -1.69 flg/m3, p < 0.001), with a minor contribution fr om CPM (m � -0.57 flg/m3, p � 0.08). The slope of the 99 1h percentile and average PM10 concentration is -5.38 and -1.76 flg/m3 respec tively from 1988 to 2009. The reduction of CPM mass concentration is sig nificantly lower than that observed in the PM10, and the trend is more variable, although, overall, a decreasing trend is seen as evident by the negative slopes of -2.36 and -0.57 for the 99th percentile and average CPM levels respectively. The relatively low coefficient of determination in the 99 1h percentile (R 2 � 0.32) is driven by the spike in 2003, due to a week of high CPM levels in October (91.0, 96.3 and Ill flg/m3 on Oct 21, 24 and 27 respectively). In Rivers ide, where the contribution of coarse particles to PM10 is generally higher than in Los Angeles and Long Beach due to the high levels of crustal materials in the sub-urban area (Cheung et a!., 20l lb), the proposed PM10. 2 .5 standard was not equivalent with the current PM10 standard. Using the data fr om 1999 to 2009, the annual 99th percentile PM10 level was higher than the current PM10 standard of 150 flg/m3 only in 2003. However, the annual 98 1h percentile CPM level was higher than the proposed CPM standard of 70 flg/m3 five times 70 (19 99-200 1, 2003 and 2006) from 1999 to 2009, suggesting that the prop osed standard is more stringent than the existing standard in areas dominated by coarse particles, and that Riverside could be in fr equent violations with the proposed PM10_ 2 _5 standard. (0) ""1 00 � � 80 j 60 Year Year Figure 4-4: Annual concentrations of: (a) PM10 fr om 1988 to 2009, and (b) CPM I PM 2 _5 from 1999-2009 in Rive rside. 4.4.2. Trends in chemical CPM corn ponents While CPM mass concentrations provide a metric of the overall mass reduction of coarse particles, the examination of CPM chemical composition provides insig hts on changes in the contribution of different sources to coarse particles, and may thus assist regulatory agencies in the desigu and implementation of more effective air quality strategies to protect public health. Figures 4.5-4.7 present concentrations of CPM chemical constituents, including organic and elemental carbon, elements and inorganic ions in downtown Los Angeles, Long Beach and Riverside, respectively. Note that the analyses described below fo cused on the study conducted in 1986 (Solomon et al., 1988), 1995-19 96 (Kim et al., 2000b) and 2008-2009 (Cheung et al., 20 l lb), all of which compiled comprehensive CPM datasets of one year or more, and their results are presented as bar charts in Figures 4.5-4.7. Although sampling of the three year-long studies was conducted not daily, but rather once every 3' ct , 6 1h or 7 t h day, the long sampling period was likely to diminish the impact of temporal variation, and could therefore provide a more 71 repres entative characterization of CPM. Other studies that were conducted in shorter time fr ames I smaller spatial scales were shown as line plots, and their results were likely to be influenced by seasonal variations of ambient CPM. Studies conducted before and after 1995 are represented by black and open symb ols, respec tively. = 1006,CAA8 100 10 (a) (b) c) • l 10 l l 0 1 • c c c a a a - � � - � i I 01 i 0 01 u u u "' 0.1 "' 0 01 "' 0.001 [L [L [L u u u 0.01 0 001 0.0001 Mass OC EC NH4+N03·S04·· Cl· Mg AI Fe Ca Tt K Mn p s v Ct N; Cu Zn Pb Figure 4-5: CPM concentrations of: (a) mass, organic and elemental carbon and inorganic ions, (b) elements of crustal origins, and (c) elements of anth ropogenic origins in downtown Los Ange les. Error bars show standard errors of the average when available. 100 (a) 10 (b) (c) l l • • l 10 0 1 • • c c I • c a a a -� • • � -� • I • I 0 1 I 001 u u u "' 0.1 "' 0 01 "' 0.001 [L [L [L u u u 0.01 0 001 0.0001 Mass OC EC NH4+N03- 804-· C!- Mg AI Fe Ca r; K Mn p s v Ct N; Cu Zn Pb Figure 4-6: CPM concentrations of: (a) mass, organic and elemental carbon and inorganic ions, (b) elements of crustal origins, and (c) elements of anth ropogenic origins in Long Beach. Error bars show standard errors ofthe average when available. 72 100 (a 0.01 = 1986,CAAB [ZZ3 1995-1996, Kill et al. , 20XI r=J 2008-20Cl9, Cheung et al, 201 1 • 1987 Sunrner, Chow et at., 1994 • 1987 Fall, Chow et al., 1994 b. 2001 Mar-Jun Sardar et aL 20)5 X 2006 Su rrrner, Po�dori et at., 2009 0 2007 Winter, Polidori et at, 2009 • Mass OC EC NH4+N03-S04-- Cl- 10 Mi c 0 p ro c 0 1 � 0 " 0.. 0 0.01 0001 (b) (c) Mi 01 c 0 p ro I 0.01 0 " 0.. 0.001 0 0 0001 Mg AJ Fe Ca Ti K Mo S V Cr Ni Cu Zn Pb Figure 4-7: CPM concentrations of: (a) mass, organic and elemental carbon and inorganic ions, (b) elements of crustal origins, and (c) elements of anth ropogenic origins in Rive rside. Error bars show standard errors ofthe average when available. Mineral dust. Many studies have shown that mineral and road dust are major components of CPM, contributing to 22-65% of CPM mass concentrations in studies conducted in Los Angeles, At hens, Helsinki, Amsterdam, and Eastern Mediterranean (Cheung et al., 2011b; Koulouri et al., 2008; Sillanpaa et al., 2006). Mineral dust comprises largely of eroded soil particles that have been mobilized and re-suspended into the atmosphere by wind and anth ropogenic activiti es. Industrial activities, such as construction and cement plant operat ions, may contribute to the emis sion and re-suspension of dust. Si, A1 and Fe are the three most abundant ma jor elements in the upper continental crust (Usher et al., 2003) and thereby may serve as tracers of mineral dust (Lough et al., 2005; Pakbin et al., 2011). Note that although Fe in the coarse fr action arises predominantly fr om crustal materials, it could be enriched fr om anth ropogenic sources such as brake wear in urban areas (Garg et al., 2000). Other crust al-dominated elements include Na, Ca, Ti, K, Mn and Mg, and their abundances vary depending on location and rock typ e. The decreasing trends of crustal materials were very consistent in the last two decades. From 1986 to 1995, levels of coarse particulate Al, Fe and Ti decreased by 34, 15 and 20% respectively in downtown Los Angeles. 73 Further reductions were observed from 1995 to 2008, when AI, Fe and Ti concentrations were reduced by 46, 49 and 70% respec tively. Such reduction trends were reinforced by the data from Sardar et al., where sampling was conducted fr om 2002 to 2003 (Sardar et al., 2005) and the average concentrations of AI, Fe and Ti stayed between those fr om Kim et al. (2000a) and Cheung et al. (20 11b ). The contribution of crustal materials and trace elements to CPM mass were higher in Ri verside (54.7%) compared with Los Angeles ( 42.6%) (Cheung et al., 2011b). In Rivers ide, CPM AI and Ti concentrations were reduced by 19 and 25%, resp ectively, from 1986 to 1995, where the corresponding reductions were 70 and 76% from 1995 to 2008. In 1986, the SCAQMD implemented Rule 111 2. 1 to limit PM emissions from cement kilns . Rule 1186 was adopted in 1997 to reduce the re-entrainrnent of fugitive dust of PM 10 emissions from paved and unpaved roads as well as li vestock operatio ns. This rule requires owners and operators of paved public roads to remove visible roadway accumulations within 72 hours of notificati ons. It also requires the paving or stabilization of heavily used unpaved public roads or the reduction of vehicular speed on such roads. In 2005, Rule 1157 was introduced to control PM10 emissions fr om aggregate and related operations by reducing various dust sources fr om loading, unloading and transferring activiti es, process equipment, paved and unpaved roads inside the facilit ies, etc. The use of dust suppressants or other control methods is required during transfer and loading activi ties. The observed high reductions of mineral dust could be due to the effecti veness of these regulations in reducing the sources and levels of re-suspension of dust part icles. Combw;tion emissions f[om vehicles. indw;tries and shi ps. Although particles from combustion sources mostly reside in the fine mode, the "upper tail" of that source function can extend to the lower range of the CPM fr action. For example, Huang and Yu (2008) demonstrated that elemental 74 carbon (EC), a tracer of vehicular emissions, experienced a bimodal pattern in both ambient and tunnel environmen ts, with a major fme mode and a minor coarse mode. Studies conducted in the Los Angeles Basin in the 20 t h century also demonstrated that, while dominant in the fine PM fr action, metals emitted by combustion sources could also contribute to a minor fraction of CPM (Krudysz et a!., 2008; Singh et a!., 2002). The tailpipe emissions of CPM is often associated with high emitting and poorly controlled vehicles (Kleeman et a!., 1999). In downtown Los Angeles, CPM elemental carbon was reduced fr om an annual average level of0.55 flg/m 3 to 0.42 flg/m 3 from 1986 to 1995 (23% reduction), and further to 0.08 flg/m 3 (80% reduction) fr om 1995 to 2008 (Figure 4.5a). The sampling sites in Riverside were located in residential areas adjacent to major roadways, and again, a similar reduction trend was observed (29% reduction fr om 1986 to 1995 and 71% reduction from 1995 to 2008) (Figure 4. 7a). The U.S. EPA requires automobiles of 1975 model year and after to be equipped with catalytic converters to control tailpipe emissions. In 1984, the Smog Check Program was implemented in California. In addition to various emission standards adopted for light-duty vehicl es, heavy-duty diesel trucks and bus engines in the 1990s, a number of programs were developed to remove the older and more polluting cars from the road. For examp le, California's Carl Moyer Program has provided funding to encourage the retrofit and replacement of diesel engines in an effort to reduce emissions fr om diesel-powered vehicles and equipment. In the last decade, the SCAQMD also adopted I amended a few rules to promote cleaner on-road vehicles used by the public sector. The high reductions ofEC, particularly after the 1990s, highlight the effective ness of various controls implemented by the regulatory agencies to control tailpipe emis sions. 75 In addition to tailpipe emissions, industrial activities also serve as potential sources of CPM in this basin. Metal processing facilities could emit high levels of heavy metals including Fe, Cu, Zn, As, Cd and Pb (Newhook et a!., 2003). In particular, smelters could release metals in the coarse size range either as stack or fugitive emis sions (Chan et a!., 1983; Harr ison and Williams, 1983), and could contribute to the high levels of Cu, Zn and Pb in 1986 as shown in Figures 4. 5-4. 7. In 1992, the SCAQMD adopted Rule 1420 to reduce emissions of lead- one of the six criteria pollutants identified by the U.S. EPA. This regulation requires all emis sion points of a lead-p rocessing facility (including facilities that produce lead-oxide, brass and bronze) to be vented to an emission collection system. Furthe rmore, facilities are required to conduct air quality monitoring at their property lines, in addition to the source-oriented monitors placed by the SCAQMD at or beyond the facilit ies' fence lines. The high reductions of Cu, Zn and Pb from 1986 to 1995 (96, 84 and 69% respectively in downtown Los Angeles) could largely result from the effective control of emissions fr om metal processing facilit ies. Oil refinery is also a major source of industrial emissions. Various coarse elements including Pb, Ce, La, Zn, V, Cu, Co, and to a lesser extent Cr , Ni and Mo could be emitted from refinery fluid catalytic cracking stacks (Campa et a!., 20 II). Using principal component analys es, Pakbin et a!. (20 II) showed that CPM-bound Ni and Cr are markers of industrial emissions, with V originating from ship emissions in the Los Angeles I Long Beach Harbor based on coarse particles sampled from 2008 to 2009. In general, the lower concentrations of these elements in the coarse mode support that they originate fr om combustion sources in this basin. In downtown Los Angeles, the reductions ofV, Cr and Ni were higher from 1995 to 2008 (> 90% reduction calculated using 1995 and 2008 data), compared with those fr om 1986 to 1995 (Figure 4.5c). In 76 Rivers ide, the corresponding reductions of Ni, Cr and V were 93, 88 and 89%, respectively, from 1995 to 2008, consistent with the observations in downtown LA. The significant reductions after 1995 could be resulting from series of regulations targeted to control emissions from stationary sources, as evident by the reductions in PM10 emissions from industrial processes, manufacturing and industrial sources fr om 1995 to 20 10 according to the emis sion inventory of the South Coast Air Basin acquired from CARB . For examp le, Rule 11 05.1, aiming to reduce PM 10 and ammonia emissions from fluid catalytic cracking units used in petroleum refine ries, was adopted in 2003. Additionally, the Los Angeles I Long Beach ports approved an incentive program that promotes the use of cleaner-burning fuel in cargo ships transiting within 40 miles of the Bay and at berth in 2008. Note that the efforts to control ship emissions began in the early 20 1h century, consistent with the historical trend of vanadium, which experienced similar annual averages in 1986 and 1995, with higher reduction rates fr om 1995 to 2008. Therefore, the high reductions ofNi, V and Cr after 1995 could be due to various programs and incentives that aim to reduce combustion-related emissions from vehicles and indust ries, as well as the efforts to reduce ship emissions from harbor activity. CPM-bound lead (Pb) also decreased considerably in the LA basin. Pb could originate from brake wear and wheel weights, gasoline exhaust and oil combustion as the extended upper tail of the fine PM mode, as well as metal processing industries (Harri son and Williams, 1983; Isakson et a!., 2001; Pakbin et a!., 2011 ). In downtown Los Angeles, Pb levels have decreased from an annual mean of 0.097 flg/m 3 to 0.030 flg/m 3 from 1986 to 1995, and further to 0. 0023 flg/m 3 in 2008. As discussed previously, the early reduction of Pb could be resulted from Rule 1420, which was adopted in 1992 aiming to control lead emissions. The latter reduction (after 1995) could 77 partly be att ributed to the complete elimination of the sales of leaded fuel for use in on-road vehicles in 1996. In Long Beach, ships (Isakson et a!., 200 1) and refineries emissions (Newhook et a!., 2003) could also be potential sources oflead. The reduction in CPM lead concentrations in Long Beach was comparable to those observed in downtown LA (0.064 flg/m 3 in 1986 to 0.0015 flg/m 3 in 2007). Recent studies, including a study of CPM in Los Angeles and Long Beach conducted in 2008 (Pakbin et a!., 20 II; Root, 2000), indicated that Pb is now predominantly sourced from vehicular abrasion. Therefore, it is likely that the Pb reductions in the last two decades were largely driven by the phase-out ofPb in gasoline vehicl es, as well as cont rolled emissions from various industries as discussed previously. Vehicle abrasion. Particles originating from vehicular abrasion contribute to road dust PM, though the mass -f raction is typ ically small. Cu, which is present in brake linings as lubricants, is often used as tracers of brake wear in areas where industrial emission ofCu is not significant (Lin et a!., 2005; Pakbin et a!., 20 11). Note that the high concentrations of Cu in 1986 likely resulted fr om emissions of metal processing indust ries, as discussed in the previous section, and therefore Cu might not be a reliable tracer of brake wear back in the 1980s. Nonethe less, a recent study in the LA basin suggested that ma jority of coarse particulate Cu was generated from the wear of brake linings based on CPM sampled from 2008 to 2009. Despite the considerable decrease of elements of combustion origins from 1995 to 2008 (Pakbin et a!., 2011), the reduction of Cu was less significant in the same period. In Los Angeles, the higher annual average concentrations of Cu in 2008 compared to the 1995 levels could be due to the inc reased emissions from tire and brake wear from the greater number of both light-duty and heavy-duty vehicles on the road. The contribution of anth ropogenic sources to elements, including Cu, is generally higher in Los 78 Angeles and Long Beach than River side (Cheung et a!., 20llb). From 1995 to 2008, the lower reduction ofCu (46%) relative to other CPM species with anth ropogenic origins (89, 88 and 93% for V, Cr and Ni respec tively) in Rivers ide further confirms the lower reduction of non -tailpipe mobile source emissions. The reduction ofCu might also be a "side-benefit" as a result of the high reduction of mineral dust in Rivers ide. Inorganic ions. In general, reductions of CPM inorganic ions were relatively lower than combustion related CPM due to their natural origins and, to a lesser extent, secondary fo rmation. Nitrate is the most abundant inorganic species in the coarse fraction, accounting for an average of 17% of CPM mass in this basin (Cheung et a!., 20llb ). In the Los Angeles Basin, nitrate is primarily formed by sea salt depletion, a process involving the reaction of nitric acid with sodium chloride (Zhuang et a!., 1999), and to a lesser extent the reactions with mineral dust and the condensation of alkaline salts on CPM surf aces when sea salt levels are low (Cheung et a!., 20llc). In downtown Los Angeles and Ri verside, the level of nitrate was reduced by 13% and 34% respectively from 1986 to 1995, while corresponding reductions were 38% and 42% from 1995 to 2008. Substantial efforts have been made to reduce emissions of nitrogen oxides, as demonstrated by the reduction ofNOx emissions from 1561 tons per day (TPD) in 1985 to 1332 TPD in 1995 and 742.2 TPD in 2010 in the South Coast Air Basin (Emi ssion Inventory ofCARB). In addition to the regulations to control mobile emissions as described previously, the SCAQMD has adopted a few rules to control NOx emissions from stationary sources, including boilers and process heaters from refineries, cement kilns, gas melting furn aces and stationary gas tur bines . Since 200 I, the LA ports incorporated a voluntary vessel speed reduction program with the objective of reducing both NOx and PM emissions by decreasing the vessel's speed to 12 knots 79 near the ports . The control ofNOx emissions may have an indirect effect on the reduction of coarse particulate nitrate by limiting the levels of its precu rsors, which could also decrease the CPM-bound nitrate associated with the hyg roscopic growth of nitrate originally in the fine PM mode (Geller et a!., 2004; Seinfeld and Pandis, 2006). In addition, the reduction of mineral dust, which could serve as a reaction site for nitrate (Usher et a!., 2003), might have also indirectly reduced the levels of nitrate in CPM. Sulf ate, on the other hand, has several potential sources, namely water- soluble gypsum and sea salt sulfate, as well as the upper tail of ammonium sulf ate, and the reduction was insignificant. Chloride, primarily originating from sea salt in the coarse mode, is used as a tracer of fresh sea salt aer osols. cr levels increased fr om 0.37f 1g/m 3 in 1995 to 0.55 flg/m 3 in 2008 in downtown Los Angeles. This could be due to yearly variations in sea salt level, as well as the reduction in nitrogen oxides emis sions, which decreased the rate of sea salt depletion and increased the levels of unreacted chloride. 4.5. Summary and Conclus ions Overall, PM10 mass concentration has decreased by approximately half from 1988 to 2009 since the PM10 standard was put in place in 1987. PM10 daily concentrations from downtown Los Angeles and Long Beach show compliance with the federal EPA PM10 standard, while violating the more stringent California standard. These two sites also demonstrated equivalency for the current PM10 and proposed PM10.2.5 standard. On the other hand, both PM10 and CPM levels were higher in Ri verside as highlighted by some violations of the federal standards in late 80s and early 90s. This site also demonstrates that the prop osed PM1 0.2.5 standard is more stringent than the current PM10 standard. The reduction trends of combustion-related CPM namely EC, V, Cr and Ni were higher after 1995, consistent with the implementation ofPM2.5 standard in 1997. The 80 concentrations of Cu, a tracer of brake wear, has decreased at a lesser rate or remained comparable relative to other anth ropogenic constituent after 1995, suggesting that the contribution of brake wear to CPM has become more significant despite the overall reduction in CPM mass since 1995. In general, the reduction of CPM mass was mostly driven by the reduction of mineral dust, while reduction of contributions from inorganic ions and non-tailpipe vehicular emission were less signif icant from 1995 to 2008. 4. 6. Acknow ledgments The authors would like to acknowledge the support of the Science to Achieve Results program of the U.S. EPA (EPA-G2006-ST AR-Q l). The authors would also like to thank Bong-Mann Kim at the Sou th Coast Air Quality Management District for providing supplementary information regarding his pro ject. 81 Chapter 5 Diurnal Trends in Coarse Particula te Matter Composition in the Los Angeles Basin 5.1. Abstract To investigate the diurnal profile of the concentration and composition of ambient coarse partic les, three sampling sites were set up in the Los Angeles Basin to collect coarse particulate matter (CPM) in four different time periods of the day (morning, midday, afternoon and overnight) in summer and winter. The samples were analyzed for total and wate r-soluble elements, inorganic ions and wate r-soluble organic carbon (WSOC). In summer, highest concentrations of CPM gravimetric mass, mineral and road dust, and WSOC were observed in midday and afternoon, when the preva iling onshore wind was stronger. In general, atmospheric dilution was lower in winter, contributing to the accumulation of air pollutants during stagnation conditi ons. Turbulences induced by traffic become a significant particle re-suspension mechanism, particularly during winter nighttime, when mixing height was lowest. This is evident by the high levels of CPM mass, mineral and road dust in winter overnight at the near-freeway sites located in urban Los Angeles, and to a lesser extent in Rivers ide. WSOC levels were higher in summer, with a similar diurnal profile with mineral and road dust, indicating that they either share common sources, or that WSOC may be adsorbed or absorbed onto the surf aces of these dust particles. In general, the contribution of inorganic ions to CPM mass was greater in the overnight sampling period at all sampling sites, suggesting that the prevailing meteorological conditions (lower temperature and higher relative humidity) favor the fo rmation of these ions in the coarse mode. Nitrate, the most abundant CPM-bound inorganic species in this basin, is fo und to be predominantly fo rmed by reactions with sea salt particles in summer. When the sea salt 82 concentrations were low, the reaction with mineral dust particles and the condensation of ammonium nitrate on CPM surfaces also contribute to the formation of nitrate in the coarse mode. 5.2. Introduction Short term exposure of high concentrations of coarse particles was related to various adverse health endpoints (Lipsett et a!., 2006). A study conducted in Chapel Hill, North Carolina also showed that small temporal increases in coarse particle concentrations could affect cardiopulmonary and lipid parameters in the adult asthmat ics, and the investigators concluded that CPM-induced health effects may have been underappreciated in susceptible populations (Yeatts et a!., 2007). To improve our understan ding of the phys ical and chemical characteristics as well as the spatial and seasonal variation of CPM, a study was conducted earlier by our group using 24-hour time integrated samples collected in 10 sampling sites at the Los Angeles Basin, and the results were presented in our previous publications (Cheung et a!., 20lla; Pakbin et a!., 20 10). It was fo und that both CPM mass concentrations and chemical composition varied considerably by season and displayed spatial heterogeneity. However, the long sampling periods (24-hour) suppr essed the observation of the evolving CPM chemical characteristics due to the indisputable influence of many important atmospheric parameter s, such as temperature, relative humidity, wind direction and speed, and mixing height, all of which vary in scales shorter than 24 hrs. The study presented here addr esses this issue by foc using on the diurnal trends of coarse particles in the Los Angeles Basin, and provi des information that will be vital in improving our understanding on how these atmospheric parameters influence the ambient coarse PM concentrations and composition. 83 5.3. Methodology 5.3. 1. Sites description Three sampling sites were selected to represent distinct geographical regions of the Los Angeles Basin. Figure 5.1 shows the locations of the three sampling sites: downtown Los Angeles (USC), Lancaster (LAN), and Ri verside (RIV). The USC site is located in an urban area in downtown Los Angeles, and is approximately 150m east ofFreeway 1- 11 0 and 20 km east ofthe Pacific Ocean. The RIV site is located in Rivers ide, about 80 km inland of the USC site. Although Ri verside is a semi-rural agricultural region, it is along the typ ical prevailing wind tra jec tory crossing the Los Angeles Basin from the coast to inland (Eiguren-Fernandez et al., 2008), and is thereby a receptor area of pollutants fr om urban Los Angeles. The site in Lancaster (LAN), located about 75 km north of the USC site and over 2 km west of highway CA-14, is characterized by a desert-like rural nature. 0 "10 20 40 Kilometers N -�- Figure 5-1: l\1ap of the 3 sampling sites. 5.3.2. Sampling time and setup The sampling campaign was conducted in both summer (July - August 2009) and winter (January - February 2010). Four distinct time periods - morning (7:00a.m. to 11:00 a.m.), 84 midday (11:00 a.m. to 3: 00 p.m.), afternoon (3:00p.m. to 7:00p.m.) and overnight (7:00p.m. to 7:00 a.m.) - were selected to reflect the effect of different atmospheric parameters on CPM sources and fo rmation mechanisms. Sampling continued for 4 to 6 weeks in each season depending on ambient concentrat ions. Samples were collected using the USC Coarse Particle Concentrator (Misra et a!., 2001) downstream of a PM 10 inlet. The USC Coarse Particle Concentrator was com posed of a virtual impactor placed upstream of a filter holder. Coarse particles were concentrated approximately 25 -fold using the virtual impactor with an intake flow rate of 50 LPM and a minor flow of 2 LPM, and then collected on Teflon filters with a diameter 47 mm and a pore size of 2.0 f!m. Four samplers were used at each site. Substrates from each time period were analyzed separately. 5.3.3. Chemical analyses Teflon substrates were weighed before and after sampling to measure the collected PM mass. They were equilibrated for 24 hours under cont rolled relative humidity (30% ± 5%) and temperature (21 'C ± 2'C) before weighing by a microbalance (Model MT 5, Mettler-Toledo Inc., Highstown, NJ). Water extracts of the Teflon filters were analyzed for water soluble organic carbon (WSOC) using a Sievers Total Organic Carbon analyzer (General Electric, Inc.) (Zhang et a!., 2008b) and inorganic ions by Ion Chromatography (Lough et a!., 2005). The same aqueous extraction was used to measure the concentration of water soluble metals and elements using Inductively Coupled Plasma-Mass Spectroscopy (ICP-MS) (PQ Excell, Thermo Element al). To determ ine the levels of total elements, the PM collected on Teflon filters was digested using a mixture of 1.0 rnL of 16 N HN0 3 , 0. 1 rnL of 28 N HF, and 0.25 rnL of 12 N HCl in a microwave digestion unit and subseq uently analyzed by magnetic-sector ICP-MS (SF-ICPMS) (Lough et a!., 85 2005). The samples were analyzed with field and laboratory blanks, filter spikes and external check standard s. On average, replicates were analyzed once every ten samples and blank levels were less than 2% of the samples ' levels . The uncertainties of some water soluble trace elements were high due to their low concen trati ons, and they were not included in the analys is. Spike recoveries were within the acceptance range of 85-ll5 %. 5.4. Results and Discussion 5.4.1. Meteorology Selected meteorological parameters in the two sampling periods are shown in Table 5 .I, with vector-averaged wind direction and speed. This information was obtained from the online data base of the California Air Resources Board. The met eorological station was located about 7 krn northeast of the USC sampling site, while the meteorological stations were located at the same site as the sampling instruments at the RIV and LAN sites . In summer, highest temperature was observed in midday. Av erage temperature was higher inland in Lancaster and Rive rside. The high temperature also extended from midday into the afternoon in the inland area. Relative humidity was highest overnight for all sampling sites . The "desert-like" characteristics of Lancaster are evident in Table 5 .I by the lower relative humidity and the higher temperature range. In summer, wind speed was highest in the midday and afternoon periods, with predominant wind direction fr om the west, which is typical in this region. Somewhat unexpectedly, the wind speed was high during the overnight period in Rivers ide. In winter, the Los Angeles Basin is characterized by lower mixing heights and frequent periods of air stagnation, resulting in the accumulation of air pollutants (Hildemann et a!., 1994). In urban Los Angeles (USC), winds originated from the west during midday and afternoon, in accordance with the typical onshore 86 flow patterns of the basin. However, this pattern was not as persistent at the two inland sites, where the predominant wind direction was from the north. The low wind speed at the two inland sites during the sampling period indicates stagnant meteorological conditions, especially during the overnight period, when the temperature was lowest. In Lancaster, the average temperature in the morning and overnight periods dropped below 10° C, highlighting its "desert-like" nature. Table 5-1: Selected meteorological parameters at the 3 sampling sites in: (a) summer and (b) winter. a) Summer Los A nge les (USC) RYersrle (RlV) Sunmer Morning Midday Afternoon Overnight Mommg Midday Afternoon T erq>erature (°C) 26.6 29. 6 25 1 194 28.7 36 31.8 Relative Humrlty (%) 539 46 464 73 3 525 33 431 Wnd Speed (m/s) 1.3 3.6 3.8 1.2 0.63 3.2 3.1 Wind Di-ectico s sw w w w w w b) Winter Los An g eles (USC) R1versrle (RIV) Writer Morning Midday Afternoon Overnight Morning Midday Afternoon Terq>erature ("C) 16.1 22.3 18.3 12.5 14.1 21.9 18.9 Relative Humrlty (%) 52.9 34.1 47.4 66.2 54.1 27.8 36.7 Wnd Speed (m/s) 20 098 20 15 02 7 02 7 13 Wi ndDi-ection NE sw w NE NE N NW 5.4.2. Coarse PM component model and data overview Lancaster (LAN) Ovemtght Mommg Mtdday Afternoon Ovenught 213 29 9 363 341 25 6 741 21 4 15 9 202 25 6 3.2 0.63 2.4 4.1 0.98 w w w sw sw Lancaster (LAN) Overnight Morning Midday Afternoon Overnight 10.9 8.2 13.6 14.2 8.2 6.4.9 61.3 43.6 42.1 62.5 0 22 009 0 31 0.18 0.49 N NW N W NW sw Chemical components were classified into 5 groups according to their common sourc es: 1) crustal materials and trace elements (CM + TE); 2) vehicle abrasion (VA) ; 3) water soluble organic carbon (WSOC); 4) sea salt (SS) and 5) secondary ions (SI). Crustal material (CM) is calculated using the following equation (Chow et al., 1994; Hueglin et al., 2005): CM = 1.89 A1 + 1.21 K + 1.43 Fe + 1.4 Ca + 1.66 Mg + 1.7 Ti + 2.14 Si ( 1) 87 where Si is estimated from AI using a factor of 3.41 (Hueglin et al., 2005). The group of trace elements (TE) includes Rb, Sr and rare earth elements such as La and Ce, etc, and is grouped with CM because of their common crustal origins. Vehicle abrasion includes traffic related emissions such as Cu, Zn, and Ba (Harrison et al., 200 1). Water- soluble organic carbon is obtained directly by the measured WSOC. The sea salt component is estimated using soluble Na +and the sea salt fr action of cr . Mg 2 +, K+, Ca 2 + and So/· fr om the typical sea water composition (Sei nfeld and Pandis, 2006), represented by the equation below: SS � Na+ + ssCl" + ssMg 2 + + ssK+ + ssCa 2 + + ssSO /" (2) The use ofNa ion as a sea salt tracer represents both fr esh and aged sea salt aer osols. The group of secondary ions consists of ammonium, non-sea salt (nss) sulfate and nitrate. Table 5.2 (a and b) shows the overview of the categorized chemical components concentrations during each time period and their 24-hour averages. Overall, crustal materials and trace elements is the most abundant category, with an average concentration of 4. 54 f1glm3 in summer and 3.85 f1glm3 in winter across the three sampling sites. In particular, levels were higher at RIV (9.76 f1glm3 and 5. 71 f1glm3 for summer and winter, respec tively) due to its semi-rural nature and receptor location. In general, the CM + TE level in Lancaster was lower than Riverside due to its remote location and lack of anth ropogenic sources and I or re-suspension of mineral and road dust. In particular, vehicles driving on dirt and I or unpaved roads in semi-rural Ri verside could contribute substantially to the generation and re-entrainrnent of mineral and road dust. In summer afternoon, the level of crustal material at LAN was similar to RIV, due to the comparable high wind speed as the main re-suspension mechanism at both sites . CPM originating fr om vehicle abrasion, on the other hand, is less abundant (on average 0.061 f1glm3, 0. 14 f1glm3 and 88 0.36 flg/m 3 for LAN, USC and RIV, respectively) compared with the crustal materi als. The VA levels at RIV were relatively high. As discussed previously, unpaved roads or paved roads with unpaved shoulders in Rivers ide could contribute substantially to the fugitive dust components of CPM. Chow et a!. ( 1992) showed that vehicle movement associated with agricultural planting and harvesting, as well as the tra nsport of agricultural products along the unpaved roads, was the primary contributor of fugitive dust in the semi -rural agricultural area of San Joaquin Valley (Chow et a!., 1992). Vehicle speed could be another factor in the re-suspension of VA (Nicholson et a!., 1989). In the morning and afternoon period in downtown LA, vehicles travel at lower speeds due to high traffic volume. On the other hand, the levels ofV A were higher in the midday at RIV and USC when traffic volume was lower. SS and SI contributed similarly at an average concentration of0.41 flg/m 3 and 0.50 flg/m 3 respectively. SS concentrations experienced a seasonal variation with higher levels in summer (on average 0.53 flg/m 3 ) than winter (on average 0.28 flg/m 3 ). WSOC concentrations were low at LAN and USC (on average 0.070 flg/m 3 and 0.077 flg/m 3 , respectively), with higher levels at RIV at 0.37 flg/m 3 Table 5-2: Av erage diurnal concentration (flg/m 3 ) of chemical components at the three sampling sites in: (a) summer and (b) winter ( ± indicates uncertainties calculated based on the analytical uncertainties and uncertainties fr om blank correct ions). a) Summer b) Winter Summer CM+TE VA ss Sl wsoc W1nter CM+TE VA ss Sl wsoc Morn1ng 1 18+0 07 0 042+0 002 031+003 0 24+0 02 0 1+0 01 Morn1ng 1 76±0 08 0 085±0 003 0 049±0 005 0 055±0 02 0 06±0 01 z Midday 28 8±015 0 084±0 004 01 9±002 0 33±0 03 0 081±0 02 z Midday 1 94±0 09 0 055±0 002 0 081 ±0 007 " 0 03±0 :3 Afternoon 799±041 0 18±0 008 0 82±0 08 0 76±0 06 031±002 <>:: Afternoon 11 8±006 0 047±0 002 0 11± 0 01 0 03±0 02 0 03±0 01 � ��r r n i v � t 0 62±0 04 0 038±0 002 0 042±0 005 0 067±0 01 0018 ±001 � ��r r n��t 1 56±0 08 0 039±0 001 0 072±0 007 0 11± 0 01 0 06±0 01 2 32 0 070 0 24 0 26 0 091 1 59 0 051 0 076 0 083 0 050 Morn1ng 0 88±0 05 0 087±0 005 0 05±0 01 01 2±001 0 046±0 01 Morn1ng 23 ±012 0 122±0 005 0 06±0 006 0 087±0 02 0 042±0 01 u Midday 349±019 0 19±0 009 0 66±0 06 0 73±0 06 01 ±002 u Midday 2 85±0 15 0 097±0 003 0 26±0 024 0 23±0 03 0 056±0 01 � Afternoon 21 4±0 12 0 076±0 003 0 66±0 06 0 64±0 05 0 085±0 02 (f) Afternoon 1 87±0 11 0 068±0 003 0 055±0 005 0 079±0 02 0 029±0 01 Overnight 0 94±0 05 0 051±0 002 0 88±0 08 0 67±0 05 0 073±0 01 :J � ��r r n�� t 6 14±0 32 031±0014 0 94±0 089 0 88±0 07 0 12±0 01 24-hr Avo 1 56 0 084 0 67 0 58 0 075 4 24 0 20 0 53 0 51 0 081 Morn1ng 7 54+0 38 0 47+0 03 0 35+0 03 0 49+0 04 0 29+0 04 Morn1ng 3 55±0 18 0 24±0 01 0 23±0 02 0 5±0 05 0 18±002 > Midday 491±025 01 5±003 0 29±0 03 0 59±0 05 0 17±0 02 > Midday 3 03±0 16 01 ±0005 01 8±002 0 26±0 03 0 12±0 01 C2 Afternoon 119±062 05 9±011 1 16±0 11 1 34±0 11 1 14±0 06 C2 Afternoon 4 35±0 23 0 55±0 03 0 2±0 02 044±0 04 0 18±002 Overnight 114±055 0 37±0 07 0 76±0 07 0 82±0 06 0 56±0 02 Overnight 7 78±0 41 0 39±0 02 0 28±0 03 1 09±0 09 0 24±0 02 24-hr Avq 9 76 0 39 0 68 0 81 0 55 24-hr Avq 5 71 034 0 24 0 75 0 20 89 5.4.3. Diurnal profiles Figures 5.2, 5.4 and 5.5 show the diurnal profile ofCPMmass and the five chemical categories in summer and winter, with the group of secondary ions presented as ammonium, nitrate and nss-s ulfate. Normalization to the corresponding 24-hr average (i.e. CPM mass I 24-hr average CPM mass concentrati ons, WSOC I 24-hr average WSOC concentrati ons, etc) are presented to highlight the diurnal trends of these species. A detailed discussion of each group is off ered in the following sections . 3 0 2 5 I � CPM /24-hr Avg CPM Summer I CPM /24-hr Avg CPM Winter 0 2 0 . "' • � . 1 5 ·" � z 1 0 " " 0 • • " " � 0 • " • • " !I " • " • 0 5 0 0 0 0 0 0 0 0 0 0 0 ! 0 ! 0 ! 0 " � " � " � � � � � � � � � � 4 0 4 0 4 0 LAN usc RIV Figure 5-2: Diurnal profiles (normalization to the 24-hr average) of CPM mass. CPM mass. In general, the diurnal profile of CPM gravimetric mass concentration closely followed the profile of wind speed in summer. The highest CPM normalized ratios were often observed in midday and I or afternoon, when the wind speed was stronger. This trend is not as persistent in winter when the wind speed was low. Despite the lower wind speed in winter, the overall CPM mass concentrations remained comparable to those of summer in this study, likely due to the staguation atmospheric conditions as discussed earlier. In winter, the overnight CPM 90 mass concentration was high at USC (normalized ratio � 1.12 ), likely due to particle re-suspension by the northeasterly wind coupled with the lower mixing height at night. At Riverside, despite the low wind speed (ca. 0.22 rn/s ), the overnight CPM mass concentration remained high. Similar results have been observed in some previous studies showing higher levels of traf fic-emitted pollutants, including EC and CO, in late night and early morning (compared with midday and afternoon) in Riverside (Polidori et a!., 2007; Snyder and Schauer, 2007), as a result of lower atmospheric dilution in the nighttime period of winter. Furtherm ore, elevated turbulence induced by increased traffic activity might become a sig nificant particle re-suspension mechanism in early morning when atmospheric dilution is low. Harri son et a!. (200 I) demonstrated that turbulence induced by vehicles provides a source strength that was approximately equal to that of exhaust emissions, and concluded that coarse particles predominantly come from vehicle-related re-suspension at the studied sampling sites in the United Kingdom. Additionally, heavy-duty vehicles are known to induce higher roadway CPM re-suspension than light duty vehicles (Charron and Harrison, 2005). Carbon monoxide, sourced primarily from vehicles in urban atmospheres, has been used as a reliable tracer of primary emission (Ning et a!., 2007). As traffic flow and pattern are consistent in both sampling periods, CO levels can be used to indicate the degree of atmospheric dilution in ambient conditio ns. Fignre 5.3 (a-c) shows the diurnal CO levels at the three sampling sites acquired from the online datab ase of the South Coast Air Quality Management District's monitoring stations located at or within 7 krn of the sampling sites employed in this study. In summer, CO concentrations were higher during daytime, in contrast to the significant overnight peak experienced at all sampling sites in winter. Although the emis sion source strength was lower 91 during nighttime, the CO levels were approximately 2-4 times higher in winter overnight compared with daytime. In general, CO levels were also higher in winter than summer, particularly during the nighttime, highlighting the much lower atmospheric dilution experienced in the overnight sampling period in winter. Therefore, turb ulence induced by vehicular traffic could become a significant mechanism of particle re-suspension that increases pollutant concentrations during the overnight period, especially considering the prevailing air stagnation conditions and the low atmospheric dilution at that time. " " i " c 0 cc � � co u 0 u " (o) 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 T1me of Day " (c) " i " c c c 0 � co � " u 0 u " " " " i " c 0 c c � � co u 0 u " (0) 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 T1me of Day 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 T1me of Day Figure 5-3: Carbon monoxide (CO) levels at: (a) Los Angeles-USC; (b) Lancaster- LAN; (c) Rivers ide-RIV. 92 Crustal materials and trace elements. Figure 5.4a shows the diurnal profile of CM + TE. In summer, the diurnal trend at Lancaster paralleled that of wind speed, peaking in the afternoon (normalized ratio � 3.4 4) when the average wind speed was strongest at 4. 1 rn/s. Crustal material and trace elements are the dominant sources of CPM at the Lancaster site which is located in a remote area away fr om fr eeways and urban pollutants. As previously discussed, RIV experienced higher CM + TE levels due to its downwind receptor location and semi -rural nature, with the highest normalized ratio in the afternoon. The average concentration was much lower in urban Los Angeles (USC), where the diurnal trend of CM + TE follows closely the CPM mass concen trati ons. In winter, overall CM + TE concentrations were lower in comparison to summer except for overnight. At USC, the relatively high wind speed, coupled with lower atmospheric dilution and traffi c-induced turbulence in the overnight period may explain the high levels of CM + TE (normalized ratio � 1.45), consistent with the peak in the overall CPM mass concentration. On the other hand, CM + TE levels remained high (normalized ratio � 1.36) despite the relatively low CPM mass and low wind speed overnight at RIV. As mentioned previously, re-suspension induced by heavy-duty vehicles might be responsible for the re-entrai rnn ent of local mineral dust, which is a dominant source ofCPM in Riverside (Cheung et a!., 20 l la). Vehicle abrasion. Traditionally, both emission control strategies and regulations have specifically targeted tail-pipe exhaust sources from vehicles. Although non-tailpipe PM is currently unregulated, it has been shown to be a siguificant source of traffic-related emissions (Querol et a!., 2004). Vehicle abrasion, such as tire and brake wear, is an important source of CPM in urban areas (Harrison et a!., 2001; Lenschow et a!., 2001). Unlike crustal mate rials, metals of anth ropogenic origins have higher solubility (Birmili et a!., 2006), and might have important 93 public health imp acts. In general, the diurnal profile of VA follows a similar pattern with CM + TE, suggesting that they are re-suspended together as soil and road dust. The high VA levels in the winter overnight period at freeway-adjacent-USC (normalized ratio � 1.53), and to a lesser degree at RIV (normalized ratio � 1.13), further suggest the contribution of particle re-suspension induced by heavy-duty vehicles as a source of these species. • CM+TE /24-hr Avg CM+TE Summer I I� VA I 24-hrAvg VA Summer I 6 CM+TE/24-hrAvg CM+TEWrnter VA I 24-hrAvg VA Wmter ' • • • • 6 6 0 • 6 6 6 " • • !\ • 6 6 6 • • • 6 6 � 6 6 6 6 • " • " • • 6 • 6 6 !\ • § � § � § � ! � � ! � � ! § f 1 j § f 1 j § f 1 § � § � § > " � > " � > " � > " � > " � > " � LAN usc RIV LAN usc RIV I � SS I 24-hr Avg SS Summer I '-' ss 1 24-hr Avg ss Wmter 1 ' • WSOC/ 24-hrAvg WSOC Summer I 6 WSOC I 24-hr Avg WSOC Wmter • • • 6 • • 6 '" 6 • � • • 6 6 6 !\ 6 6 6 • • 6 • 6 • • 0 6 6 • • 6 6 • 6 6 6 • • • " 6 � 8 � � 8 � � 8 � - j � - j � - j � § § § > " 0 > " 0 > " 0 � � 8 � � � 8 � � � 8 � § j � § j � § j � > " 0 > " 0 > " 0 LAN usc �v LAN usc RIV Figure 5-4: Diurnal profiles of: (a) crustal materials and trace elements; (b) vehicle abrasion; (c) water soluble organic carbon and (d) sea salt. Normalization to the 24-hr average is presented. Water soluble org anic carbon. WSOC has been shown to be assoc iated with the production of reactive oxygen species in biological cells, which play a sig nificant role in PM-induced health eff ects . Despite detailed investigation ofWSOC in many studies, very limited literature exists 94 that addr essed the sources of WSOC in the CPM mode. Some studies indicated that organic materials in this size range consist of biological mate rials, semi-volatile organic species, organic acids, as well as organic ions (Bauer et a!., 2008; Falkovich et a!., 2005). The diurnal trend of WSOC (Figure 5.4 c) may give us some insights on the possi ble sources of the water-soluble fr action of organic carbon in this basin. In summer, WSOC follows a similar diurnal trend with CM + TE and VA, suggesting that WSOC either share common origins with mineral and road dust, or that WSOC may be adsorbed or absorbed onto these dust particles and re-suspended subsequently during PM sampling. RIV and LAN experienced the highest WSOC concentrations in the afternoon in summer (normalized ratio � 2.08 and 3.38 for RIV and LAN, respectively), which indicates that the sources of WSOC in these inland areas might be ass ociated with upwind pollutants. This is further conf rrrn ed by the low levels of WSOC in the afternoon of the winter period, when the lower wind speed may limit the transport of upwind air pollutants, thereby eliminating the peak concentrations observed at RIV and LAN in summer. Sea salt. The diurnal profile of sea salt aerosols in CPM is shown in Figure 5.4 d. In summer, the strong prevailing onshore wind facilitated the transport of sea salt particles inland. The diurnal trend of sea salt concentration varied among the three locations depending on their relative distances from the ocean. The sea salt levels at USC, located around 20 krn east of the Pacific Ocean, started to inc rease in midday (normalized ratio � 0. 99) and remained high overnight (normalized ratio � 1.32), consistent with the high wind speed observed during daytime at that site. The Riverside site (RIV), which is around 70-100 krn inland of the coast and is along the air mass tra jectory of this basin, had the highest level of sea salt in the afternoon with a normalized ratio of I. 70. In the morning and midday, the regional transport of sea salt from coast to inland was 95 limited due to the low wind speed, resulting in lower sea salt concentrati ons. The high sea salt concentration observed overnight (normalized ratio � 1.11 ) might be resulting from the strong wind in the afternoon that transported particles fr om the coast to inland areas . We observed a similar diurnal pattern at the other inland site (Lancaster), with the highest level of sea salt in the afternoon (normalized ratio � 3.39 ). In winter, in addition to the lower mixing height in the overnight period, the high sea salt level at USC (normalized ratio � 1. 77) could be due to the relatively strong wind in the aftern oon. Furt hermore, given the hyg roscopic nature of marine aerosols, the higher relative humidity at night might have facilitated the growth of these particles fr om the upper range of fine PM to the lower range of the coarse fraction (Tursic et a!., 2006), thereby contributing to the high levels of coarse particulate sea salt aerosol in the overnight period. Secondary ions. Figure 5.5 (a-c) presents the diurnal profile of ammonium, nitrate and nss-s ulfate ions, which com posed the category of secondary ions. In general, secondary ions follow a similar pattern with sea salt, with higher concentrati ons in summer and less in winter except in the overnight period. In an urban atmosphere, the concentrations of both nitrate and sulfate depend on source strengths of their acid forms. HN0 3 and H2S04, the precursors of nitrate and sulf ate respec tively, are fo rmed secondarily fr om the emission of vehicular combustion in the Los Angeles Basin. The high correlation (R � 0.80) between sulf ate and nitrate further confirms that they share a common origin. In cooler months, less photochemical oxidation of S02 and N02, decreases the source strengths of sulfate and nitrate part icles. Therefore, levels of nitrate and sulfate were generally lower in winter than summer, with the exception of the overnight period. Ammonium, present in much lower concentrati ons, follows a similar trend with nitrate and 96 sulfate. The low ammonium level reported in this study is likely to be predominantly in the form of ammonium sulfate, which is a dominant species in fme PM, with an upper "tail" in its size distribution extending to the lower range of the coarse mode. (a) I � NH4+ I 24-hr Avg NH4+ Summer I NH4+ I 24-hr Avg NH4+ Wrnter (b) I � N03- I 24-hr Avg N03- Summer I N03- I 24-hr Avg N03- Wmter • & ' • !! • � , z 6 • 6 " 6 • 6 6 • • • • " • • • 6 6 • 6 • 6 " • 6 • 6 6 6 • 6 6 6 6 6 • " § � § � � � � � 8 � � 8 § � § � 1 � § � 1 � § § j § j z " z " � z " � z " z " z " LAN usc �v LAN usc RIV '" I • nss-804-- 1 24-hrAvg nss-804-- Summer I lC, nss-804-- 124-hrAvg nss-804--Wmter ' ' " • !! � , 6 z 6 • • 6 • • • • • 6 • !I 6 6 • !, 6 6 § � § � § � § � 1 � § � 1 � § � 1 � z " � 0 z " � 0 z " � 0 LAN usc RIV Figure 5-5: Diurnal profiles of: (a) ammonium; (b) nitrate and (c) non-sea salt sulfate. Normalization to the 24-hr average is presented. 6 • In the winter overnight period at USC, the concentration of nitrate reached 705 ng/m 3 . This high level likely results from the reaction of nitric acid with sea salt aer osols, which had a similar peak in the same time period. The reaction of nitric acid with sea salt to form sodium nitrate, a process known as chloride depletion, is observed previously in this basin (Hughes et a!., 1999), 97 and will be discussed in more detail in the next section. Further inland at the Ri verside site (RIV), the overnight peak of the inorganic ions was more significant than those for the sea salt aeros ol and PM mass, suggesting that in addition to the reaction with sea salt, other fo rmation mechanisms might be present at this site. With the abundant crustal materials detected overnight, it is possible that these secondary ions react on the surface of mineral dust particles (Usher et a!., 2003). At LAN, the trends of secondary ions follow closely with sea salt and crustal materials in summer. As mentioned in the previous section, wind speed is the ma jor mechanism re-suspending PM in Lancaster. In winter, the low wind speed prevents regional transport, coupled with the lack of local precursors of acidic gases, contributes to the lower concentrations of inorganic aerosols at Lancaster. Overall, the overnight peaks for inorganic species are more significant compared with the corresponding PM mass, suggesting that in addition to the lower mixing height and/ or stagnation condition s, the meteor ological parameters during nighttime might facilitate the fo rmation and I or re-entrai rnn ent of these inorganic ions. Furt hermore, nitrate containing dust particles of> I f!m were also fo und to be highly hydrophilic (Shi et a!., 2008). The uptake of nitric acid on mineral dust particles tends to increase with increasing relative humidity (Vlasenko et a!., 2006), which is consistent with the high levels of nitrate and mineral dust at the Ri verside site overnight in this study. On the other hand, the gas/particle partitioning of ammonium salts is highly dependent on atmospheric conditions (Sei nfeld and Pandi s, 2006). The lower temperature at night fa vors the fo rmation of ammonium salts in the particle phase, consistent with the higher levels of these inorganic ions observed in the overnight period. 98 5.4.4. Seasonal and spatial correlations Table 5-3: Correlation coefficient between selected species in: (a) summer and (b) winter. so• ,, Dy Yb WSM< RH ,, .. wsoc c; N03· ,, Dy Yb WSM< WSC• Cl N03- SQ._. 0.51 0.68 0 . 63 0.52 0.65 0.65 0.82 0.90 oso NH4+ K+ AI. 0.33 0.96 088 O.:Y 0.60 0.45 0.� 0.70 OS7 o.n on 0.68 O.til on 0.59 o.43 0.36 025 088 � cu 0.70 0.68 O.SS 0.44 0.44 O.lS 0.49 0.48 0.52 0.62 0.62 0.41 0.44 0.44 031 024 0.25 0.� 0.73 0.71 0.84 "" .. 0 . .55 0.62 O.:.D 0.52 0.32 0.50 O.Jfi 0.65 0.18 0.50 "' Dy 0.71 0.77 0.13 038 035 0.34 0.46 0.48 0.47 058 055 0� 038 035 0.36 Yb WSN; WSCt WISNa RH Temp VecW 0,8:1 0.84 0.79 027 0.12 0.13 0.13 0.34 0.11 0.68 0 26 0 .37 - 0.21 0.23 0.48 0.91 0.17 0.19 0.43 -O.ll 0.45 0..52 0.90 0.86 029 0.31 -0.07 0j) 0.34 0.88 0.74 025 0.46 -0.27 O.Jil 0.16 0.27 021 020 0.17 0.17 M6 0.49 037 0.44 -0.25 0.25 0.48 0.67 0.75 0.19 0.'19 0.82 0.89 0.86 028 0.22 0.02 o.os 0.72 0.88 094 097 0.95 0.97 0.78 0.86 0.61 0 ()¢ 0.28 0.2:5 � 93 0� 0.815 o.u o::u o.n on 098 096 o.93 o.so 0.69 o.87 0.67 0.73 0.97 098 095 0.93 0.89 0.68 0.86 0.75 -0.10 O.Jil O.lS 0.76 -0.00 0.38 0.:?6 "' 029 0.81 0.64 036 0.91 0.50 0.82 O.SO 0.84 0.84 Q_� 092 091 0.24 0.89 069 0 . .55 0.69 033 0.92 0.72 0.81 0.86 098 0.89 0.98 0.68 on o . .ss o.90 086 o.78 o.86 059 0.23 0.82 0.66 0.57 0.65 0.40 0_49 0.85 0.88 0.87 079 081 O.t& 0.84 0.48 � 0.97 0.74 ' LQ,H_ 093 0.16 0.63 0.66 071 0.78 0 73 0.79 0.67 0.69 0 . 44 0.22 0.12 000 0.11 0.68 0.74 0.64 0.65 0.41 0.56 0.61 -0.2:5 0 . .55 0.44 093 089 0.87 0.83 0.69 0.88 0.78 -0.01 0.31 0.42 Dy Yb WS Mg WS C• n<<N• 00¢ 0.25 0.12 -0.82 -0.23 0.44 0.69 0.63 090 0.94 0.84 0 . .55 0.!9 O . .:P 0.90 051 0.41 0.64 0 ::8 0.516 Tn np J:qh -O.D3 .Q_)') -024 0.19 0.71 0.92 093 0.86 0.97 0.6:: 0.89 0.94 0.77 0.98 0.98 0.86 0 � 0.64 003 O . .U -0.15 -O.W 0.37 -0.14 0.0+ 0.71 0.92 0.18 O."".U 0.83 0.5:2 0.99 0.!9 0.84 053 0.43 0.61 O.J3 0.516 o_64 o.63 on o.92 o.89 o. 7 4 o.48 0.61 0.92 0.12 0.58 O.S2 0 . .55 0.91 027 -O.ol O.JJ 0.34 -0.14 -0.22 0.91 0.00 0.52 -0.19 0.32 o_74 o.ss 093 o.89 0.99 0.12 o.83 o.90 o.42 035 .o.o1 oro 0.91 0.82 096 O.SO 0.91 0.69 0.15 0.84 0.41 0. :::9 -OD2 -O.O:S 0.78 0.93 091 0.88 0.97 0. 7 4 0.83 0.90 Q_j) 0.49 -0.19 -0.04 I..M!L 063 -0 .18 082 0.8J 0.74 0.53 0.56 0.65 033 0.19 080 0. 6:: 0.11 0.61 O."XJ 0.74 0.41 0.25 0.84 0.� 0.86 0.64 0.93 0.92 0.49 0.56 OD6 -0.02 ODl 0.16 -023 O.l2 � 95 071 0.73 0.85 0.47 0.44 -023 -0.18 o . .ss o.os 0.18 o. :ll 0.24 .om -0.33 O."XJ 0.8:> 0.88 0.00 0.37 -OD8 -O.o7 o_45 OSfi om 0.33 -0.10 0� o.m 0.20 -0.00 -0.78 QQ;I 0.3J Table 5.3 presents the correlation coefficient (R) of selected individual species in summer and winter using the pooled data from the three sampling sites. In both seasons, tracers of vehicle abrasion (Cu, Zn and Ba) are highly correlated with soil dust tracers of AI, Ti and Fe (R values range from 0.67 to 0.97) and also the earth elements ofRb and Cs (R values range from 0.53 to 0.93), highlighting their common origins of re-suspended dust. Sodium ion, on the other hand, is highly correlated with chloride ion, (R= 0.82 and 0.98 for summer and winter, respectively), confirming that sodium ion is a reliable tracer of sea salt aer osols. Table 5.4 presents the correlation analys is, structured by location, using pooled summer and winter data points. Non-sea-salt sodium (nss-Na), calculated as the difference between WS Na and sea-salt fraction ofNa estimated using Cl ion, represents aged aero sols with sea salt origins . Nss-Na exhibited 99 high correlations with inorganic ions and low correlations with soil dust tracers at USC, and to a lesser extent RIV, suggesting that the non-sea-salt fraction of sodium is mainly derived from the reaction with these ions. In the site of Lancaster, CPM is primarily re-suspended by wind, as evidenced by the high correlations between individual species and wind speed (R values range fr om 0.64 to 0.89). The effe ct of relative humidity and temperature on these CPM species are also more significant in Lancaster compared with the other two sites, highlighting that meteorology plays a significant role in ambient coarse particle levels in that area. Table 5-4: Correlation coefficient between selected species in: (a) Lancaster (LAN); (b) Los Angeles (USC) and (c) Riverside (RIV). �""'-' Cl- N03- SOl- - N •+ NH4+ A1 Ti p, ,c u Zn B• Rb c, nssM nssCa nssNa. RH Tem p l(b) w.;oc Cl- N03- 3)4. . N •+ NH4+ A1 Ti p, Cu Zn B• Rb c, nssMg nssCa nssNa RH Tem p ' w.DC " �. N•+ NH 4+ A! i Fo [ z � "' Rb r · nssMg ns!Ca nssNa RH ' "" U.l. C< 01 �;; s ns g nss a nss a ��:�;��:��;:�;�� =�:� ��:; � � � � = � � � � = = � � � � � m � � � m = u � � = � m - � � � m 0.86 0.72 083 082 041 0.32 0.72 080 064 0.59 0.95 1.00 ..042 048 0.72 � = = m u � � � = � = � � m ""' �-� �;� �� ; �-� �-� ::; :;; 091 0.90 083 0.87 0.72 005 OJQ 0.64 0.69 t..22L �36 052 0.82 0.9 5 �-" 0.38 0.77 ..041 048 0.72 -U97 -0 5 0.78 N<Jj. :SU4- N •+ NH.,._ AI ' ' u Ln "' Hb C> nssMg nss t..: a nssNa. HH ' "" w.; , _,. " "' 0 . 4 UH u.oo 0.43 04>l )<5() 034 0.40 0.51 0"' " '"' U<. 0.52 002 003 0.36 0<50 � 048 048 057 049 044 0.63 054 053 0 .7 4 0.65 045 036 -025 �0.16 o.as o.94 o.n 0.70 0.66 0.7 0.62 055 0.59 0.73 0.70 0.95 0.79 0.95 0.13 0.04 0.49 0.81 0.81 0.12 O.D3 OIJI 0.18 000 O.DI 0.17 0.12 0.92 0.53 0.93 0.18 0.22 0 . ... 0.74 0.16 0.14 Ol!l 0.17 009 0.21 0)5 024 095 0.55 0.'/1 037 .Q.Ql 025 -0.18 -0.22 -0.18 -0.14 -0.21 -0.16 -0.12 -0.13 O.Ol 0.26 0.83 0.42 -0.07 0.12 � 091 092 0.92 0.10 :.� O.Ol 0.86 O .lll �-04 -0.19 0.16 0.87 097 0.92 ow 0.19 0.79 0.01 .0.11 -0.15 0.17 0.88 094 0.98 O.Ol 0.99 O J 8 0.8.5 O.C>; 0.04 -0.28 0.06 � �-� � � � � �;: -�% �.� : � \'; 0.13 0.13 092 095 o_., 082 005 Ol!l -044 -0.11 L.R& O.lli5 0.87 0.13 �-06 -0.12 O,l!l 0 . 78 0.84 0.12 .om -0.19 0.11 � 0 � [)21 -Q03 [).38 0.45 0.12 -0.1.5 0.18 [)J2 006 Q.34 -0'1 -0.48 0..54 •+ ' ' u n . ' � · n" . n" . = � 0. 94 0.98 om ozz 0.6 O.ll <50 0 . .57 0.17 061 .61 093 0.87 0.97 -0.19 0.44 0..5.5 0.93 0.94 O.Jl 093 0.73 0.71 0.54 O.lO 0 . 42 0.70 0.75 0.93 0.94 0.89 o.rn 029 0 . 47 0.73 0.70 0.62 0.78 0.63 0.75 0.4>l 0.47 0 . 62 056 0.66 0.72 0.81 0.63 0.12 0.14 0 . 42 098 -O.CO 095 IJ73 061 04& 03& 0.19 078 0.69 -O.CO � 069 054 0..53 o . .. 014 070 0.65 0.32 0.56 0.42 O.lO "' ' 008 0.34 � 0.58 o ... 042 ""' "·" 9 0.81 0.67 0.66 091 0.89 0.62 0..51 Olli5 087 0.91 UL : � :;; :� 052 0.68 � 099 0.88 097 � 0 � 0 . 00 0.29 -0.13 U.>'<> "·" "·"' 0.79 0.82 0.66 0.68 0.81 048 u_, :� " -" 0.41 0.43 0..54 0 . 54 O.D5 � : � : � -� -0.13 -0.17 0.<3 om 0.35 053 0.1 0.10 0.75 0.31 0.52 -0.04 0:!> -0.12 051 058 04& 056 -0.47 .(1.19 037 0.3& 0.12 0.06 -006 .(10.5 :� :� -0.47 .(1.35 015 0.12 -008 �08 0.46 04>l 0.10 0.19 053 0.58 -0.84 .(1.78 r o-w 100 5.4.5. Chloride depletion Chloride depletion plays an important role in atmospheric processes as it changes the deliquescence points and optical properties of coarse particles (Finlayson-Pitts, 2003). Many studies have reported the depletion of chloride along coastal areas . Yet, the fa ctors aff ecting the degree of depletion are not completely understood. Zhuang et al. (1999) attributed the depletion of sea salt to both HN03 and H 2 S04 in Hong Kong, and observed that the rate of depletion was dependent on relative humidity, particle size and the relative abundance ofCa 2 + and Na+. Another study in Newark, New Jersey concluded that nitrate, and to a lesser extent sulfate and organic acids, were responsible for chloride depletion in that area, and less depletion was observed when the air masses were directly originated from the Atlantic Ocean (Zhao and Gao, 2008a). In th is study, the degree of chloride depletion was examined at the three sampling sites in different periods of the day, enabling us to understand the fa ctors that influence th is depletion process. & MorncnstO·)t{NOH e Ovcn,.&l'lt IO·J•(NOJ.I Ahcmoon (CII :1: • • X M1dd<�v (CI-)•(NO).) A Morn.ng(O) e Ovffn•Jhl(O) v 0.24 1<·028 R1•0.93 10 12 14 16 18 20 (N;u) neq/m1 25 20 � A Motn•ns (O·)t(NOJ·) • Ov«ncl'lt (O·JotN03·) • Ahemoon [CI-) • .. v•044.--035 R1 095 )( M ldd.Jy (CI·J•I"Ol•) • Momllll (Q-) e Ovt'ffM& ht (O·I 10 ll 14 16 I. W (Na+)neq/m' Figure 5-6: Chloride depletion and nitrate replacement scatter-plots in a) summer and b) winter. Dark legends refer to the sum of [Cl-] + [N03 -], and gray legends refer to [Cl-] . Figure 5. 6 (a and b) shows scatter plots of [Cl.] against [Na +], in the unit of molar equivalent concentrati ons. The average observed [Cl.]/ [Na+] equivalent ratio is lower in summer (slope = 101 0.24) compared with winter (slope � 0.4 4), indicating that more depletion is observed in summer due to enhanced photochemistry and higher concentrations of acidic gas-phase species under higher temperature. Less depletion, as shown by the higher [Cl"]I[Na l equivalent ratio, as presented in Figure 5.6 and 5. 7, was observed in the overnight sampling period in both seasons due to lower levels of acidic gases at night (Wall et a!., 1988), which further conf mns that the concentration of acidic gas is an important parameter influencing the degree of depletion in this basin. As depletion primarily results from nitric acid reacting with sodium chloride to form sodium nitrate and hydr ochloric acid, the ratios of the sum of [Cl"] and [NO;] I [Na +] is also plotted as the dark legend in Figure 5.6. The regr ession slope in summer (slope � 0. 99) falls close to the theoretical seawater ratio of [Cl"]I[Na+] (slope � 1.17), with a R 2 of0.85, confirming the role of nitrate in chloride depletion, and that nitrate in CPM is predomina ntly formed by the depletion reaction in summer. The slightly lower slope indicates that a minor fr action of the depletion might be caused by other species such as nss-so.z· and organic acids, as shown in other studies (Zhuang et a!., 1999). In winter, both the sea salt concentration and the rate of depletion were lower. The regression slope of sum of [Cl"] and [NO;] I [Na +] is 1.40, indicating that excess nitrate, generated by means other than sea salt depletion, is present. In addition to chloride depletion, the two other ma jor fo rmation mechanisms of coarse mode nitrate particles are: I) reactions of nitric acid on soil dust PM, and 2) condensation of ammonium nitrate on CPM surf aces, which would be expected in cooler winter mont hs. Noble and Prather (19 96) demonstrated the associ ation of both marine and soil particles to super-micron nitrate particles in Rive rside, CA. Many studies have also observed the presence of sulfate and nitrate on 102 soil particles (Wu and Okada, 1994; Zhang et a!., 2000; Zhuang et a!., 1999). A study conducted in Italy suggested that NO; shif ts to the super-micron range in episodes of high coarse dust concentrations by adsorption ofHN0 3 onto dust particles (Putaud et a!., 2004). The abundance of both soil and sea salt particles would therefore explain the high nitrate concentrations in this study. Figure 5.7 shows the diurnal concentrations of excess ammonium, as well as equivalent ratios of [Cl"] I [Na+] and sum of [Cl"] and [NO;] I [Na+]. When sea salt concentration was high, the [Cl"] + [NO;] I [Na +] ratio was within 20% of the theoretical value of 1.17. In some periods when sea salt concentration was low in summer (LAN midday, PIU morning, RIV morning and midday), a small fr action of excess nitrate was observed, probably due to the analytical uncertainty associated with low concentra tions . In summer, ammonium concentrations were not high enough to fully neutralize sulf ate in CPM with the exception of the USC overnight period. In winter, NIL+ I nssso .z· > 2 for most of the time periods, indicating the presence of excess ammonium, as illustrated in Figure 5. 7b. At RIV, high levels of excess nitrate were observed (equivalent ratio of the sum of [Cl"] + [NO;] I [Nal > 2) when levels of excess ammonium were also high, suggesting the presence of ammonium nitrate at that site. As previously mentioned, many studies observed the uptake of nitric acid on mineral dust particles, particularly in the form of Ca(N0 3 ) 2 and Mg(N0 3 ) 2 (Usher et a!., 2003). In this study, levels of nss-Ca (on average 97.6 nglm3 and 78.9 nglm3 in summer and winter respe ctively) were much higher than nss-Mg (on average 8. 70 nglm3 and 4.22 nglm3 in summer and winter respectively). Correlation coefficient (R) between nitrate and nss-Ca and nss-Na is 0.81 and 0.63 at RIV resp ectively, suggesting that both chloride depletion and the reaction with soil particles are important fo rmation mechanisms 103 of coarse particulate nitrate at this site. These results imply that the excess nitrate present in winter at Riverside can be in the form of soil-assoc iated calcium nitrate, as well as ammonium nitrate, in addition to NaN0 3 generated by sea salt depletion. 5 Tf (a i) ) -- � ==== =;� ---- � 3 0 lc::::J ExcessArrmoni um!on I • JCI-J4Na+J 6 JCI-J<JN03-V!Na>J 10 b) I = Excess Ammo"'" ion I • JCI-�Na<J f'.. [CI-J<JN03-�Na+J 2.5 ['.. 20 0 � ct ['.. ['.. ['.. [',. ['.. ['.. �� . . 1-l �"" � ,Jel 6 6 15 � 6 � [',. � 6 6 6 <7 6 1.0 w 6 [',. 6 ·R 05 • • • • • • • • • • 0.0 g> f 8 � g> f 8 � ·s E E ·s E E J!l � J!l � " " :,: " " :,: c � c c c � g> � g g> � g "' g> � g E E E s :g J!l � s :g J!l � s :g J!l � " :1 :;: " :1 :;: " i :;: LAN usc RIV LAN usc 6 ['.. ['.. �r.l • g> f 8 s E J!l " " :,: RIV ['.. • � E � 0 3 � ct c "' ro 2 -� w Figure 5-7: Diurnal profile of excess ammonium ion concentration (nglm3) and molar equivalent ratios of [Cl-] I [Na+] and sum of [Cl-] and [N03-] I [Na+] in: (a) summer and (b) winter. 5. 5. Summary and Conclus ions The diurnal profile of chemical groups and constituents of CPM differs sig nificantly in summer and winter. Highest concentrations of CPM gravimetric mass, crustal materials and trace elements, vehicle abrasion, and water soluble organic carbon were observed in summer midday and I or afternoon when the wind speed was higher for all the three sampling sites . On the other hand, high levels of sea salt particles and secondary ions were experienced in the summer overnight period at USC. In winter, the Los Angeles Basin is characterized by fr equent stagnation condition s, and vehicle-induced turbulence becomes a significant mechanism of particle re-suspension when mixing height was low, as highlighted by the high concentrations of CPM mass, as well as soil and road dust in the winter overnight period at the near fr eeway sampling sites. Nitrate, which has been shown to be the most abundant species in the CPM, is 104 predominantly fo rmed by chloride depletion in summer. A significant nitrate peak is observed in the afternoon at the inland sites (LAN and RIV), consistent with the diurnal profile of sea salt concen trations. In winter, high levels of nitrate were fo und during overnight. Due to the overall lower sea salt levels in winter, the reactions with sea salt and mineral dust particles, as well as the condensation of ammonium nitrate on PM surfaces are the dominant fo rmation mechanisms of nitrate in the coarse particles. 5. 6. Acknowledgements This study was funded by the Science to Achieve Results program of the United States Environmental Protection Agency (EPA-G2006-ST AR -Ql). The authors would also like to thank the staff at the Wisconsin State Laboratory of Hygiene for the ass istance with chemical analys es. 105 Chapter 6 Diurnal Trends in Oxidative Potential of Coarse Particula te Matter in the Los Angeles Basin and Their Relation to Sources and Chemical Com position 6.1. Abstract To investigate the relationship between sources, chemical composition and redox activity of coarse particulate matter (CPM), three sampling sites were setup up in the Los Angeles Basin to collect ambient coarse particles at four time periods (morning, midday, afternoon and overnight) in summer 2009 and winter 2010. The generation of reactive oxygen species (ROS) was used to assess the redox activity of these particles. Our results present distinct diurnal profiles of CPM-induced ROS formation in the two seasons, with much higher levels in summer than winter. Higher ROS activity was observed in the midday I afternoon during summertime, while the peak activity occurred in the overnight period in winter. Crustal materi als, the major component of CPM, demonstrated very low water-solubility, in contrast with the modestly water- soluble anth ropogenic metals, including Ba and Cu. The wate r-soluble fr action of four elements (V, Pd, Cu and Rh) with primary ant hropogenic origins displayed the highest ass ociations with ROS activity (R 2 >0.60). Our results show that coarse particles generated by anth ropogenic activities, despite their low contribution to CPM mass, are important to the biological activity of CPM, and that a more targeted control strategy may be needed to protect the public health fr om these toxic CPM sources . 6.2. Introduction A review article of over 30 epidemiological studies on the effect off me and coarse particles on mortality and morbidity revealed evidence that CPM had similar, or stronger short-term effe cts as fine PM on asthma and respiratory hospital admissions, as well as chronic obst ructive 106 puhnonary disease (Brunekreef and Forsberg, 2005). In particular, the investigators called for consideration in studying and regulating CPM separately fr om fine particles. Whereas the temporal, spatial and toxicological characteristics associated with PM 2 .5 and PM10 is more thoroughly investigated, there is a dearth of equivalent information on coarse particles. To develop cost- effe ctive air quality regulations for protecting the public health fr om CPM exposure, it is essential to establish the linkage between their sources and chemical composition to toxicity. Although a number of studies investigated the relationship between sources and toxicity in ultrafine PM, PM 2 5 and PM10 (Hu et a!., 2008; Shafer et a!., 2010; Verma et a!., 2009a), very limited literature has examined the effects of CPM indep endently. Unlike the dominance of organic materials and secondary ions in fm e particles (Arhami et a!., 2009; Ning et a!., 2007), metals and elements are the core components of CPM in both rural and urban enviro rnn ents (Cheung et a!., 20llb; Hueglin et a!., 2005; Sillanpaa et a!., 2006). Furthermore, metals in the coarse fr action primarily arise fr om mineral and road dust, in contrast to the combustion origins of the ultrafine and accumulation particles. Since chemical characteristics directly affe cts the generation ofPM-assoc iated toxicity (Cho et a!., 2005; Schwarze et a!., 2007; Shafer et a!., 2010), the different chemical composition of CPM might induce toxicity differently than other fr actions of PM. The generation of ROS and the subsequent induction of oxidative stress plays a signif icant role in adverse health outcomes related to particle exposure (Nel, 2005; Tao eta!., 2003). A number of studies have shown that PM components can induce pro-inflammatory responses in human airways through their ability to form ROS (Barnes, 1990; Nel et a!., 200 1). The presence of transition metals could enhance the fo rmation of hydroxyl radical- a strong oxidizing species, 107 via Fenton or Fenton-like reactio ns. In particular, water-soluble species, especially metals, have been shown to be the key drivers ofROS generation (Goldsmith et a!., 1998; Prophete et a!., 2006). On the other hand, key atmospheric parameters such as mixing height, wind speed and direction, as well as emis sion sources and their strengths all vary in scales shorter than 24 hou rs, which could result in the diurnal variation of chemical composition and thus redox activity. The temporal variability of CPM-induced toxicity has important implications in terms of exposure and risk asses sment of CPM exposure, and will help determine the actual impact of CPM toxicity on public health. In this study, CPM was collected at four daily time periods (morning, midday, afternoon and overnight) at three distinct sampling sites to study the diurnal characteristics of chemical composition and oxidative potential of ambient coarse particles. The CPM-induced generation ofROS was measured by an in-vitro bioassay and used to evaluate the toxic activity of these particles. The goals of this study are to determine the diurnal variability of CPM-induced oxidative potential and to identify specific CPM species I source classes (if any) that drive the ROS activity. This information will ultimately help the regulatory communities to desigu more targeted and effective control strategies to protect the public health fr om CPM exposure. 6.3. Methodology Details of the sampling location, equipment and methods are described in the previous chapter (Chapter 5). 6.3. 1. Chemical analyses To determine the concentration of total metals and elements, the substrates were digested in an acid mixture (1.0 rnL of 16 N HN0 3 , 0. 1 rnL of 28 N HF, and 0.25 rnL of 12 N HCl) in a Teflon digestion bomb using a microwave -assisted digestion unit. The digestates were 108 subs equently analyzed by high resolution inductively coupled plasma-mass spectro scopy techniques (HR-ICPMS) (S hafer et a!., 2010). The levels of water-soluble metals and elements were quantified using the same HR-ICPMS technique, with the ext raction conducted using 10 rnL ofMilli-Q water (Millipore, Bedford, MA, USA) fo llowed by filtration using 0.45 f!m filt ers. The water extracts were also analyzed by a Sievers total organic carbon analyzer (General Electric, Inc.) (Zhang et a!., 2008b) and ion chromatography (Lough et a!., 2005) to determine the levels of water-soluble organic carbon (WSOC) and water-soluble ions, resp ectively. The generation of ROS was quantified using an in-vitro assay described in greater detail elsewhere (Landreman et a!., 2008; Zhang et a!., 2008b ). Briefly, substrates were extr acted with 1.0 rnL ofMilli-Q water for 16 hours, in the dark, on a shaker table. After removing the filters, samples were centrifuged at 6600 RPM for I minute and the supernatant were filtered through 0.22 f!m polypropylene syringe filte rs. The aqueous solutions were then buff ered using !Ox concentrated Salts Gluco se Medium. The buffered PM extract solutions were subsequently split to prepare aliquots of diluted and non-diluted samples in triplicates. The aliquots were mixed with 2'7'- dichlorofluore scin diacetate and then added to rat alveolar macrophage cells (cell line NR8383) that were previously plated into 96-well plates, and incubated at 37'C for 2 hours. The plates were read 5 times (at 0, 30, 60, 90, 120 minutes) using a Cytoflour II automated fluor escence plate reader at a wavelength of 485/530nrn. The samples were analyzed along with posi tive (Zymosan, urban dust extracts) and negative (method blanks) controls. ROS activity was reported as the increase in the flu orescence intensity of the PM samples relative to that observed in the cont rols. The results were reported in units of Zymosan equivalen ts. Samples were analyzed with laboratory and field blanks, and all reported values have been method and field 109 blank corrected. On average, field blank levels contributed to less than 2% of the samples ' levels, with the exception of four water-soluble trace elements (Th, Lu, Cs and Sn), which experienced high field blanks concen trations, and were excluded in the analys is. Total uncertainties were determined based on the analytical uncertainties and uncertainti es of field blanks . 6.4. Results and Discussion 6.4. 1. Overview Chemical species were categorized into : (1) crustal materials and trace elements (CM + TE); (2) abrasive vehicular emissions (A VE); (3) water-soluble organic carbon (WSOC); ( 4) sea salt (SS) and (5) secondary ions (SI). Their diurnal profiles have been presented in the Chapter 5, and only a brief summary is described here. It should be noted that although coarse particulate Fe is a major soil constituent, it could be enriched by anthropogenic sources in urban areas (Gietl et a!., 2010). The inclusion of Fe in the crustal material group might have overestimated the contribution of soil fr om Fe. Overall, the group of CM + TE, primarily com posed of crustal elements of Al, Fe and Si, was the most abundant category of CPM at all three sampling sites (avg.�4.20 flg/m3 ). Levels of these crustal materials were higher at RIV (avg.�7.7 flg/m3) than LAN (av g.�2.0 flg/m3) I USC ( avg. �3.4 flg/ m3), due to the semi-rural receptor nature of the area. Abrasive vehicular emissions refer to the emissions fr om the abrasion processes of tire wear, brake linings, catalyst deterioration, etc. Metals and elements that originated predominantly fr om vehicular abrasion (Cu, Ba and Zn) were much less abundant (a vg.�O .l9 flg/m3 ). Sea salt levels were higher in summer (avg.�0.53 flg/m3) than winter (avg.�0.28 flg/m3 ). Secondary ions, primarily com posed of nitrate, and to a lesser extent sulfate and ammonium, had an overall average of0.50 flg/m3, with lower 110 concentrations at LAN. WSOC concentrations were low at LAN (avg.=0.071 !lg/m3 ) and USC (avg. =0.078 11g/m3 ), with higher levels at RIV (avg.=0.65 11g/m3 ). The diurnal profiles of CPM differ substantially in summer and winter (Figure 6. 1), due to the combined effe cts of different source strengt hs, meteorology and primary re-suspension mechan isms. Levels ofCPM mass, mineral and road dust, as well as WSOC were highest in summer midday I afternoon, concurrent with the higher wind speed during those periods . In winter, atmospheric dilution was lower due to stagnation conditions during the sampling period, resulting in the accumulation of air po llutants . In particular, turbulence induced by vehicular movement (especially heavy-duty vehicles) became a ma jor re-suspension mechanism of CPM in the calm and stagnant atmosphere during winter nighttime (Cheung et al., 2011c ). As a result, levels of mineral and road dust were elevated at the near-f reeway USC site, and to a lesser extent the RIV site during winter overnight. This finding is consistent with many other studies that showed vehicular movement could contribute substantially to the re-entrainment of coarse mode aerosols (Charron and Harris on, 2005; Harri son et al., 200 1; Pakbin et al., 2010). r:::z:::J Crusta\materialandtrace element C=:l Abrasive 11ehicular emssion c::::JJ Water·sol uble organic carbon 12.0 (a) -S eas art ! 1 10 0 c:::::J Secondary ion 8.0 I II "E 6 0 "' 4.0 � -.'0 ll [ c 2 0 lT 0 t � t t 1! 12 t t 8 t t 0.8 t t t 0.4 t t t 0 0 t 2' >- 8 13, 2' >- 8 13, 2' >- 8 13, E g i � � g � � E g � � 6 6 � � " " " li � " li 0 LAN usc RIV 8.0 (b) 7.0 6.0 5.0 M- 4 .0 � 30 ::!.. 2.0 T1 § 1.0 H H lT ll e � 0.5 8 0. 4 03 02 0.1 0.0 +----J-IJ,RLw,L-.JJ.J,al...LU,II.ljill,llL!JJ,aLill\RL rn ! 8 §, c � · � � 8 " ., li LAN usc RIV Figure 6-1: Diurnal profile of chemical composition in: (a) summer and (b) winter. Error bars represent analytical uncertainti es. 111 6.4.2. Water solubility of elements Particle solubility is an important phys ico-chemical property that affe cts its bioavailability to human cells (Costa and Dreher, 1997). The water-soluble components of particles are more easily bioavailable, and they have been shown to be the key drivers ofPM-induced toxicity in both in-vivo (Roberts et al., 2007) and in-vitro studies (Goldsmith et al., 1998). Figure 6.2 shows the overall water solubility of selected metals and elements of the CPM sampled in this study, calculated as the ratio of the water-soluble to the total elemental concentration for each species. In general, the trend of water solubility was similar seasonally and spatially, and the first, second and third quartiles of solubility are presented. Species with more than 25% of data (i.e. 2> 7 out of 24 data points) under detection limit (2 x total uncertai nties) were not included in this analys is. Using such criteria, 12 out of 45 elements, mostly rare earth elements, were excluded due to the trace levels of the water-soluble fr action of these species. As seen in Figure 6.2, elements with sea salt origins, such as sodium (Na), magnesium (Mg) and calcium (Ca), are quite soluble. Although not shown, the fr action of water- soluble Na was highest in the afternoon and overnight periods in summer, parallel to the high sea salt levels observed. The median wate r-soluble fr action ofNa was also highest at USC due to its proximity to the coast compared to RlV and LAN, leading to higher fr actions of sea salt particles at that site. While metals and elements are the dominant components of ambient coarse particles (Cheung et al., 20 llb; Sillanpaa et al., 2006), most of these crustal materials are insoluble in water. This is evident fr om the very low solubility of these crustal materials including AI (overall median�0.48%) and Ce (overall median� 0.60%). One particular exception would be Ca, since gypsum (CaS04 ·2H 2 0) is moderately water- soluble. Calcium nitrate, resulting fr om the reaction of mineral dust (or sea salt) and nitric acid, is highly 112 water-soluble as well (Cheung et al., 2011c; Usher et al., 2003). On the other hand, elements of anth ropogenic origins experienced moderate solubility, consistent with previous published work (Birmili et al., 2006). Metals primarily originated fr om brake wear, such as Ba and Cu, have a median solubility of 15% and 12%, respec tively. Zn, with several potential anth ropogenic sources including industrial emissions as well as tire and brake wear (Pakbin et al., 2011), also exhibited moderate solubility with a median of 21%. 1.0,-------------------------------, 0.8 +---------------------------------1 0.6 c .2 1l 0.4 .:: 0.2 0.0 Ca Na s Figure 6-2: Water solubility of selected metals and elements, calculated across three sites, four periods and two seasons. 18 1 quartile, median and 3r d quartile is shown. Species with > 25% of data points under detection limit (2 x total uncertain ties) were excluded. 6.4.3. ROS activity ROS is a collec tive term to describe species that contains oxygen and is highly reactive, such as oxygen and hydroxyl radicals. ROS activity, assessed by an in-vitro method in this study, represents the oxidative potential of ambient coarse particles. This cellular method provides a comprehensive evaluation ofPM-induced ROS activity as it takes account into both the direct (by the PM aqueous extracts) and indirect (resulting fr om cellular stimulation) fo rmation ofROS, and provides an assessment of total ROS activity resulting fr om hydroxyl radical, peroxide, 113 superoxide radical, and peroxynitrite (S hafer et al., 20 10). Note that ROS is naturally produced in cells as a "byproduct" of normal metabolic activity - and is held in check by cytoplasm antioxidants (e.g. glutat hione).This baseline activity is accounted for in the assay by a series of negative controls I blanks . The ROS activity of the PM measured I reported is that in excess of baseline and thereby allows for effective comparison of PM oxidative potenti al. We perform the assay under conditions where the ROS produced is not overtly toxic to the macrophage cells to ensure that cell viability issues do not impact our assessment of the relative and absolute oxidative potential of the PM. 50 (a) �- ROS Activly Sunmet I r:=:J ROS Activ�yWI11et � flt I� 11 � � 'i � c 15. 15. 0 � :; � � � � � � � � � 0 0 0 LAN usc RIY 2000 � Q_ rn 11500 :0 c rn � 1000 N :o!: � 500 � s (( (b) l � - ROS Activi� Sunmer c::::::J ROS Activity" Winter � l l I 'i 0 � � � � � 0 � LAN usc F; � h � E � E g � g RIV Figure 6-3: Diurnal profile ofROS activity on a: (a) air volume basis and (b) PM mass basis. Figure 6.3 (a-b) shows the diurnal profile ofROS activity (expressed per air volume and per PM mass basis) of the three sampling sites in summer and winter. ROS activity expressed per volume of air represents the toxicological activity imparted on the en vironment by these particles, which, in some respect, makes it more relevant in the context of population exp osure, compared to the more traditional representation of toxicity on a per PM mass basis. On a per air volume basis, ROS activity was generally higher in summer than winter. In summer, ROS activity peaked 114 in midday I afternoon. At USC, ROS activity was highest in midday at 26.2 fig Zymosan unitslm3, while the peak occurred in the afternoon at RIV (23.9 fig Zymosan unitslm3) and LAN (16.3 fig Zymosan unitslm3 ). In winter, higher ROS activity was observed overnight at USC (5.2 fig Zymosan unitslm3) and RIV (24.6 fig Zymosan unitslm3 ). This finding indicates significant diurnal and seasonal variations in the toxicity of ambient coarse particles, which should be considered in the air-quality related regulation and policy. In general, the diurnal trend ofROS activity is similar on both per mass and per volume basis, with the exception of USC in summer. The mass-based ROS activity of summer overnight ( 1440 fig Zymosan unitslmg of CPM) was at a similar level with the midday (1140 fig Zymosan unitslmg of CPM) and afternoon (11 90 fig Zymosan unitslmg of CPM) activity, in contrast to the lower overnight activity, compared to midday I afternoon, on a per volume basis. This suggests that the mass fr action of CPM emitted by sources that drive ROS generation was higher overnight, either due to an increase in their emission strengt hs, or to a reduction in the fr action of non-ROS active components . 6.4.4. Association between ROS activity and water-soluble elements To elucidate the relationship between ROS activity and individual elements (and possi bly classes of CPM sources), a linear regre ssion analys is was perf ormed. Table 6.1 shows the coefficient of determination (R-squared) between ROS (Zy mosan unitslm3) and the water- soluble (WS) fr action of selected elements (nglm3 ). High correlations (R 2 >0.60) were observed for WS vanadium (V), palladium (Pd), copper (Cu) and rhodium (Rh), which are indicative of anth ropogenic sources (Pakbin et a!., 2011). These high ass ociations were consistent in summer and winter, as illustrated in Fignre 6.4 using WS Cu as an example. On the other hand, lower ass ociations were found for WS Fe and AI, the tracers of crustal materials (Cheung et a!., 20 lib), 115 probably because of the very low water solubility of these elemen ts, and I or the low intri nsic redox properties of crustal elements in the coarse fr acti on. The regres sion analysis was repeated using concentrations of total elements as independent variables. Although not shown, the coefficients of determination between ROS activity and total metals all fall below 0.56. The overall lower associations are consistent with the fact that the bioassay was conducted using filtered aqueous extracts, and only the water- soluble elements were exposed to the macrophage cells. Table 6-1: Coefficient of determination (R 2 ) between ROS activity and selected water-soluble (WS) elemen ts. VVS-Ba WS-N1 WS-Pb WS-Fe WS-T1 WS-AI ROS 0.16 0.23 0.03 0.42 0.14 0.16 ws.v 0 36 0 26 00 4 0 39 0 15 0 27 WS-Pd 0 27 0 26 0 06 0.35 0 11 03 4 WS-Cu 0.33 0.29 0.08 0.29 0.12 0.28 WS-Rh 0 14 03 4 0 11 0 25 007 05 3 WS-Ba 0.08 0.01 0.07 0.04 0.25 WS-Ni 0.05 0.04 0.19 WS-Pb 0.29 0 28 WS-Fe 0.17 WS-TI 00 6 Platinum group elements (PG Es) ofP d and Rh are commonly used in automobile catalytic converter s. They are believed to be emitted to the atmosphere fr om mechanical and thermal stresses in operation (Zereini et al., 2001), and their presence have been observed in both the fine and coarse fr actions of ambient PM in many other areas (Kanits ar et al., 2003; Rauch et al., 2001). Rauch et al. (2005) observed the presence of super-micron-PGEs (as a maj or or minor component on Al/Si oxide) in automobile exhaust and urban air in Sweden using scanning electron microscopy. In this study, the overall average concentration of total Pd and Rh was 0.0438 ± 0.041 ng/m 3 and 0.0138 ± 0.0 15 ng/m 3 , respectively, with higher concentrations observed during USC winter overnight and in the afternoon and overnight period at RIV in winter. Traffi c-induced turbulence might be a ma jor re-suspension mechanism of CPM during winter nighttime when 116 atmospheric dilution was lowest, leading to the high levels of the traffic-related emissions in these periods, as discus sed in a greater detail in Chapter 5. �E 25 � � 20 1 N 15 :J � " 1 0 � 0 " ,., . . y= 041x +76 R 2 =07 0 Water-soluble Cu (nglm ' ) ' " "' "" "' --- --:- -::- -:-:cc- --- --, y=04 7x-02 8 ' ! 1 5 2:t o j " 0 " " · . p 2 =09 5 Water -soluble Cu (ng/m ' ) Figure 6-4: Correlations between measured ROS activity and water-soluble Cu in: (a) summer and (b) winter. CPM- bound copper is typically produced by the wear of brake pads and linings (Garg et a!., 2000; Hjortenkrans et a!., 2007; Lin et a!., 2005). Cu is used as high-temperature lubricant in brake linings, and is emitted primarily fr om mechanical wear (Sanders et a!., 2003). The overall average concentration of total Cu was 154 (±150) ng/m 3 , and it could be considered as the most significant element with anthropogenic origins in CPM. Higher levels were observed at USC and RIV , where the sites are more heavily influenced by vehicular emissions and re-suspension. In particular, agricultural activities and vehicle movement on dirt, unpaved roads, or paved roads with unpaved shoulders, could generate and re-entrain a significant amount of road dust in semi-rural regions (Chow et a!., 1992), and might contribute to the high levels of Cu at the RIV site. The sources of vanadium are more dive rse. A recent study conducted in the Los Angeles Basin suggested that even in CPM, V could be generated from ship emissions (Pakbin et a!., 2011). A study in the Los Angeles-Long Beach Harbor area also revealed significant levels (ca. 117 1-2 ng/m 3 ) ofV in the accumulation and coarse PM mode, and suggested that V could be originated fr om emissions of bunker-fuel and oil combustion (Krudysz et al., 2008). Additionally, Schauer et al. (2006) demonstrated the presence of vanadium in brake wear and road dust. Overall, CPM-b ound vanadium, with an overall average concentration of0.93 ng/m 3 , could be generated fr om a variety of sources. The high correlations between the water- soluble fr action ofV and Pd (R 2 �0.86), to a lesser extent Cu (R 2 �0.67), suggest that water-soluble V could be largely originated fr om vehicular abrasive emissions in the coarse fr action of PM in this study. Certain organic compounds, such as quinones and nitro-P AHs, are capable of generating superoxide radicals (Cho et al., 2005), which are responsible for the recycling of the precursors of Fenton (or Fenton-like) reactions, and play a critical role in subsequent form ation of the highly reactive hydroxyl radicals . In particular, previous studies have demonstrated high correlations between WSOC I OC content of PM and the consumption of dithiothreitol (DTT), an assay that measures the generation of superoxide radicals (Verma et al., 2009a; Verma et al., 20 11). Nonet heless, these studies were conducted based on fine or quasi -ultrafine particles, and the role of organics on the toxicity of CPM is not well studied. While fine-particulate WSOC consists of polar organic compounds and could originate from vehicular emissions (Kawamura and Kaplan, 1987) and secondary form ation (Verma et al., 2009a), its source could be different in the coarse mode. In addition to the condensation of semi-volatile species on coarse particles, WSOC could be distinctly originated fr om biological materials, which is the maj or component of CPM- bound OC in this basin (Cheung et al., 20l lb). The WSOC level of this study was reported previously, and it was highest at RIV among the three sampling sites (Cheung et al., 20 llc ). Correlation between ROS and WSOC is moderate (R 2 �0.31) in this study. The site-specific correlation at RIV, 118 which experienced the highest WSOC concentrati ons, was moderately low as well (R 2 �0.22), suggesting that the high WSOC content at RIV was driven by non-ROS active soluble orga nics. Note that the term WSOC refers to a mixture of the water- soluble fr action of organic compounds, which is likely to be originated fr om different sources with different physical and chemical properties. Therefore, it is possible that some of the compounds within the WSOC, although in low concentra tions, could compri se redox-active materials that initiate the fo rmation ofROS. Overall, the role of organics in CP M-induced toxicity is rather unclear due to the dominance of biological material in OC I WSOC fr acti on. The organic content in CPM and its relationships with oxidative potential is a subject that may require further investigation. In summary, the higher associations between ROS and the water- soluble fr actions ofV, Pd, Cu and Rh suggest that, although in smaller mass fr actions compared with crustal materi als, these species could be primarily responsible for the oxidative potential of ambient CPM. In particular, vehicular abrasive emission-the common origin ofV, Pd, Cu and Rh in the coarse mode-could be a potential source driving CPM toxicity. The high internal correlations among these four species do not allow further investigation of their independent contributi on. The use of more sophisticated statistical analys is methods, such as Principal Component Analysis (PCA) or Positive Matrix Factorization (PMF), which could potentially provide more insightful results linking ROS and CPM species I sources was precluded by the limited number of data points generated in this study. 6.4.5. Corn parisons with other studies A number of studies have been conducted to evaluate the oxidative potential of aero sols of different size fr actions and distinctive origins . Table 6.2 shows a summary of the studies that 119 quantify ROS generation using the same method (in- vitro bioassay) as described in section 6.3. In an earlier study conducted in the downtown Los Angeles area, water-soluble V, Ni and Cd were the significant predictors-which have intrinsic redox properties- of ROS activity in quasi-u ltraf ine particles (Verma et al., 2009a). Hu et al. (2008) investigated the chemical and toxicological characteristics of PM in the coarse, accumulation and quasi -ultrafine mode in the Los Angeles-Long Beach Harbor area, and revealed that organic carbon, water- soluble V and Ni were linked to the generation ofROS. In 2007, fine particles were collected before and after a wildfire event to characterize the impact of wildfire on PM in Los Angeles, and high associ ations between ROS activity and a group of wate r-soluble metals including Cr, Ba, Pb, Fe, Ni, and V were reported (Verma et al., 2009b ). Based on the three studies conducted in the Los Angeles Basin, ROS generation was linked to emission of combustion origins, particularly heavy oil combustion (V and Ni). The results fr om our work, fo cusing solely on CPM, revealed that species of abrasive origins could be responsible for generation ofROS activity in the coarse mode. Table 6-2: Summary of studies that employed the same in-vitro bioassay as our study to examine oxidative potential in ambient particulate matter. Author and Published Year Location Time/Period Size Fractions Association: High Rl (>0.6 1 and p<0.05 Current study Los Angeles Basin, CA Diurnal samples in summer 2009 and winter Coarse 2010 Water-soluble V, Pd, Rh and Cu Zhang el at., 2008 Denver, CO Daily samples for one year in 2003 Iron source and soil dust source apportioned using water-s oluble elements, and water soluble carbon factor Hu el at., 2008 Verma el at., 2009 Verma el at., 2009 Shafer et at., 2010 Long Beach and Weekly composites on daily samples from Coarse, accumulation . �� Angeles, Mar-May 2007 and quasi-ullrafine OC, Water-soluble V, Nl Los Angeles, Integrated samples in Oci-Nov of2007, PM2.s Water-soluble Cr, Ba, Pb, Fe, Ni, V CA during and after wildfire Los Angeles, AM and PM periods in June and August Quasi-ultrafine (< 180 Water-soluble Ca, S, V, Cd, Ni CA 2008 nm) Lahore, Samples collected every sixth day for a full PMm wat er -soluble Mn, Co; Pakistan year from January 2007 to January 2008 PM 10 and PM2.s PM2.5: water -soluble Mn, Cd, Ce 120 A few studies evaluating PM-to xicity using other methods yielded results similar to our study. To investigate the toxicity of PM originating from contrasting traffic profiles in Europe, Gerlofs-Nijland et a!. (2007) sampled coarse and fine PM, and reported associations between toxicological responses and brake wear (Cu, Ba) in coarse particles. In 2007-2008, Godri et a!. (20 II) collected size- fr actionated particles at roadside and urban background school sites in London, and evaluated the oxidative potential using the depletion of ascorbate and glutat hione. The highest activity (per volume) was fo und for PM ranging from 1.9 to 10.2 f!m. Oxidative potential, assessed using glutathione depletion, also inc reased with inc reasing particle size, consist ent with the inc reased levels of Fe, Ba and Cu, which were likely emitted by brake wear (Godri et a!., 2011). Wang et a!. (2010) conducted a study on coarse mode aerosols collected in Los Angeles and Rivers ide, California, and suggested that the ma jority of hydrogen peroxide is mediated by soluble Fe, Cu and Zn. A recent study also found that water-soluble Cu collected in San Joaquin Valley, California is a potent source of generating hydrogen peroxide and hydroxyl radical in fine and coarse particles (Shen and Anastasio, 20 II). Furthermore, Fe but particularly Cu were linked to hydroxyl radical generating properties, measured by electron paramagnetic resonance, in fine particles collected in Paris (Baulig et a!., 2004). Wate r-soluble Cu was also shown to be a maj or contributor to the generation of hydroxyl radical, characterized by the consumption of ascorbic acid and the fo rmation of dihydroxyb enzoate, based on results fr om diesel exhaust particles and ambient particles (PM < 180 nm) (DiS tefano et a!., 2009). The abovementioned studies suggested that Cu was an important indicator of toxicity. In particular, mechanical wear of automobiles, a maj or source of soluble Cu in CPM, might represent a source that induces the generation of ROS in this size fr action. 121 6.5. Summary and Conclusions Overall, the ROS activity of coarse particle experienced distinct diurnal variati on. In summer, ROS activity peaked in midday I afternoon when wind speed was higher. Higher ROS activity was observed overnight in winter as CPM was re-suspended primarily by traffi c-induced turbulence. It should be noted that the CPM -related oxidative potential described here represents the toxicity of coarse particles in an outdoor enviro rnn ent. Particle int rusion from outdoor to indoor is lower in coarse than fine PM (Abt et a!., 2000). Polidori et a!. (2009) showed that the indoor-to-outdoor ratio was lowest for elemental coarse particles in comparison to those of quasi -ultrafine and accumulation part icles. The diurnal trend of oxidative potential presented here needs to be combined with personal exposure models in various microenvirornnents to estimate the actual toxicity imparted by coarse particles on public health. Based on our results, regulating coarse particles using PM10. 2 _5 or PM10 mass standards might not be eff ective in cont rolling the toxic sources of CPM. The elements that drive ROS activity in the coarse fr action are mainly anth ropogenic, and contribute to a relatively low mass fr action in that size range. More targeted regulation may therefore be needed to protect public health fr om toxic sources of coarse particles. 6. 6. Acknowledgements The study was supported by United States Enviro rnnental Protection Agency under the Science to Achieve Results program (EPA-G2006-STAR-Q l) to University of Southern California. The authors would also like to thank the staff at the Wisconsin State Lab of Hygiene (WSLH) for chemical and toxicological analysis of the PM samples. 122 Chapter 7 Conclusions 7.1. Characteristics of Coarse Particles 7. 1.1. Mass concentration The annual average of CPM mass concentrations ranged fr om 9.37 to 13.4 fig I m3 at the 10 sampling sites based on data fr om the USC comprehensive CPM study (details described in Chapters 2 and 3). Concentrations were 2 to 4 times higher in summer than winter. In particular, higher CPM levels were observed at the two Riverside sites in warmer months, when high wind speed facilitated regional transport and enhanced local re-suspension of coarse mode aerosols. In winter, the CPM mass concentrations were generally lower due to the lower wind speed, with the exception of the Long Beach site, where port-related activiti es contributed to both the emissions and re-suspension of coarse particles throughout the year. The overall correlation coefficients (R) between PM10. 2 _5 and PM10 vary between 0.71-0.91 among the 10 sampling sites, indicating that PM10 is a good surrogate of coarse particles. In general, higher correlations are observed in spring and fa ll (R ranging from 0.80-0.98). In summer, the associ ation between PM10. 2 _5 and PM10 was lower (R ranging from 0.40-0. 74) due to the high contribution of secondary aerosol form ation to fine PM as fa cilitated by photo-chemical reactions in warmer months in this basin (Pakbin et a!., 20 10). Additionally, the strong onshore wind fa cilitated the transport of marine aerosols inland in summer, resulting in significant contribution of sea salt aerosols to the CPM mass. In winter, the lower correlation (R ranging fr om 0.40-0.85) was driven by lower levels of CPM due to lower wind speed which decreased its contribution to PM10. In general, the overall associations between PM10. 2 _5 and PM10 were good, with seasonal variations due to the diverse sources of fine and coarse particles. 123 7.1.2. Chemical composition As presented in Chapter 2, Chapter 4 and Chapter 5, the chemical composition of coarse particles -which mostly comprise of mineral and road dust, and to a lesser extent inorganic ions and organic matter - is diff erent fr om those of fine PM. The ma jor origins of fine particles are fo ssil fuel combustion and photochemical reactions in urban atmospheres, leading to the high levels of carbonaceous compounds and secondary ions in the fm e PM mode as previously reported in the Los Angeles Basin. In contrast to anthropogenic origins of fine particles, CPM arises primarily fr om natural sources such as mineral dust, as well as fr esh and aged sea salt aerosols. Although the concentrations of crustal materials and organic matter varied among seasons, their relative contributions to CPM were consistent throughout the year, suggesting the source strength of soil dust and the associated biota is not significantly affected by the seasonal variation of met eorological conditi ons. On the other hand, both the concentrations and the fr actions of sea salt (both fr esh and aged) experienced seasonal variat ions. Their levels were higher in spring and summer, when the prevailing onshore wind was strong and regional transport fa cilitated the transport of marine aerosols fr om coast to inland. In addition to the seasonal variation, the diurnal variation of the chemical composition is significant. In summer, the concentrations of CPM mass, mineral and road dust, and WSOC were highest in the midday I afternoon, in parallel with the high wind speeds during those periods . In winter, the levels of mineral and road dust, sea salt and inorganic ions peaked overnight, when atmospheric dilution was low and traff ic-induced turbulences became a dominant re-suspension mechanism. This is particularly evident at the near-f reeway sites in Los Angeles and Rive rside. Overall, the different sources and fo rmation mechanism, as well as the strength of the primary re-suspension 124 mechanism of coarse and fine PM are the key drivers of their differential mass concentrations observed. 7.1.3. Health eff ects The generation of reactive oxygen species is used to characterize the oxidative potential of coarse particles. The CPM- induced ROS activity displayed distinct diurnal profiles in summer and winter, with higher levels in the warmer mont hs. Higher ROS activity was observed in the midday I afternoon sampling periods (11 a.m. to 3 p.m. and 3 p.m. to 7 p.m. respectively) during summertime, while the peak activity occurred in the overnight period (7 p.m. to 7 a.m.) in winter. Crustal materials, the most dominant component of CPM, displayed very low water solubili ty. On the other hand, metals generated fr om anthropogenic activities such as tire and brake wear experienced modest water solubility, which could increa se their bioavailability to human cells. Correlation analys is suggests that water-soluble fr action ofV, Pd, Cu and Rh are the key drivers of ROS activity. These constituents primarily originate fr om anth ropogenic activities of coarse particles. Despite their low contributions to the mass concentrati ons, they are important to the biological activity of CPM. 7.2. Discussions and Recommendations 7.2.1. Limitations of current investigation This investigation aimed to identify the linkage between source, chemical composition and toxicity of coarse particles, and to help the regulatory community to establish cost- effective PM regulations. Through the current investigation, the chemical mass composition, as well as their spatial and temporal characteristics is well characterized in the Los Angeles Basin, thereby providing valuable scie ntific information for environmental policy decision making. Nonet heless, 125 due to the higher settling velocity and heterogeneity of coarse particles, the results generated fr om this study might only be applicable in the Los Angeles Basin. Since the sources and fo rmation mechanism of coarse particles could vary considerably spatially, extrapolation of the findings developed based on this investigation to other areas needs to be proceeded with cautions . On the other hand, the toxicity of coarse particles is evaluated using the in-vitro ROS assay method, as described in a greater detail in Chapter 6. This bioassay is selected due to its sens itivity to elements, which are abundant in CPM. More importantly, this assay has been shown to be associated with airway and systematic biomarkers (using exhaled NO and plasma IL-6 respectively) in 60 human subjects (Delfino et a!., 20 10), whereas other commonly used molecular assays, such as the dithiothreitol (DTT) assay and the ascorbate assay, did not correlate with the observed health outcomes of that study. Theref ore, the use of the cellular ROS assay is very relevant in terms of health effects of airborne PM pollutants and thus is particularly appropriate for the scope of this study. Noneth eless, although this bioassay is designed to be responsive to ROS generated by a wide range of mechanisms and active species (peroxide, hydroxyl radical, super-oxide radial, organic -peroxides), and thereby providing a comprehensive evaluation of the oxidative potential induced by PM, the high levels of biological constituents present in coarse partic les, which might be partly responsible for CPM-induced adverse health effects , might not be characterized in this bioassay. Therefore, it is important to note that the toxicity data generated fr om this investigation represent only one of the many ways of evaluating PM-induced toxicity. 7.2.2. Im plications on epidemiological studies In epidemiological studies, the mass concentrations of coarse particles are estimated either 126 using direct (based on cascade sampling or impactors) or indirect (based on the difference between the measured PM10 and PM 2 .5 levels) measurements . The latter of which is affected by two measurement errors. The higher spatial heterogeneity of CPM, compared to PM10 and PM 2 .5, might also introduce uncertainty in the exposure characterization of coarse particles (Moore et a!., 20 10; Pakbin et a!., 20 I 0). In addition, the infiltration of ambient coarse particles in an indoor environment is moderate to low, resulting in lower levels of CPM indoor (Polidori et a!., 2009). Therefore, the actual exposure dose of coarse particles might be much lower than the ones estimated based on the measured ambient levels in sub jects that spent a consideration amount of time indoor. All of the abovementioned conditions need to be carefully considered in the design phase, as well as the exposure assessment of CPM in epidemiological studi es. 7.2.3. Recommendations for future research Exposure to bioaerosols, which are predominantly comprised of plant pollen, spores and microorgani sms, can result in allergic, toxic and infectious responses in a large fr action of the population (Levetin and Perry, 1995; Ross et a!., 2002; Targonski et a!., 1995). The results of current investigation suggest that the biological material is a ma jor component of the organic fr action of CPM. Therefore, further characterization of bioaerosols including the quantification of fungi, bacteria, plant pollen, and spore materials might provide important knowledge on the toxicity induced by the biological fr action of CPM. As discussed in Chapter 6, the results fr om this study, fo cusing solely on CPM, as well as a few other studies evaluating PM toxicity using other methods (Gerlofs-Nijland et a!., 2007; Godri et a!., 20 II), revealed that species of abrasive origins could be responsible for generation of ROS activity in the coarse mode. Therefore, road dust - comprising of crustal materials enriched by 127 both tailpipe and non-tailpipe vehicular emissions - could be an important source of coarse particle that generates redox-active constituents . In addition to the in-vitro evaluations of toxicity, in-vivo studies will be useful to evaluate the toxicity of road dust and I abrasive materials generated fr om dynamometer studi es. Further characterization on the physico-chemical and toxicological properties of road dust is also recommended. 7.2.4. Recommendations on coarse particle regulation In contrast to the anthropogenic origins of fine part icles, CPM predominantly arises fr om natural sources including crustal and biological materi als. Due to the distinct sources and fo rmation mechanism of fine and coarse PM, regulating coarse particles under the mass-based PM10 standard might not be appropriate. The past and current PM standards are effective in controlling particles of combustion origins, resulting in siguificant reductions of fm e PM in the last decade in this basin. In contrast, the levels of CPM remained similar in the last few years. In the PM NAAQS review in 2006, the U.S. EPA has proposed to use a new PM10. 2 .5 standard in replacement of the PM10 standard to regulate coarse particles. The proposed PM10. 2 .5 standard, which is desigued to be comparable to the current PM10 standard, displayed equivalency in the urban areas of the Los Angeles Basin. On the other hand, in sub-urban or rural areas, where both the wind speed and source strength of crustal materials are higher, the proposed standards are more stringent. Therefore, the proposed PM10. 2 .5 standard is likely to impose a greater impact in rural areas, where CPM concentrations are mostly driven by windblown dust. Although it is more appropriate to control coarse particles using a PM10. 2 .5 standard, PM10 is generally a good surrogate of coarse particles based on data collected in the Los Angeles Basin, where the levels of coarse and fm e PM are comparable. Using oxidative potential as a metric to evaluate toxicity, 128 constituents that drive ROS activity of CPM are mainly anth ropogenic in origin, and their contributi ons to CPM mass are low. Therefore, using a mass-based standard, regardless of its fo rm, may not be effective in reducing the CPM sources that drive ROS activity. Based on the results of this investigation, it is recommended to regulate coarse particles using target contro ls. Prioritizing the critical PM sources that are responsible for inducing adverse health effects may be the most effective approach to protect the public health fr om CPM exposure. 129 Bibliography Abbey, D.E., Lebowitz, M.D., Mills, P.K., Petersen, F.F., Beeson, W.L., Burchette, R.J., 1995. Long-Term Ambient Concentrati ons of Particulates and Oxidants and Development of Chronic Disease in a Cohort of Nonsmoking California Residen ts. Inhal Toxicol 7, 19-34. Abt, E., Suh, H.H., Catalano, P., Koutrakis, P., 2000. Relative contribution of outdoor and indoor particle sources to indoor concentra tio ns. Environ Sci Te chnol 34, 35 79-3587. Andrew s, E., Saxena, P., Musarra, S., Hildemann, L.M., Koutrakis, P., McMurry, P.H., Olmez, 1., White, W.H., 2000. Concentration and composition of atmospheric aerosols fr om the 1995 SEA VS experiment and a review of the closure between chemical and gravimetric measuremen ts. J Air Waste Manage 50, 648-664. Arhami, M., Ming uillon, M.C., Polidori, A., Schauer, J.J., Delf ino, R.J., Sioutas, C., 2010. Organic compound characterization and source apportionment of indoor and outdoor quasi -ultrafine particulate matter in retirement homes of the Los Angeles Basin. Indoor Air 20, 17-30. Arhami, M., Sillanpaa, M., Hu, S.H., Olson, M.R., Schauer, J.J., Sioutas, C., 2009. Size-Segregated Inorganic and Organic Components of PM in the Communities of the Los Angeles Harbor. Aerosol Sci Tech 43, 145-160. American Lung Association. State of the air: 2011, Available at: http: //www.stateofthe air.or g. Accessed November 30, 20 11. Barnes, P.J., 1990. Reactive Oxygen Species and Airway Inflammat ion. Free Radical Biology and Medicine 9, 235-243. Bauer, H., Schueller, E., Weinke, G., Berger, A., Hitzenberger, R., Marr, I.L., Puxbaum, H., 2008. Significant contributions of fungal spores to the organic carbon and to the aerosol mass balance of the urban atmospheric aerosol. Atmospheric Environment 42, 5542-554 9. 130 Baulig, A., Poirault, J.J., Ausset, P., Schins, R., Shi, T.M., Baralle, D., Dorlhene, P., Meyer, M., Lefevre, R., Baeza-Squiban, A., Marano, F., 2004. Physicochemical characteristics and biological activities of seasonal atmospheric particulate matter sampling in two locations of Paris . Environ Sci Te chnol 38, 5985 -5992. Birch, M.E., Cary, R.A., 1996. Elemental carbon-b ased method for monitoring occupational exposures to particulate diesel exhaust. Aerosol Sci Tech 25, 221-2 41. Birrnili, W., Allen, A. G., Bary, F., Harris on, R.M., 2006. Trace metal concentrations and water solubility in siz e-f ractionated atmospheric particles and influence of road traffic. Environ Sci Techn ol 40, 1144 -1153. Biswas, S., Verma, V., Schauer, J.J., Sioutas, C., 2009. Chemical speciation of PM emissions from heavy-duty diesel vehicles equipped with diesel particulate filter (DPF) and selective catalytic reduction (SCR) retrofi ts. Atmospheric Enviro rnn ent 43, 1917-1925. Brunekreef, B., Forsberg, B., 2005. Epidemiological evidence of effe cts of coarse airborne particles on health. European Respiratory Journ al 26, 309-318. Cahill, T.M., 2010. Size-Resolved Organic Speciation of Wintertime Aerosols in California's Central Valley. Enviro rnn ental Science & Te chnology 44, 23 15-2321. Campa, A.M. S.d.!., Moreno, T., Rosa, J.d.!., Alast uey, A., Querol, X., 20 11. Size distribution and chemical composition of metalliferous stack emissions in the San Roque petroleum refinery complex, southern Spain. Journal of Hazardous Materials, 713 -722. Castillejos, M., Bor ja-Aburto, V.H., Dockery, D. W., Gold, D.R., Loomis, D., 2000. Airborne coarse particles and mortality. Inhal Toxicol l2, 61- 72. Castranova, V., Ma, J.YC., Yang, H.M., Antonini, J.M., Butterworth, L., Barger, M.W., Roberts, J., Ma, J.K.H., 2001. Effect of exposure to diesel exhaust particles on the susceptibility of the lung to infection. Environ Health Persp 109, 609-612. Castro, L., Freeman, B., 2001. Reactive oxygen species in human health and disease. Nutrition 17, 161-165. 131 Chan, W.H., Vet, R.J., Lusis, M.A., Skelton, G.B., 1983. Airborne Particulate Size Distribution Measurements in Nickel Smelter Plu mes. Atmospheric Environment 17, 117 3-1181. Charron, A., Harris on, R.M., 2005. Fine (PM 2.5) and coarse (PM2.5-10) particulate matter on a heavily trafficked London highway: Sources and processes. Environ Sci Technol 39, 7768-7776. Chen, L., Yang, W., Jennison, B.L., Omaye, S. T., 2000. Air particulate pollution and hospital admissions for chronic obstruc tive pulmonary disease in Reno, Nevada. Inhal Toxicol 12, 28 1-298. Chen, Y, Yang, Q. Y, Krewski, D., Burnett, R. T., Shi, YL., McGrail, K.M., 2005. The effect of coarse ambient particulate matter on first, second, and overall hospital admissions for respiratory disease among the elderly. Inhal Toxicol 17, 649-6 55. Cheng, Y, Brook, J.R., Li, S.M., Leithead, A., 20 11. Seasonal variation in the biogenic secondary organic aerosol tracer cis-pinonic acid: Enhancement due to emissions fr om regional and local biomass burning. Atmospheric Envirornn ent 45, 7105-7112. Cheung, K., Daher, N., Kam, W., Shafer, M.M., Ning, Z., Schauer, J.J., Sioutas, C., 2011a. Spatial and Temp oral Variation of Chemical Composition and Mass Closure of Ambient Coarse Particulate Matter (PM10-2.5) in the Los Angeles Area. Atmospheric Enviro rnn ent, doi: 10. 10 16/j.a tmosenv .2011. 1002. 1066. Cheung, K., Daher, N., Kam, W., Shafer, M.M., Ning, Z., Schauer, J.J., Sioutas, C., 20 11b. Spatial and temporal variation of chemical composition and mass closure of ambient coarse particulate matter (PM10-2.5) in the Los Angeles area. Atmospheric Environment 45, 2651-2662. Cheung, K., Shafer, M.M., Schauer, J.J., Sioutas, C., 20 11c. Diurnal trends in coarse particulate matter composition in the Los Angeles Basin. Journal ofEnviro rnnental Monitoring DOl: 10.10 39/clem102 96f. Cheung, K., Shafer, M.M., Schauer, J.J., Sioutas, C., 2012. Historical trends in the mass and chemical species concentrations of coarse particulate matter in the Los Angeles Basin and relation to sources and air quality regulation s. J Air Waste Manage 62, 541-556. 132 Cheung, K.L., Ntzia christos, L., Tzarnkio zis, T., Schauer, J.J., Samaras, Z., Moore, K.F., Sioutas, C., 2010. Emissions of Particulate Trace Elements, Metals and Organic Species fr om Gasoline, Diesel, and Biodiesel Passenger Vehicles and Their Relation to Oxidative Potenti al. Aerosol Sci Tech 44, 500-5 13. Cho, A.K., Sioutas, C., Miguel, A.H., Kumagai, Y, Schmitz, D.A., Singh, M., Eiguren-Fernandez, A., Froines, J.R., 2005. Redox activity of airborne particulate matter at different sites in the Los Angeles Basin. Environ Res 99, 40-47. Chow, J.C., Watson, J.G., Fujita, E.M., Lu, Z.Q., Lawson, D.R., Ashbaugh, L.L., 1994. Temporal and Spatial Variations ofPrn(2.5) and Prn(10) Aerosol in the Southern California Air-Quality Study. Atmospheric Environment 28, 2061-2080. Chow, J.C., Watson, J.G., Lowenthal, D.H., Chen, L.W.A., Zielinska, B., Mazzoleni, L.R., Magliano, K.L., 2007. Evaluation of organic markers for chemical mass balance source apportionment at the Fresno Supersite. Atrnos Chern Phys 7, 1741-1754. Chow, J.C., Watson, J.G., Lowenthal, D.H., Solomon, PA., Magliano, K.L., Zirnan, S.D., Richards, L.W., 1992. Prn10 Source Apportionment in California San-Joaquin Valley. Atmospheric En vironment Part a-General Topics 26, 3335-3354. Costa, D.L., Dreher, K.L., 1997. Bioavailable transition metals in particulate matter mediate cardiopulmonary in jury in healthy and compromi sed animal models. Environ Health Persp 105, 1053-1060. Daher, N., Ruprecht, A., Invernizzi, G., De Marco, C., Miller-Schulze, J., Heo, J.B., Shaf er, M.M., Schauer, J.J., Sioutas, C., 20 11. Chemical Characterization and Source Apportionment of Fine and Coarse Particulate Matter Inside the Ref ectory of Santa Maria Delle Grazie Church, Horne of Leonardo Da Vinci's "Last Supper". Environ Sci Technol 45, 10344-10353. Delfino, R.J., Sioutas, C., Malik, S., 2005. Potential role oful trafine particles in associations between airborne particle mass and cardiovascular health. Environ Health Persp 11 3, 934-946. 133 Delfino, R.J., Staimer, N., Tjoa, T., Arhami, M., Polidori, A., Gillen, D.L., George, S.C., Shafer, M.M., Schauer, J.J., Sioutas, C., 2010. Ass ociations of Primary and Secondary Organic Ae rosols With Airway and Systemic Inflammation in an Elderly Panel Cohort. Epidemiology 21, 892-902. Dentener, F.J., Carmichael, G.R., Zhang, Y, Lelieveld, J., Crutz en, P.J., 1996. Role of mineral aerosol as a reactive surface in the global trop osphere. J Geophys Res-Atmos 101, 22869-22889. DiStef ano, E., Eiguren-F ernandez, A., Delfino, R.J., Sioutas, C., Froines , J.R., Cho, A.K., 2009. Determination of metal-based hydroxyl radical generating capacity of ambient and diesel exhaust particles. Inhal Toxicol 21, 731-738. Docherty, K.S., Stone, E.A., Ulbrich, I.M., DeCarlo, P.F., Snyder, D.C., Schauer, J.J., Peltier, R.E., Weber, R.J., Murphy, S.M., Seinf eld, J.H., Grover, B.D., Eatough, D.J., Jimenez, J.L., 2008. Apportiornn ent of Primary and Secondary Organic Aerosols in Southern California during the 2005 Study of Organic Aerosols in Ri verside (SOAR-1). Environ Sci Technol 42, 7655- 7662. Dockery, D. W., Pope, C. A., 1994. Acute Respiratory Effects of Particulate Air-Pollution. Annual Review of Public Health 15, 107-132. Dockery, D.W., Pope, C.A., Xu, X.P., Spengler, J.D., Ware, J.H., Fay, M.E., Ferri s, B. G., Speizer, F.E., 1993. An Asso ciation between Air-Pollution and Mortality in 6 Unite d-States Cities. New England Journal of Medicine 329, 1753-1759. Dutton, S.J., Ra jagopalan, B., Veda!, S., Hannigan, M.P., 2010. Temporal patterns in daily measurements of inorganic and organic speciated PM2. 5 in Denver. Atmospheric Environment 44, 987-998. Eiguren-Fernandez, A., Miguel, A.H., Lu, R., Purvis, K., Grant, B., Mayo, P., Di Stef ano, E., Cho, A.K., Froines, J., 2008. Atmospheric formation of 9, 10-phenanthraquinone in the Los Angeles air basin. Atmospheric Enviro rnnent 42, 2312-2319. Eldering, A., Solomon, P.A., Salmon, L.G., Fall, T., Cass, G.R., 1991. Hydrochloric-A cid - a Regional Perspective on Concentrations and Formation in the Atmosphere of Southern California. Atmospheric Enviro rnnent Part a-General Topics 25, 2091-2102. 134 Falkovich, A.H., Graber, E.R., Schkolnik, G., Rudich, Y, Maenhaut, W., Artaxo, P, 2005. Low molecular weight organic acids in aerosol particles from Rondonia, Brazil, during the biomass-burning, transition and wet periods . Atmos Chern Phys 5, 781-797. Falkovich, A.H., Schkolnik, G., Ganor, E., Rudich, Y, 2004. Adsorption of organic compounds pertinent to urban enviro rnnents onto mineral dust particles. J Geophys Res-Atmos 109, -. Fine, PM., Chakrabarti, B., Krudysz, M., Schauer, J.J., Sioutas, C., 2004a. Diurnal variations of individual organic compound constituents of ultrafine and accumulation mode particulate matter in the Los Angeles basin. Environ Sci Technol 38, 1296-1304. Fine, PM., Shen, S., Sioutas, C., 2004b. Inf erring the sources off me and ultrafine particulate matter at downwind receptor sites in the Los Angeles basin using multiple continuous measuremen ts. Aerosol Sci Tech 38, 182-195. Finlayson-Pitts, B.J., 2003. The trop ospheric chemistry of sea salt: A molecular-level view of the chemistry ofNaCI and NaBr. Chemical Reviews 10 3, 4801- 4822. Folkrnann, J.K., Risom, L., Hansen, C.S., Loft, S., Moller, P, 2007. Oxidatively damaged DNA and inflammation in the liver of dyslipidernic ApoE( -/-) mice exposed to diesel exhaust particles. Toxicology 237, 134-144. Garg, B.D., Cadle, S.H., Mulawa, PA., Groblicki, PJ., Laroo, C., Parr, G.A., 2000. Brake wear particulate matter emissions. Environ Sci Techn ol 34, 4463-4469. Gauderrnan, W.J., McConnell, R., Gilliland, F., London, S., Thomas, D., Avo!, E., Vora, H., Berhane, K., Rappaport, E.B., Lurmann, F., Margolis, H. G., Peters, J., 2000. Associ ation between air pollution and lung function growth in southern Calif ornia children. American Journal of Respiratory and Critical Care Medicine 162, 1383-1390. Gauderrnan, W.J., Vora, H., McConnell, R., Berhane, K., Gilliland, F., Thomas, D., Lurrnann, F., Avoli, E., Kunzli, N., Jerrett, M., Peters, J., 2007. Effect of exposure to traffic on lung development from 10 to 18 years of age: a cohort study. Lancet 369, 571-577. Geller, M.D., Fine, PM., Sioutas, C., 2004. The relationship between real-time and time-integrated coarse (2.5-10 mu m), interrnodal (1-2.5 mu m), and fine (< 2.5 mu m) particulate matter in the Los Angeles Basin. J Air Waste Manage 54, 1029-1039. 135 Geller, M.D., Kim, S., Misra, C., Sioutas, C., Olson, B.A., Marple, VA., 2002. A metho dology for measuring size-dependent chemical composition ofultrafine particles. Aerosol Sci Tech 36, 748-762. Gerlofs-Nijland, M.E., Dorrnans, J.A.M.A., Bloemen, H.J.T., Leseman, D.L.A.C., Boere, A.J.F., Kelly, F.J., Mudway, I.S., Jimenez, A.A., Donaldson, K., Guastadisegni, C., Janssen, N.A.H., Brunekreef, B., Sandstrom, T., Cassee, F.R., 2007. Toxicity of coarse and fme particulate matter from sites with contrasting traffic profiles. Inhal Toxicol 19, 1055-1069. Gertler, A.W., Gillies, J.A., Pierson, W.R., 2000. An assessment of the mobile source contribution to PM10 and PM2.5 in the United States . Water Air and Soil Pollution 123, 203-214. Gietl, J.K., Lawrence, R., Thorpe, A.J., Harris on, R.M., 2010. Identification of brake wear particles and derivation of a quantitative tracer for brake dust at a maj or road. Atmospheric Environment 44, 141- 146. Godri, K.J., Harrison, R.M., Evans, T., Baker, T., Dunster, C., Mudway, I.S., Kelly, F.J., 20 11. Increased Oxidative Burden Associ ated with Traffic Component of Ambient Particulate Matter at Roadside and Urban Background Schools Sites in London. Plos One 6. Goldsmith, C.A.W., Imrich, A., Danaee, H., Ning, Y, Kobzik, L., 1998. Analys is of air pollution particulate-mediated oxidant stress in alveolar macroph ages. Journal of Toxicology and Environmental Health-Part A 54, 529-545. Goodman, A.L., Underwood, G.M., Grassian, V.H., 2000. A laboratory study of the heterogeneous reaction of nitric acid on calcium carbonate part icles. J Geophys Res-Atmos 105, 29053- 29064. Harrison, R.M., Williams, C.R., 1983. Physicochemical Characterization of Atmospheric Trace-Metal Emissions from a Primary Zinc-Lead Smelter. Sci Total Environ 31, 129-140. Harrison, R.M., Yin, J.X., Mark, D., Stedman, J., Appleby, R.S., Booker, J., Moorcroft, S., 200 1. Studies of the coarse particle (2.5-10 mu m) component in UK urban atmo spher es. Atmospheric En vironment 35, 3667-3679. 136 Hays, M.D., Gullett, B., King, C., Robinson, J., Preston, W., Touati, A., 20 11. Characterization of Carbonaceous Ae rosols Emitted from Outdoor Wood Boilers. Energy & Fuels 25, 5632-5638. Herckes, P, Engling, G., Kreidenweis, S.M., Collett, J.L., 2006. Particle size distributions of organic aero sol constituents during the 2002 Yosemite Aerosol Characterization Study. Environ Sci Technol 40, 4554 -4562. Hildernann, L.M., Mazurek, M.A., Cass, G.R., Sirnoneit, B.R. T., 1994. Seasonal Trends in Los-Angeles Ambient Organic Aerosol Observed by High-Resolution Gas-Chromatography. Aerosol Sci Tech 20, 303-317. Hinds, W.C., 1999. Aerosol Technology: Propert ies, Behavior, and Measurement of Airborne Particles, 2nd ed. Wiley-Inte rscience, New York. Hjortenk rans, D.S .T., Bergback, B. G., Haggerud, A.V., 2007. Metal emissions from brake linings and tires: Case studies of Stockholm, Sweden 1995/1998 and 2005. Environ Sci Technol 41, 5224-5230. Ho, K.F., Lee, S.C., Cao, J.J., Chow, J.C., Watson, J.G., Chan, C.K., 2005. Seasonal variations and mass closure analysis of particulate matter in Hong Kong. Sci Total Environ 355, 276-287. Hoek, G., Brunekreef, B., Goldbohrn, S., Fisc her, P, van den Brandt, PA., 2002. Asso ciation between mortality and indicators of traffic-related air pollution in the Nether lands : a cohort study. Lancet 360, 1203-1209. Hu, S., Polidori, A., Arharni, M., Shaf er, M.M., Schauer, J.J., Cho, A., Sioutas, C., 2008. Redox activity and chemical speciation of size fractioned PM in the communities of the Los Angeles-Long Beach harbor. Atrnos Chern Phys 8, 6439-6451. Huang, X.F., Yu, J.Z., 2008. Size distributions of elemental carbon in the atmosphere of a coastal urban area in South China: characte ristics, evolution processes, and implications for the mixing state. Atrnos Chern Phys 8, 5843-5853. Hueglin, C., Gehrig, R., Baltensperger, U., Gysel, M., Monn, C., Vonrnont, H., 2005. Chemical characterisation ofPM2. 5, PMlO and coarse particles at urban, near-city and rural sites in Switzerland. Atmospheric Environment 39, 637-651. 137 Hugh es, L.S., Allen, J.O., Kleeman, M.J., Johnson, R.J., Cass, G.R., Gross, D.S., Gard, E.E., Galli, M.E., Morrical, B.D., Fergenson, D.P., Dienes, T., Noble, C.A, Silva, P.J., Prather, K.A., 1999. Size and composition distribution of atmospheric particles in southern California. Environ Sci Technol 33, 3506-3515. Hwang, I., Hopke, P.K., Pinto, J.P., 2008. Source apportionment and spatial distributions of coarse particles during the regional air pollution study. Environ Sci Technol 42, 352 4-3530. Isakson, J., Pe rsson, T.A., Lindgren, E.S., 2001. Identification and assessment of ship emissions and their effe cts in the harbour of G( o )over-circleteborg, Sweden. Atmospheric Environment 35, 3659-3666. Jia, YL., Clements, A.L., Fraser, M.P., 2010. Saccharide composition in atmospheric particulate matter in the southwest US and estimates of source contributi ons. J Aerosol Sci 41, 62-73. Kanitsar, K., Koellensp erger, G., Harm, S., Limbeck, A., Puxbaum, H., Stingeder, G., 2003. Determination ofPt, Pd and Rh by inductively coupled plasma sector field mass spectrometry (ICP-SFMS) in size-classified urban aerosol samples. Journal of Analytical Atomic Spectrometry 18, 239-246. Karageorgo s, E.T., Rapsomanikis, S., 2007. Chemical characterization of the inorganic fraction of aerosols and mechanisms of the neut ralization of atmospheric acidity in Athens, Greece. Atmos Chern Phys 7, 301 5-3033. Kawamura, K., Kaplan, I.R., 1987. Motor Exhaust Emissions as a Primary Source for Dicarboxylic-Acids in Los-Angeles Ambient Air. Environ Sci Technol 21, 105-110. Khalili, N.R., Scheff, P.A., Holsen, T.M., 1995. Pah Source Fin gerprints for Coke Ovens, Diesel and Gasoline-Engine s, Highway Tunnels, and Wood Combustion Emissions. Atmospheric Environment 29, 533-542. Kim, B.M., Teff era, S., Zeldin, M.D., 2000a. Characterization ofPM2.5 and PMIO in the South Coast Air Basin of southern Californi a: Part I - Spatial variat ions. J Air Waste Manage 50, 2034-2044. 138 Kim, B.M., Teff era, S., Zeldin, M.D., 2000b. Characterization ofPM.2.5 and PMIO in the Sou th Coast Air Basin of sou thern Californi a: Part 2 - Temporal variat ions. J Air Waste Manage 50, 2045-2059. Kleeman, M.J., Hug hes, L.S., Allen, J.O., Cass, G.R., 1999. Source contributions to the size and composition distribution of atmospheric partic les: Sou thern California in September 1996. Environ Sci Technol33, 4331-4341. Kleeman, M.J., Robert, M.A., Riddle, S.G., Fine, PM., Hays, M.D., Schauer, J.J., Hannigan, M.P, 2008. Size distribution of trace organic species emitted fr om biomass combustion and meat charbroiling. At mospheric Environment 42, 3059-3075. Kleinman, M. T., Sioutas, C., Froines, J.R., Fanning, E., Hamade, A., Mendez, L., Meacher, D., Oldham, M., 2007. Inhalation of concentrated ambient particulate matter near a heavily trafficked road stimulates antigen-induced airway responses in mice. Inhal Toxicol 19, 117 -126. Koulouri, E., Saarikoski, S., Theodosi, C., Markaki, Z., Gerasopoulos, E., Kouvarakis, G., Makela, T., Hillamo, R., Mihalopoulos, N., 2008. Chemical composition and sources of fine and coarse aerosol particles in the Eastern Mediterran ean. At mospheric Environment 42, 6542-6550. Krudysz, M.A., Froines, J.R., Fine, PM., Sioutas, C., 2008. Intra-community spatial variation of size- fractionated PM mass, OC, EC, and trace elements in the Long Beach, CA area. At mospheric Environment 42, 5374-5389. Kuhnala, M., Vehkamaki, H., Petaja, T., Dal Maso, M., Lauri, A., Kerminen, V.M., Birrnili, W., McMurry, PH., 2004. Formation and growth rates of ultrafine atmospheric partic les: a review of observatio ns. J Aerosol Sci 35, 143-176. Landreman, A.P, Shafer, M.M., Hemming, J.C., Hannigan, M.P, Schauer, J.J., 2008. A macrophage-based method for the assessment of the reactive oxygen species (ROS) activity of atmospheric particulate matter (PM) and application to routine ( daily-24 h) aerosol monitoring st udies. Aero sol Sci Tech 42, 946-957. 139 Lenschow, P., Abraham, H.J., Kutzner, K., Lutz, M., Preuss, J.D., Reichenbacher, W., 2001. Some ideas about the sources ofP MlO. At mospheric Environment 35, S23-S33. Levetin, E., Perry, C., 1995. Evaluation of Airborne Fungi with a Dichloran-Glycerol Agar. Journal of Allergy and Clinical Immu nology 95, 168-168. Li, N., Sioutas, C., Cho, A., Schmitz, D., Misra, C., Sernpf, J., Wang, M. Y, Oberley, T., Froines, J., Nel, A., 2003. Ultrafine particulate pollutants induce oxidative stress and mitochondrial damage. Environ Health Persp 111 , 455-460. Lin, C. C., Chen, S.J., Huang, K.L., Hwang, W.I., Chang-Chien, G.P., Lin, W.Y, 2005. Characteristics of metals in nano/ultrafrne/f rne/ coarse particles collected beside a heavily trafficked road. Environ Sci Technol39, 811 3-8 122. Lips ett, M.J., Tsai, F. C., Roger, L., Woo, M., Ostro, B.D., 2006. Coarse particles and heart rate variability among older adults with coronary artery disease in the Coachella Valley, California. Environ Health Persp 114, 1215-1220. Lough, G. C., Schauer, J.J., Park, J.S., Shafer, M.M., Derninter, J.T., Weinstein, J.P., 2005. Emissions of metals associated with motor vehicle roadways . Environ Sci Technol39, 826-836. Lun, X.X., Zhang, X.S., Mu, YJ., Nang, A., Jiang, G.B., 2003. Size fractionated speciation of sulfate and nitrate in airborne particulates in Bei jing, China. At mospheric Environment 37, 2581-2588. Mader, B. T., Schauer, J.J., Sei nf eld, J.H., Flagan, R.C., Yu , J.Z., Yang, H., Lim, H.J., Turpin, B.J., Deminter, J.T., Heidemann, G., Bae, M.S., Quinn, P., Bates, T., Eatough, D.J., Huebert, B.J., Bertram, T., Howell, S., 2003. Sampling methods used for the collection of parti cle-phase organic and elemental carbon during ACE-Asia. At mospheric Environment 37, 1435- 1449. Marcazzan, G.M., Va ccaro, S., Valli, G., Vecchi, R., 200 1. Characterisation ofP MlO and PM2.5 particulate matter in the ambient air of Milan (Ital y). At mospheric Environment 35, 4639-46 50. 140 McConnell, R., Berhane, K., Gilliland, F., London, S.J., Vora, H., Avo!, E., Gauderrnan, W.J., Mar golis, H. G., Lurrnann, F., Thomas, D.C., Peters, J.M., 1999. Air pollution and bronchitic symptoms in Southern California children with asthma. Environ Health Persp 107,757-760. Miguel, A.H., Eiguren-Fernandez, A., Jaques, P.A., Froines, J.R., Grant, B.L., Mayo, P.R., Sioutas, C., 2004. Seasonal variation of the particle size distribution of polycyclic aromatic hydrocarbons and of major aerosol species in Claremont, California. At mospheric Environment 38, 3241-3 251. Minguillon, M.C., Arhami, M., Schauer, J.J., Sioutas, C., 2008. Seasonal and spatial variations of sources off me and quasi-ultrafine particulate matter in neighb orhoods near the Los Angeles-Long Beach harbor. At mospheric Environment 42, 7317-7328. Misra, C., Geller, M.D., Shah, P., Sioutas, C., Solomon, P.A., 200 1. Development and evaluation of a continuous coarse (PM10-PM 2.5) particle monitor. J Air Waste Manage 51, 1309-13 17. Moore, K.F., Verma, V., Ming uillon, M.C., Sioutas, C., 2010. Inter- and Intra-Community Var iability in Continuous Coarse Particulate Matter (PMl0 -2.5) Concentrations in the Los Angeles Area. Aero sol Sci Tech 44, 526-540. Neiburger, M., Wurtele, M.G., 1949. On the Nature and Size of Particles in Haze, Fog, and Stratus of the Los-Angeles Region. Chemical Reviews 44, 321-335. Nel, A., 2005. Air pollution-related illnes s: Effects of part icles. Science 308, 804-806. Nel, A.E., Diaz-Sa nchez, D., Li, N., 200 1. The role of particulate pollutants in puhnonary inflammation and asthma : evidence for the involvement of organic chemicals and oxidative stres s. Current opinion in pulmonary medicine 7, 20-26. Newhook, R., Hirtle, H., Byrne, K., Meek, M.E., 2003. Releases from copper smelters and refineries and zinc plants in Canada: human health exposure and risk characterization. Sci Total Environ 301, 23-41. Nicholson, K. W., Branson, J.R., Giess, P., Cannell, R.J., 1989. The Effects of Vehicle Activity on Particle Resuspension. J Aero sol Sci 20, 1425-1428. 141 Ning, Z., Geller, M.D., Moore, K.F., Sheesley, R., Schauer, J.J., Sioutas, C., 2007. Daily variation in chemical characteristics of urban ultrafine aero sols and inference of their sourc es. Environ Sci Technol41, 6000-6006. Noble, C.A., Prather, K.A., 1996. Real-time measurement of correlated size and composition profiles of individual atmospheric aerosol particles. Environ Sci Technol30, 2667-2680. Oberdorster, G., Sharp, Z., At udorei, V., Elder, A., Gelein, R., Kreyling, W., Cox, C., 2004. Translocation of inhaled ultrafrne particles to the brain. Inhal Toxicol 16, 437-445. Pakbin, P., Hudda, N., Cheung, K.L., Moore, K.F., Sioutas, C., 2010. Spatial and Temporal Var iability of Coarse (PMI0-2. 5) Particulate Matter Concentrations in the Los Angeles Area. Aero sol Sci Tech 44, 514 -525. Pakbin, P., Ning, Z., Shafer, M.M., Schauer, J.J., Sioutas, C., 2011. Seasonal and Spatial Coarse Particle Elemental Concentrations in the Los Angeles Area. Aero sol Sci Tech 45, 949-U l5 6. Paode, R.D., Shahin, U.M., Sivadechathep, J., Holsen, T.M., Franek, W.J., 1999. Source apportio rnnent of dry deposited and airborne coarse particles collected in the Chicago area. Aero sol Sci Tech 31, 473-486. Pelicano, H., Carney, D., Huang, P., 2004. ROS stre ss in cancer cells and therapeutic implicat ions. Drug Resistance Updates 7, 97- 11 0. Peters, A., Ve ronesi, B., Calder6n-Garciduefias, L., Gehr, P., Chen, L.C., Geiser, M., Reed, W., Rothen-R utishauser, B., Schiirch, S., Schulz, H., 2006. Translocation and potential neu rological effe cts of fine and ultrafine particles a critical update. Particle and Fibre Toxicology 3:13. Polidori, A., Arharni, M., Sioutas, C., Delfino, R.J., Allen, R., 2007. Indoor/outdoor relations hips, trends, and carbonaceous content of frn e particulate matter in retirement homes of the Los Angeles basin. J Air Waste Manage 57, 366-379. Polidori, A., Cheung, K.L., Arharni, M., Delfino, R.J., Schauer, J.J., Sioutas, C., 2009. Relationships between siz e-f ractionated indoor and outdoor trace elements at four retirement communities in southern California. Atrnos Chern Phys 9, 4521-4536. 142 Pope, C. A., Dockery, D.W., 2006. Health effects of fine particulate air pollution: Lines that connect. J Air Waste Manage 56, 709-742. Prophete, C., Maciejczyk, P., Salnikow, K., Gould, T., Larson, T., Koenig, J., Jaques, P., Sioutas, C., Lippmann, M., Cohen, M., 2006. Effects of select PM -asso ciated metals on alveolar macrophage phosphorylated ERK.l and-2 and iNOS expr ession during ongoing alteration in iron homeostas is. Journal of Toxicology and Environmental Health-Part a-Current Issues 69, 935-95 1. Putaud, J.P., Van Dingenen, R., Dell'Acqua, A., Raes, F., Matta, E., Decesari, S., Facchini, M.C., Fuzzi, S., 2004. Size-seg regated aerosol mass closure and chemical composition in Monte Cirnone (I) during MINA TROC. At mospheric Chemistry and Physics 4, 889-902. Qiu, H., Yu, LT., Tian, L., Wang, X., Tse, L.A., Tarn, W., Wong, T.W., 2012. Eff ects of Coarse Particulate Matter on Emergency Hospital Admi ssions for Respiratory Diseases : A Time Series Analys is in Hong Kong. Environ Health Perspect. Querol, X., Alastuey, A., Ruiz, C.R., Artinano, B., Hansson, H. C., Harris on, R.M., Buringh, E., ten Brink, H.M., Lutz, M., Brockmann, P., Straehl, P., Schneider, J., 2004. Speciation and origin ofPMIO and PM2.5 in selected European citi es. At mospheric Environment 38, 65 47- 6555. Rauch, S., Hemond, H.F., Barbante, C., Owari, M., Morrison, G.M., Peucker-Ehrenbrink, B., Wass, U., 2005. Importance of au tomobile exhaust catalyst emissions for the depostion of platinum, palladium, and rhodium in the Northern Hemisphere. Environ Sci Technol39, 8156-8162. Rauch, S., Lu, M., Morr ison, G.M., 200 1. Heterogeneity of platinum group metals in airborne particles. Environ Sci Techn ol 35, 595-599. Roberts, J.R., Young, S.H., Castranova, V., Antonini, J.M., 2007. Soluble metals in residual oil fly ash alter innate and adaptive pulmonary immune responses to bacterial infection in rats . Toxicology and Applied Pharmacology 221, 306-319. 143 Rogge, W.F. , Hildemann, L.M., Mazurek, M.A., Cass, G.R., Sirnoneit, B.R.T., 1993a. Sources of Fine Organic Aero sol .3. Road Dust, Tire Debris, and Organometallic Brake Lining Dust Roads as Sources and Sinks. Environ Sci Techn ol 27, 1892-1904. Rogge, W.F. , Hildemann, L.M., Mazurek, M.A., Cass, G.R., Sirnoneit, B.R. T., 1993b. Sources of Fine Organic Aero sol .4. Particulate Abrasion Products fr om Leaf Surfaces of Urban Plants . Environmental Science & Technology 27, 2700-27 11. Rogge, W.F. , Hildemann, L.M., Mazurek, M.A., Cass, G.R., Sirnoneit, B.R.T., 1998. Sources of fine organic aero sol. 9. Pine, oak and synthetic log combustion in residential firep laces. Environ Sci Techn ol 32, 13-22. Rogge, W.F. , Hildemann, L.M., Mazurek, M.A., Cass, G.R., Simone!!, B.R.T., 1991. Sources of Fine Organic Aero sol .I. Charbroilers and Meat Cooking Operat ions. Environ Sci Technol25, 1112-1125. Rogge, W.F. , Mazurek, M.A., Hildemann, L.M., Cass, G.R., Sirnoneit, B.R.T., 1993c. Quantification of Urban Organic Aerosols at a Molecular-Level - Identification, Abun dance and Seasonal-Variation. At mospheric Enviro rnnent Part a-General Topics 27, 1309-1330. Root, R.A., 2000. Lead loading of urban streets by motor vehicle wheel weig hts . Environ Health Persp 108, 937-940. Ross, M.A., Persky, V. W., Chung, J., Curtis, L., Ramakrishnan, V., Wadden, R.A., Hryhorczuk, D.O., 2002. Effect of ozone and aeroallergens on the respiratory health of asthmat ics. Archives of Environmental Health 57, 568-578. Ruzer, L.S., Harley, N.H., 2004. Aerosols Hand book: Measurement, Dosimetry, and Health Effects , 1st ed. CRC Pres s. Samy, S., Robinson, J., Hays, M.D., 2011. An advanced LC-MS (Q-TOF) technique for the detection of amino acids in atmospheric aero sols. Analytical and Bio analytical Chemistry 401, 310 3-3 11 3. Sanders, P.G., Xu, N., Dalka, T.M., Maricq, M.M., 2003. Airborne brake wear debris : Size distributi ons, composition, and a compa rison of dynamometer and vehicle tests . Environ Sci Techn ol 37, 4060-4069. 144 Sardar, S.B., Fine, PM., Sioutas, C., 2005. Seasonal and spatial variability of the size-resolved chemical composition of particulate matter (PMlO) in the Los Angeles Basin. J Geophys Res-Atmos ll O, -. Schauer, J.J., Kleeman, M.J., Cass, G.R., Sirnoneit, B.R.T., 2002. Measurement of emissions from air pollution sourc es. 4. C-1-C-27 organic compounds from cooking with seed oils. Environ Sci Technol36, 567-575. Schauer, J.J., Lough, G. C., Shafer, M.M., Christensen, W.F., Arndt, M.F., DeMinter, J., Park, J.-S., 2006. Characterization of Metals Emitted from Motor Ve hicles . Health Effects Institute 133, 1-88. Schnelle-Kreis, E., Sklorz, M., Peters, A., Cyrys, J., Zimmermann, R., 2005. Analys is of partic le-assoc iated semi-volatile aromatic and aliphatic hydrocarbons in urban particulate matter on a daily basis. At mospheric Enviro rnnent 39, 7702-7714. Schwartz, J., Dockery, D. W., Neas, L.M., Wypij , D., Ware, J.H., Spengler, J.D., Koutrakis, P, Speizer, F.E., Ferris, B. G., 1994. Acute Effe cts of Summer Air-Pollution on Respiratory Symptom Reporting in Children. American J oumal of Respiratory and Critical Care Medicine 150, 1234-1242. Schwarze, PE., Ovrevik, J., Hetland, R.B., Becher, R., Cassee, F.R., Lag, M., Lovik, M., Dybing, E., Refsnes, M., 2007. Importance of size and composition of particles for effects on cells in vit ro. Inhal Toxicol 19, 17-22. Sei nf eld, J.H., Pandis, S.N., 2006. At mospheric Chemistry and Physics -F rom Air Pollution to Climate Change, Second edition ed. John Wiley & Sons, Inc, New Yo rk. Shafer, M.M., Perkins, D.A., Antkiewicz, D.S., Stone, E.A., Quraishi, T.A., Schauer, J.J., 2010. Reactive oxygen species activity and chemical speciation of size -f ractionated atmospheric particulate matter from Lahore, Pakistan: an important role for transition meta ls. Journal of Environmental Monitoring 12, 704-7 15. Sheesley, R.J., Schauer, J.J., Bean, E., Kenski, D., 2004. Trends in secondary organic aerosol at a remote site in Michigan's upper peninsula. Environ Sci Techn ol 38 , 6491-6500. 145 Shen, H., Anast asio, C., 2011. A Comparison of Hydroxyl Radical and Hydrogen Peroxide Generation in Ambient Particle Extracts and Laboratory Metal Solutions . At mospheric Environment doi: 10. 1016 /j.atmos env.20 11. 10.006. Shi, Z., Zhang, D., Hayashi, M., Ogata, H., Ji, H., Fujiie, W., 2008. Influences of sulfate and nitrate on the hyg roscopic behaviour of coarse dust particles. At mospheric Enviro rnn ent 42, 822-827. Sillanpaa, M., Hillamo, R., Saarikoski, S., Frey, A., Perrn anen, A., Makkonen, U., Spolnik, Z., Van Grieken, R., Branis, M., Brunekreef, B., Chalbot, M.C., Kuhlbusch, T., Sunyer, J., Kerrninen, V.M., Kulmala, M., Salonen, R.O., 2006. Chemical composition and mass closure of particulate matter at six urban sites in Europe. At mospheric Environment 40, S212 -S223. Simoneit, B.R. T., 1986. Characterization of Orga nic-Cons tituents in Aerosols in Relation to Their Origin and Tr ansport - a Review. International J oumal of Enviro rnn ental Analytical Chemistry 23, 207-237. Simoneit, B.R.T., 1999. A review of biomarker compounds as source indicators and tracers for air pollution. Enviro rnnental Science and Pollution Research 6, 159-169. Simoneit, B.R. T., Kobayashi, M., Mochida, M., Kawamura, K., Lee, M., Lim, H.J., Turpin, B.J., Komazaki, Y, 2004. Composition and major sources of organic compounds of aerosol particulate matter sampled dur ing the ACE-Asia campaign. J Geophys Res-Atmos 109. Singh, M., Jaques, PA., Sioutas, C., 2002. Size distribution and diurnal characteristics of particle-bound metals in source and receptor sites of the Los Angeles Basin. At mospheric Environment 36, 1675-1689. Smith, R.L., Spitzner, D., Kim, Y, Fuente s, M., 2000. Threshold dependence of mortality effe cts for fine and coarse particles in Phoenix, Arizona. J Air Waste Manage 50, 1367-1379. Snyder, D.C., Schauer, J.J., 2007. An inte r-comparison of two black carbon aero sol instruments and a semi-continuous elemental carbon instrument in the urban envirornn ent. Aerosol Sci Tech 41, 463-474. 146 Solomon, PA., Fall, T., Salmon, L., Cass, G.R., 1989. Chemical Characteristics ofPrnlO Aerosols Collected in the Los-Angeles Area. Japca-the Journal of the Air & Waste Management Association 39, 154-163. Solomon, PA., Fall, T., Salmon, L., Lin, P, Va squez, F., Cass, G.R., 1988. Acquisition of acid vapor and aerosol concentration data for use in dry deposition studies in the South Coast Air Basin, in: 25, E.Q.L. R.N. (Ed.). California Air Resources Board. Solomon, PA., Sahnon, L.G., Fall, T., Cass, G.R., 1992. Spatial and Temporal Distribution of At mospheric Nitric-Acid and Particulate Nitrate Concentrations in the Los-Angeles Area. Environ Sci Technol26, 1594-160 I. Sram, R.J., Binkova, B., Rossner, P, Ru bes, J., Topinka, J., Dejmek, J., 1999. Ad verse reproductive outcomes from exposure to environmental mutag ens. Mutation Research-Fundamental and Molecular Mechanisms of Mutagenesis 428, 203-215. Stayner, L., Dankovic, D., Smith, R., Steenland, K., 1998. Predicted lung cancer risk among miners ex posed to diesel exhaust part icles. Am J Ind Med 34, 207-219. Tao, F., Gonzalez-Flecha, B., Kobzik, L., 2003. Reactive oxygen species in puhnonary inflammation by ambient particula tes. Free Radical Biology and Medicine 35, 327-340. Targonski, PV., Persky, V.W., Ramekrishnan, V., 1995. Effect ofEnviro rnn ental Molds on Risk of Death from Asthma during the Pollen Season. Journal of Allergy and Clinical Immun ology 95, 955- 96 1. Taylor, S.R., McLennan, S.M., 1985. The Continental crust: Its Composition and Evolution. Blackwell, Oxf ord. Terzi, E., Argy ropoulos, G., Bougatioti, A., Mih alopoulos, N., Nikolaou, K., Samara, C., 2010. Chemical composition and mass closure of ambient PMIO at urban sit es. Atmospheric Environment 44, 223 1-2239. Turpin, B.J., Lim, H.J., 2001. Species contributions to PM2.5 mass concentrat ions: Revisiting common assumptions for estimating organic mass. Aero sol Sci Tech 35, 602-610. 147 Tursic, J., Podkrajsek, B., Grgic, I., Ctyroky, P., Berner, A., Dusek, U., Hitzenberger, R., 2006. Chemical composition and hyg roscopic properties of size-s egregated aerosol particles collected at the Adriatic coast of Slovenia. Chernosphere 63, 1193-1202. U.S.EPA, 2004. Air Quality Criteria for Particulate Matter. Office of Research and Development, National Center for Enviro rnn ental Asses sment, U.S. Environmental Protection Agency, Research Triangle Park, NC. U.S.EPA, 2006. National Ambient Air Quality Standards for Particulate Matter; Proposed Rule, 40 CFR Part 50. US Office of the Federal Register, National Archives and Records Adm inistration, Washington, DC. Usher, C.R., Michel, A.E., Grassian, V.H., 2003. Reactions on mineral dust. Chemical Reviews 103, 4883- 4939. Ver ma, V., Ning, Z., Cho, A.K., Schauer, J.J., Shafer, M.M., Sioutas, C., 2009a. Redox activity of urban quasi -ultrafine particles fr om primary and secondary sources . At mospheric Environment 43, 6360-6368. Ver ma, V., Pakbin, P., Cheung, K.L., Cho, A.K., Schauer, J.J., Shafer, M.M., Klei rnn an, M. T., Sioutas, C., 2011. Phys icochemical and oxidative characteristics of semi-volatile components of quasi -ultrafine particles in an urban atmosphere. At mospheric Environment 45, 1025-1033. Ver ma, V., Polidori, A., Schauer, J.J., Shafer, M.M., Cassee, F.R., Sioutas, C., 2009b . Phys icochemical and Toxi cological Profiles of Particulate Matter in Los Angeles during the October 2007 Southern California Wildf ires . Environ Sci Techn ol 43, 954-960. Vlasenko, A., Sjogren, S., Weingartner, E., Stemmler, K., Gaggeler, H. W., Ammann, M., 2006. Effect of humidity on nitric acid uptake to mineral dust aerosol part icles. Atm os Chern Phys 6, 2147-2160. Wall, S.M., John, W., Ondo, J.L., 1988. Measurement of Aerosol Size Distributions for Nitrate and Major Ionic Species. At mospheric Environment 22, 1649- 1656. Wang, G., Kawamura, K., Xie, M., Hu, S., Gao, S., Cao, J., An, Z., Wang, Z., 2009. Size-dist ributions of n-alkanes, PAHs and hopanes and their sources in the urban, mountain and marine atmospheres over East Asia. Atm os Chern Phys 9, 8869-8882. 148 Wang, Y, Arellanes, C., Curti s, D.B., Paulson, S.E., 2010. Probing the Source of Hydrogen Peroxide Associated with Coarse Mode Aero sol Part icles in Southern California. En viron Sci Techno l4 4, 4070-4075. Winiwarter, W., Bauer, H., Caseiro, A., Puxbaum, H., 2009. Quantifying emissions of primary biological aero sol particle mass in Europe. At mospheric Environment 43, 1403-1409. Wu, P.M., Okada, K., 1994. Nature of Coarse Nitrate Particles in the Atmosphere -a Single, Particle Approach. At mospheric Enviro rnn ent 28, 2053-2060. Yeatts , K., Svendsen, E., Crea son, J., Alexis, N., Herbst, M., Scott, J., Kupper, L., Wi lliams, R., Neas, L., Cascio, W., Devlin, R.B., Peden, D.B., 2007. Coarse particulate matter (PM2.5-10) affects heart rate variability, blood lipids, and circulating eosinophils in adults with asthma. Environ Health Persp 115, 709-714. Yoshizumi, K., Hoshi, A., 1985. Size Distributions of Ammonium-Nitrate and Sodium-Nitrate in At mospheric Aer osols. Environ Sci Techno! 19, 258- 261. Yu e, Z. W., Fraser, M.P., 2004. Polar organic compounds measured in fine particulate matter during TexAQS 2000. At mospheric Enviro rnn ent 38, 3253- 326 1. Zereini, F., Wi seman, C., Alt, F., Messerschmidt, J., Muller, J., Urban, H., 200 1. Platinum and rhodium concentrations in airborne particulate matter in Germany from 1988 to 1998. Environ Sci Technol35, 1996-2000. Zhang, D.Z., Shi, G.Y, Iwasaka, Y, Hu, M., 2000. Mixture of sulfate and nitrate in coastal atmospheric aerosols: individual particle studies in Qingdao (36 degrees 04 ' N, 120 degrees 21 'E ), China. At mospheric Enviro rnnent 34, 2669-2679. Zhang, T., Claeys, M., Cachier, H., Dong, S.P., Wang, W., Maenhaut, W., Liu, X.D., 2008a. Identification and estimation of the biomass burning contribution to Beijing aerosol using levoglucosan as a molecular marker. At mospheric Enviro rnnent 42, 7013-7021. Zhang, YX., Schauer, J.J., Shafer, M.M., Hannigan, M.P., Dutton, S.J., 2008b. Source apportio rnnent of in vitro reactive oxygen species bioassay activity from atmospheric particulate matter. Environ Sci Technol42, 7502-7509. 149 Zhang, YY, Muller, L., Winterhalter, R., Moortgat, G.K., Hoffmann, T., Posch!, U., 2010. Seasonal cycle and temperature dependence of pinene oxidation products, dicarboxylic acids and nit rophenols in fine and coarse air particulate matter. Atrnos Chern Phys 10, 7859-7873. Zhao, YL., Gao, Y, 2008a. Acidic species and chloride depletion in coarse aerosol particles in the US east coast. Sci Total Environ 407, 541- 547. Zhao, YL., Gao, Y, 2008b. Mass size distributions of water-soluble inorganic and organic ions in size-seg regated aerosols over metropolitan Newark in the US east coast. At mospheric Environment 42, 4063-4078. Zhuang, H., Chan, C.K., Fang, M., Wexler, A.S., 1999. Formation of nitrate and non-sea-salt sulfate on coarse part icles. At mospheric Enviro rnn ent 33, 4223-4233. Zygadlo, J.A., Maestri, D.M., Grosso, N.R., 1994. Alkane Distribution in Epicuticular Wax of Some Solanaceae Species. Biochemical Systematics and Ecology 22, 203-209. !50

Abstract (if available)

Abstract To advance our understanding on the relationship between the sources, chemical composition and toxicity of coarse particles, two comprehensive investigations were conducted in the Los Angeles Basin from 2008 to 2010 to characterize the physico-chemical and toxicological properties of ambient coarse particulate matter (CPM). The first study features of a year-long sampling campaign at 10 sampling sites throughout the basin in an attempt to study the spatial and seasonal characteristics of coarse particles. An intensive study, focusing on the diurnal trends, was conducted at 3 sampling sites to examine how the changes in meteorological conditions throughout a day may affect the source strength and formation mechanisms of coarse mode aerosols. The results from these two studies have become an invaluable tool to advance our understanding on the causes of the temporal and spatial variation of CPM concentrations and chemical composition, and to identify the linkage between toxicity levels and source-specific chemical constituents in CPM. This investigation thereby provides valuable scientific information for environmental policy decision making, and a strong scientific basis to develop cost-effective strategies to protect the public health from CPM exposure.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Size-resolved particulate matter (PM) in urban areas: toxico-chemical characteristics, sources, trends and health implications

PDF

Toxicological characteristics of particulate matter in an urban environment and their linkage to the source-specific constituents

PDF

Particulate matter (PM) exposure for commuters in Los Angeles: chemical characterization and implications to public health

PDF

Investigation of physico-chemical characteristics of particulate matter from vehicular sources

PDF

Analysis of sources and profiles of organic carbon in ambient particulate matter across fine and coarse sizes and introducing an optical technique for real-time urban dust measurement

PDF

Development of novel techniques for evaluating physical, chemical and toxicological properties of particulate matter in ambient air

PDF

Measurement and methods of assessing the impact of prevalent particulate matter sources on air quality in southern California

$Chemical and toxicological characteristics and historical trends of size-fractioned particulate matter from traffic-related emissions in Los Angeles$

PDF

Chemical and toxicological characteristics and historical trends of size-fractioned particulate matter from traffic-related emissions in Los Angeles

PDF

Characterization of black carbon: from source to evolution of physical and optical properties in the atmosphere

$Exposure assessment and source apportionment of size fractions of airborne particulate matter$

PDF

Exposure assessment and source apportionment of size fractions of airborne particulate matter

PDF

Influence of atmospheric processes on the physico-chemical properties of primary particles and their impact on public exposure

PDF

Oxidative potential of urban atmospheric particles: spatiotemporal trends and associations with source-specific chemical components

PDF

Chemical and toxicological characteristics of particulate matter in urban environments with a focus on its sources, associated health impacts and mitigation policies

PDF

Physico-chemical characteristics and sources of ambient PM mass and number concentrations and their associated toxicity, and development of novel techniques for high time-resolution measurement o...

PDF

Investigating the temporal trends, sources, and toxicity of ambient particulate matter (PM) in different metropolitan environments, and development of a novel aerosol generation setup for inhalat...

$Physico-chemical properties and source apportionment of size-fractionated airborne particulate matter in urban areas with implications for public health$

PDF

Physico-chemical properties and source apportionment of size-fractionated airborne particulate matter in urban areas with implications for public health

PDF

Impact of urban source emissions on ambient particulate matter (PM) composition and associated toxicity in local and regional scales using source appointment models

PDF

Assessing indoor and outdoor air quality by characterizing the physicochemical and toxicological properties of particulate matter using…

PDF

Investigating the role of urban emission sources on redox-active PM compounds and the chemical analysis of the standardized diesel exhaust particles

PDF

Ultrafine particle concentration inside vehicles: models for exposure assessment

Asset Metadata

Creator Cheung, Kalam (author)

Core Title Temporal, spatial and toxicological characteristics of coarse particulate matter in an urban area and relation to sources and regulations

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Environmental Engineering

Publication Date 06/21/2012

Defense Date 05/22/2012

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag coarse particle,health effects,OAI-PMH Harvest,particulate matter,reactive oxygen species,regulations

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Sioutas, Constantinos (committee chair), Chen, Jiu-chiuan (committee member), Fruin, Scott (committee member), Henry, Ronald C. (committee member)

Creator Email kalam@ucla.edu,kalamche@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-49632

Unique identifier UC11290021

Identifier usctheses-c3-49632 (legacy record id)

Legacy Identifier etd-CheungKala-897.pdf

Dmrecord 49632

Document Type Dissertation

Rights Cheung, Kalam

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA