Aerosol and Air Quality Research, 10: 193–211, 2010 Copyright © Taiwan Association for Aerosol Research ISSN: 1680-8584 print / 2071-1409 online doi: 10.4209/aaqr.2009.06.0042
Particulate Air Pollution in Mexico City: A Detailed View Elizabeth Vega1*, Silvia Eidels1, Hugo Ruiz1, Diego López-Veneroni1, Gustavo Sosa1, Eugenio Gonzalez1, Jorge Gasca1, Virginia Mora1, Elizabeth Reyes1, Gabriela Sánchez-Reyna1, Rafael Villaseñor1, Judith C. Chow2, John G. Watson2, Silvia A. Edgerton3 1
Instituto Mexicano del Petróleo, Eje Central Lázaro Cárdenas Núm. 152. Col. San Bartolo Atepehuacan. Delegación Gustavo A. Madero. 07730, México, D. F. 2 Desert Research Institute, 2215 Raggio Parkway, Reno, NV 89512-1095, USA 3 National Science Foundation, Arlington, Virginia, USA ABSTRACT A detailed view of particulate air pollution in Mexico City resulting from several analyses of data collected during four field campaigns is given. The resulting database from the March 1997 campaign was used to identify and quantify the contribution of particles to the atmosphere from different emission sources, to estimate transport and dispersion of pollutants and to estimate the amount of secondary organic carbon. Emission inventories of PM10, PM2.5 and ammonia were also calculated. Data from March 1997 and March 2002 campaigns were used to estimate aerosol optical properties and its impact on visibility, and measurements during March and November 2001 and March 2002 were used to obtain particle mass size distributions. Results showed that major contribution of PM2.5 were the mobile source emissions with 45%. Computer simulations were run for a wind blown dust episode and results agreed with observations. Particle concentrations were found to be inversely related to transport wind speed, and the highest pollutant dispersion was in the afternoon as calculated with the ventilation index. An average of 25% of the total organic carbon in the PM2.5 was associated to secondary organic aerosol estimated with an empirical model. Particles of sizes between 0.1 and 1.0 μm accounted for the highest mass concentrations and were associated mainly to primary emissions. The earlier visible light absorption peaks that appeared in the diurnal patterns were attributed to the elevated elemental carbon vehicular emissions during the heavy traffic hours while the later light scattering peaks were attributed to secondary aerosol formed photochemically in the atmosphere. Dairy and non-dairy cattle were the dominant sources according to the calculated emission inventory (EI) for ammonia. PM2.5 mobile sources derived from the EI were 11% whereas those estimated with the receptor model were 45%. Keywords: Particles; Fine aerosols; Air pollution; Mexico City.
INTRODUCTION Air pollution is a persistent and pervasive environmental problem with health implications and economic costs on society. Particles suspended in air from both anthropogenic and natural sources have been estimated to account for a large fraction of air pollutants in Mexico City (Miranda et al., 1996; Vega et al., 1997; Aldape et al., 2005; Vega et al., 2009). Although PM10 concentration has decreased by half during the past 20 years (SMA, 2007), its average concentration is still relatively high when compared to other megacities and to international health standards.
*
Corresponding author. Tel.: (5255) 9175-6919; Fax: (5255) 9175-8272 E-mail address:
[email protected]
Particles with diameters between 0.1 and 1.0 μm have long residence times in the atmosphere, are responsible for visibility reduction and have the potential to produce respiratory diseases. In addition, they can be transported for long distances at different altitudes and therefore contribute to local and global climate changes. With government expenditures on pollution controls estimated to be in several millions of dollars per year and slow progress toward compliance with air quality standards, there is a pressing need for better and more effective abatement strategies. An integrated approach which includes the measurement, chemical characterization and modeling of production and transport of particles with different sizes is helpful in understanding the air pollution problem and its local and regional impact. The Instituto Mexicano del Petróleo (IMP), in collaboration with several national and international institutions, conducted the project Investigación sobre Materia Particulada y Deterioro
194
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
Atmosférico-Aerosol and Visibility Evaluation Research (IMADA-AVER). The main objectives of this project were to characterize the nature and causes of particulate concentrations and visibility impairment in and around Mexico City to further characterize the major sources contributing to significant chemical components of PM10, PM2.5, and light extinction. The IMADA field campaign took place in the winter of 1997 (Doran et al., 1998; Edgerton et al., 1999; Chow et al., 2002) which included sampling of particles from different types of land use located within Mexico City Metropolitan Area (MCMA) (Fig. 1 and Table 1). In this paper efforts were directed to complement the published results of the IMADA-AVER project within the frame of its objectives. The resulting database has been used to: identify and quantify the contribution of particles to the atmosphere from different emission sources; estimate pollutant transport and to evaluate the behavior of the daytime mixing height, transport wind speed and direction and ventilation index; estimate the amount of secondary organic carbon; and determine aerosol optical properties and its impact on visibility. Additionally the emission inventories of PM10 and PM2.5 were calculated as they were necessary to simulate air quality. The results of the aforementioned as well as those derived from data on particles size distribution taken during March 2002 and from the mass
concentration of particles collected by sizes during March and November 2001 and March 2002 are presented in this paper. To date a considerable number of publications have been produced as a result of the recent research developed in MCMA during the Megacity Initiative: Local and Global Research Observations (MILAGRO) in 2006. The initiative was focused on aerosols due to their contribution to climate change (Forster et al., 2007) and their harmful effect on human health (Osornio-Vargas et al., 2003). According to this research, an important fraction of particulate matter is constituted by carbonaceous species (Edgerton et al., 1999; Querol et al., 2008; Aiken et al., 2009) and mostly as organic carbon. Modeling was also carried out for the production of secondary organic carbon (Volkamer et al., 2006) and the simulated results were compared with the analysis of the evolution of aerosols with photochemical age (Kleinman et al., 2008). Other studies on organic carbon have shown the contribution of primary and secondary sources (Schauer, 2003; Johnson et al. 2005; Yokelson et al., 2007; DeCarlo et al., 2008; Stone et al., 2008; Aiken et al., 2008; 2009). A number of stateof-the-art techniques were applied during MILAGRO field campaign in 2006 to fully characterize primary and secondary organic aerosol in the MCMA (Hennigan et al., 2008; Moffet et al., 2008a, b; Stone et al., 2008) however,
Fig. 1. Mexico City Metropolitan Area study zone (in UTM) sampled between February 23 and March 22, 1977. Base sites: Xalostoc (XAL), La Merced (MER), Cerro de la Estrella (CES), Tlalnepantla (TLA), Netzahualcóyotl (NET) and Pedregal (PED). Regional sites: Cuautitlan (CUA), Teotihuacan (TEO), Chalco (CHA) and Universidad Nacional Autonoma de Mexico (UNAM).
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
195
Table 1. Sampling sites and locations. Samples considered in this study were collected during February 23 and March 22, 1997. Traffic Road Fuel source Activity Site Site/Sector frequency characteristics Xalostoc (XAL) NE Paved and Gasoline and diesel old Industrial Heavy unpaved and new vehicles Merced (MER) C Paved Heavy-duty diesel Residential, commercial, Heavy buses and Light duty services diesel vehicles Old and new gasoline Residential, commercial Heavy Base sites Cerro de la Estrella (CES) SE Paved and unpaved vehicles Tlalnepantla (TLA) NW Paved Gasoline vehicles Industrial, residential Light Netzahualcóyotl (NET) NE Paved and Old and new gasoline Commercial, residential Heavy unpaved and diesel vehicles Pedregal (PED) SW Paved Gasoline vehicles Residential Light Cuautitlán (CUA) NW Unpaved Old and new gasoline Agricultural, industrial Heavy and diesel vehicles Teotihuacan (TEO) NE Unpaved Old and new gasoline Commercial, residential, Light and diesel vehicles agricultural Regional Chalco (CHA) SE Unpaved Old and new gasoline Residential, Agricultural Light sites and diesel vehicles Universidad Nacional Paved and Gasoline vehicles Residential Light Autonoma de Mexico unpaved (UNAM) SW during the IMADA-AVER study those techniques were not available; thus, primary and secondary organic aerosols estimation by an alternative approach is still worthwhile. The contribution of primary and secondary aerosol to light scattering and absorption (Paredes-Miranda et al., 2009) as well as measurements of aerosol absorption and scattering at the urban area of Mexico City and rural sites (Marley et al., 2009) during the MILAGRO campaign have also been reported. METHODS Particle Sampling During the IMADA-AVER project, particles were sampled within the MCMA between February 23 and March 22, 1997. Six core-sites (Xalostoc (XAL), La Merced (MER), Cerro de la Estrella (CES), Tlalnepantla (TLA), Netzahualcóyotl (NET), and Pedregal (PED)) with different land use were located within the city. Four regional-scale sites (Cuautitlan (CUA), Teotihuacan (TEO), Chalco (CHA) and Universidad Nacional Autónoma de México (UNAM)) on the periphery of the city were selected to represent nearby agricultural areas within 50 km from the Mexico City. Results of chemical composition and gravimetric analysis have been previously described (Chow et al., 2002a; Vega et al., 2002, 2003) and site description and methodologies were mentioned elsewhere (Edgerton et al., 1999). Two different particles sizes (PM2.5 and PM10) from the six core-sites were collected with sequential filter samplers (mean flow of air intake of 113 L/min). Daily 24-h samples were taken at the TLA, NET, and PED sites. Diurnal 6-h samples (00:00–06:00, 06:00–12:00, 12:00–18:00 and
18:00–24:00 h) were taken daily, four times per day, at the XAL, MER, and CES sites. At regional sites daily 24-h PM10 samples were collected with Minivol samplers, which operate with a mean flow of 5 L/min. Particles were also sampled in a 10-stages (10, 5.6, 2.5, 1.8, 1.0, 0.56, 0.32, 0.18, 0.10 and 0.056 µm) MOUDI sampler (Micro-Orifice Uniform Deposit Impactor, model 110, MSP Co.) operating at a mean flow of 27 L/min. The MOUDI sampler, placed 3 m above ground at La Merced (MER), collected particles every 24 h for 10 days in March 2001, 6 days in November–December 2001 and 12 days in March 2002. Particles for gravimetric and elemental analyses were collected on Teflon filters (2 μm pore). An aluminum foil, placed at the MOUDI zero-stage with 18µm cut-size, was covered with a thin layer of silicon grease to reduce particle bounce (Hinds, 1998). Particle Composition and Chemical Mass Balance Model Gravimetric analysis was performed by weighing the filters before and after sampling. Filters were previously stabilized at a temperature of 21.5 ± 0.5°C and relative humidity of 25–35%. Particles were also analyzed for different chemical species. X-ray fluorescence was used for elemental analysis. Soluble ions were determined on quartz filters by ion chromatography (Cl-, NO3- and SO4=), atomic absorption spectrometry (Na+ and K+), and colorimetry (NH4+). Organic carbon and elemental carbon were determined by thermal optical reflectance. Results of chemical composition and gravimetric analysis have been previously described (Chow et al., 2002; Vega et al., 2002). Particle mass and composition were analyzed to determine the origin and the contribution of major sources of PM2.5 by using chemical mass balance (CMB) receptor model
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
196
and to provide inputs for air quality models. The following Eq. (1) expresses the relationship between the concentrations of the chemical species measured at the receptor with those emitted by the source: Ci =
P
Fij ⋅ S j ∑ j= i
(1)
where Ci = ambient concentration of the species i measured at the receptor site; P = Number of sources j that contribute; Fij = fraction of the emissions of the species i from the source j; and Sj = impact of source j to the receptor (calculated source contribution). Different particulate matter source profiles were determined in Mexico City and further details about the source sampling have been previously described (Vega et al., 2009) and similar sampling procedures of sources and application of the model were mentioned elsewhere (Watson et al., 1994; Vega et al., 2000). Primary and Secondary Organic Carbon Primary OC (OCP) is emitted directly to the atmosphere, whereas secondary OC (OCS) is formed in the atmosphere through chemical reactions of reactive organic gases and subsequent gas-to-particle partitioning process (Chu, 2005). Elemental carbon (EC) is produced during the incomplete combustion of fossil fuels and is linearly correlated with OCP, since they are sub products of the combustion. The semi-empirical EC tracer method has been applied successfully as an alternative approach to estimate secondary organic aerosol production (Turpin et al., 1991; Turpin and Huntzicker, 1995; Castro et al., 1999; Yu et al., 2009). In this work this method was applied to observations gather at three sites in the urban core of the MCMA (XAL, MER and CES). The total organic carbon (OCT) in the aerosol phase is given by the sum of OCP and OCS. Therefore, if the OCP is determined accurately, the secondary production can be estimated by OCS = OCT – OCP
(2)
To estimate OCP accurately, a set of linear regression analyses was performed with the available data of OCT and EC. The best linear regression fit is when the contribution of OCS is low or zero, suggesting that OCT is mainly primary. Therefore, OCT = OCP = a + b × EC and, b = (OCT/EC)P
(3) (4)
Emission Inventory of PM10, PM2.5 and Ammonia The emission inventory (EI) is the result of emission factor times the activity, being the emission factor the amount of pollutants released to the atmosphere per unit of activity. The activity is the variable with highest uncertainty as it is defined in terms of production, the use of fuel or distance traveled among others. In general, the
emission factors are obtained from the AP-42 compilation of the Environmental Protection Agency (1995, 1995a). First EI for Mexico City was reported in 1989 (DDF et al., 1996) since then, efforts have been made to obtain itemized and more precise calculations, which is the case of the 1998 EI. In this EI, estimation for PM10 was included for the first time with a total of 19,889 tons per year released from 14 industrial entries, 7 area sources including the emissions due to wind erosion and 12 mobile sources. Industrial sources represented 15.5% of the total; the area sources 8.5%, the erosive sources contributed 40%, and the mobile sources 36% among others (CAM, 2001). It should be mention that this EI did not include PM2.5 and ammonia, which is an important secondary inorganic aerosol precursor. Therefore in this work the emission inventories of ammonia, PM10 and PM2.5, including dust re-suspension from paved and unpaved roads, were calculated for February 1997 to support transport and source apportion modeling. Meteorology This section focuses on the factors which determine dilution and transport of pollutants such as mixing height, transport wind speed (defined as average wind speed in the mixing height) and direction and ventilation index (VI) (Tennekes, 1974), which are essential in the study of pollution dynamics. In particular, determination of the atmospheric mixing layer (ML) and the wind direction and velocity are critical in determining pollutant concentration and transportation. Table 2 summarizes the meteorological parameters measured at the regional scale sites: CUA at the north, TEO at northeast, CHA at east and UNAM at southwest, during 1 to 20 of March 1997. Wind was measured with wind profilers while sondes were launched daily in most stations to measure atmospheric pressure, dry-bulb temperature and relative humidity. To calculate the ML the profiles of potential temperature and mixing ratio were plotted, and the height of the mixing layer was identified with the base of an elevated capping inversion or stable layer, and when a significant reduction in air moisture occurred (Seibert et al., 2000). Mixing height (MH) is used to quantify the extension in which vertical mixing takes place, whereas the average vector wind in the boundary layer determines the expected direction and speed at which air pollutants will be dispersed. One variable that relates the mixing height and transport wind speed and direction is the VI, which is the product of both variables. The VI indicates the relative degree of air circulation for a specific area and time period, and it is a measurement of the ability of the atmosphere to clean itself and to disperse pollutants. Wind patterns were modeled with the Regional Atmospheric Modeling System (RAMS) (Pielke et al., 1992; ASTER Division, 1994, 1999; Tremback and Walko, 2000). The simulation domain consisted of three nested grids: the outermost grid included part of the Gulf of Mexico and the Pacific Ocean, and the innermost grid included the MCMA basin and the surrounding mountain ranges (Fig. 2). Grid parameters are summarized in Table 3.
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
197
Table 2. Meteorological parameters measured at Cuautitlan (CUA), Teotihuacan (TEO), Chalco (CHA) and Universidad Nacional Autónoma de México (UNAM) during 1 to 20 of March 1997. Sondes were launched daily in most stations which measured atmospheric pressure, dry-bulb temperature and relative humidity. Sampling site Sector Elevation (m) Wind profiler Radio-sound CUA N 2252 X X TEO NE 2275 X X CHA E 2248 X X UNAM SW 2274 X X * Includes pressure, temperature, humidity and wind measurements.
Sodar X X
Surface sampling X X X X
Total* 76 78 82 77
Fig. 2. Grid configuration for the RAMS simulation. a): macroscale (Grid 1), mesoscale (Grid 2) and microscale (Grid 3) nested grids. b): microscale grid with elevations in m. Table 3. Grid parameters used in the Regional Atmospheric Modeling System (RAMS) during 1 to 20 of March 1997. Number of cells
Horizontal resolution (km)
Coarse
56 × 56
32 × 32
Intermediate
50 × 50
8×8
Fine
54 × 54
2×2
Grid Size
Dispersion Modeling A 48-hour air quality simulation was run for the period March 5–6, 1997 to estimate wind erosion and dust emission owing to the contrasting meteorological features and distinct air quality conditions for these days. The CALMET/CALPUFF modeling system was selected to simulate meteorological conditions and pollutant transport of a wind blown dust event in the MCMA. Three main components comprise this modeling system: CALMET (a diagnostic 3-D meteorological model (Scire et al., 1997)), CALPUFF (the transport and dispersion model (Scire et al., 2000)) and CALPOST (a post-processing package). In this study only surface meteorological observations and upper air soundings were used to drive CALMET.
Vertical resolution (m) 43 grid points (x) between ground level and 23,206 m. Grid spacing: Lx = L(x-1) × 1.1 where L0: 50 m, L30: 4600 m
Constant parameters Soil: salty sand Rugosity: 2.0 m Albedo: 0.15 Surface water temperature: 280 K
A wind erosion equation approach was applied to generate the necessary input to a transport-dispersion simulator for the selected event that caused regional particulate concentrations to exceed the 120 µg/m3 ambient standard. The algorithm that predicted dust concentrations was developed with data taken from wind tunnels and field studies to produce a wind-erosion-prediction equation (Hagen et al., 1996). Details on the wind erosion equation and the parameters used in this equation for the MCMA soil sources can be found elsewhere (GDF, 2001). A geographical information system (GIS) was incorporated to obtain a series of maps to be used as a framework for the system over which the project’s results can be displayed and manipulated under a unique
198
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
consistent and flexible database. Each municipal local authority was digitalised in Universal Transverse Mercator (UTM) coordinates, and then inserted in the system as separate entities, maximising the system’s flexibility (ESRI, 1999; Ellsworth-Davis, 2001). Aerosol Optical Properties and Visibility Aerosol optical properties are expressed in terms of light scattering (bsp) and absorption (bap) coefficients and depend on particle size distribution, chemical composition and refractive indices. Visibility (Vis), as given by the well known Koschmieder relationship Vis = 3.912/bext, is related to total light extinction. The total light extinction coefficient (bext) is the sum of the scattering and absorption coefficients of gases (bsg and bag) and particles (bsp and bap) and can be expressed as: bext = bag + bsg + bap + bsp = beg + bep,
(5)
where: beg = bag + bsg is the extinction coefficient of gases and bep = bap + bsp is the extinction coefficient of particles. The gaseous contributions to light extinction are limited in the visible region to absorption by NO2 and to Rayleigh scattering due to O2 and N2. Particle scattering is produced by a wide range of aerosol compositions whereas particle absorption is only due to elemental carbon (EC) in the visible region (Rosen et al., 1978; Yasa et al., 1979; Japar et al., 1986; Adams et al., 1990). In urban atmospheres, fine particles are usually the main contributors to visible light extinction (Adams et al., 1990a). A study of the optical properties of fine particles and their contribution to total light extinction as well as their impact on visibility was performed for the period 28 February to 10 March 1997 at the downtown La Merced (MER) and the residential Pedregal (PED) sites, and for the period 11–25 March 2002 at MER site. The first sampling period corresponds to 11 days of the IMADA-AVER intensive field campaign (Edgerton et al., 1999) and the second to one of the 4 field campaigns conducted under the project “Integrated assessment to understand the particulate matter air pollution problem in Mexico City and its coupling to regional and global impacts through the use of regional air quality models”. Diurnal (6:00–18:00 h) visible light absorption and scattering coefficients of gases and particles were determined using wavelengths of 530 nm and 550 nm for the 1997 and 2002 periods of study, respectively. The gaseous coefficients were calculated using the temperature and NO2 concentration data of the atmospheric automated monitoring network (RAMA). The aerosol coefficients were obtained, for the 1997 period, from measurements of light absorption with an aethalometer (Magee Scientific, Co., Model AE-16) operating at 880 nm and measurements of light scattering with an integrating nephelometer (Optec Inc., Model NGN2) operating at 530 nm while, for the 2002 period, they were obtained from Mie theory using the particle size distribution in the diameter range 15.7–791 nm measured with a Scanning Mobility Particle Sizer, SMPS, (TSI,
Model 3936 L10) and the refractive index determined as described in Eidels-Dubovoi (1993). RESULTS AND DISCUSSION Chemical Composition and Sources of PM. Basic statistics for average concentrations of PM10 and PM2.5 during February 23 to March 22, 1997 are shown in Table 4. Highest average 24-h values for both PM2.5 and PM10 were observed at the NE sector (XAL and NET) and decreased to the SW (PED), consistent with the highest industrial and traffic sources of the NE sector of the MCMA (Vega et al., 2002). The largest concentration variability for PM10 occurred in the eastern and central sectors and decreased at PED and XAL. A comparison for 24-h and 6-h sampling periods shows that there was considerable variability (up to 30 %) at sampling periods below 24 h at all sites (Vega et al., 2002). Time series analysis of the sampling sites shows that the PM2.5 USA 24-h average standard (65 µg/m3) was exceeded four times at the eastern (NET) and northeastern (XAL) sectors. In turn, the 120 µg/m3 Mexican standard for PM10 was exceeded 3 times at each site (NET and XAL) and once at CES, as a result of the prevailing meteorological conditions during the sampling campaign which favors highest particle concentrations during weak high-pressure systems and when mixing layer heights were lowest. The PM2.5/PM10 ratio suggests relatively highest concentrations of coarse particles at the eastern sectors relative to the southwest (PED) and central (MER) sites, which are due to dust resuspension. The chemical composition of PM2.5 was dominated by carbon, sulfate, nitrate, ammonium and crustal components, with site- and time-dependent variations. Carbonaceous aerosols comprised over half of the PM2.5 mass while 30% were secondary inorganic aerosols and 15% geological material. In contrast, PM10 was dominated by geological material, 30% carbonaceous aerosols and 17% secondary inorganic aerosols (Chow et al., 2002). Chemical mass balance model results (Vega et al., 2009) showed significant spatial variations in source contributions among the six sites (Table 5), which are influenced by local soil types and land use. Geological contributions (dust) to PM2.5 varied from 9% at the residential area of PED site to over 38% at NET. Mobile sources dominated the PM2.5 mass at XAL, MER and CES with an average of 57%, and with an average of 33% at NET, PED and TLA. These results agree with those reported by Stone et al. (2008) where motor vehicles, accounted for 49% of ambient OC in the urban area of Mexico City and 32% on the rural area. Secondary inorganic aerosols were dominated by ammonium sulfate with an average of 14%, showing the highest percentage at PED (22%) and the lowest at NET (8%). Ammonium nitrate showed a more homogeneous distribution at all sites, except at XAL. Incineration, characterized by Zn, Pb and Cl-, which represent industrial metal emissions or open waste burning, was observed at all sites varying its contribution from 11– 2%, as expected, the industrial area (TLA and XAL) showed the highest contribution. Pb and Zn were also
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
199
Table 4. Average PM10 and PM2.5 concentration basic statistics in the Base and Regional sites during February 23 to March 22, 1997. Location Merced (MER) Xalostoc (XAL) Pedregal (PED) Tlalnepantla (TLA) Cerro de la Estrella (CES) Nezahualcóyotl (NET) Teotihuacan (TEO) Ajusco (AJU) Cuautitlán (CUA) Chalco (CHA)
PM10 57.3 ± 22.0 104.2 ± 33.2 39.4 ± 12.2 58.9 ± 11.1 61.2 ± 20.7 112.7 ± 57.8 46.1 ± 37.7 47.5 ± 17.9 44.6 ± 13.2 73.5 ± 26.9
PM2.5 36.0 ± 9.7 44.8 ± 11.5 21.6 ± 7.0 30.7 ± 8.9 29.8 ± 9.4 52.4 ± 31.5 ND ND ND ND
PM2.5/PM10 0.61 ± 0.14 0.46 ± 0.11 0.57 ± 0.12 0.51 ± 0.22 0.49 ± 0.12 0.44 ± 0.13 -
Table 5. 24-h average sources contribution (µg/m3) of PM2.5 determined by chemical mass balance receptor model at Xalostoc (XAL), La Merced (MER), Cerro de la Estrella (CES), Tlalnepantla (TLA), Netzahualcóyotl (NET) and Pedregal (PED) during March, 1997. XAL 49 ± 3 31.59 ± 4.94 7.89 ± 0.78 5.04 ± 0.73 3.87 ± 0.30
MER 38 ± 3 22.38 ± 3.34 5.43 ± 0.84 4.50 ± 0.65 0.98 ± 0.47 2.19 ± 0.18 0.42 ± 2.46 0.06 ± 0.26
CES 32 ± 3 15.83 ± 10.40 4.86 ± 1.19 4.35 ± 0.88 1.10 ± 0.80 1.22 ± 0.13 0.85 ± 2.96 2.92 ± 2.12 0.11 ± 0.26
associated with the industrial areas in northern Mexico City (Johnson et al., 2005). Cooking contributions to PM2.5 were observed only at MER, CES, PED and TLA and ranged from 9% at CES to 5% at PED. Open or biomass burning contributed at CES, NET and PED with 9%, 5% and 13%, respectively. Results from the MILAGRO campaign suggest that biomass burning contributed significantly to both gas and particles pollution, although the magnitude of the contribution of this source is quite variable (Moffet et al., 2008b; Querol et al., 2008). Other results reporting daily contribution of biomass burning to OC ranged from 5–26% at the Mexico City urban site and 7–39% at the rural site (Stone et al., 2008). Finally, emissions of heavy oil combustion, characterized by V and Ni, for power generation or local industrial contributions were detected at low levels (0.2%) and large uncertainties, indicating collinearity with other sources. Particle Mass Size Distribution Particles collected in March 2001 and 2002 and November–December 2001 with the MOUDI sampler at La Merced site grouped together in three different modes (Fig. 3) The first mode in the nucleation range (diameters < 0.1 μm) consisted of particles generated from combustion processes (Seinfeld and Pandis, 1998). The second mode in the accumulation range (diameters between 0.1 and ~2 μm)
TLA 34 ± 2 11.25 ± 8.22 3.86 ± 0.74 6.38 ± 0.82 0.44 ± 0.63 3.73 ± 0.28 3.33 ± 1.76 0.55 ± 0.20
NET 58 ± 3 19.78 ± 9.33 22.30 ± 1.34 4.71 ± 0.78 0.84 ± 0.67 0.89 ± 0.09
PED 25 ± 1 7.97 ± 3.49 2.14 ± 0.39 5.48 ± 0.55 0.92 ± 0.29 0.40 ± 0.04 1.97 ± 1.21 3.24 ± 1.29 0.12 ± 0.08
3.18 ± 2.58 0.08 ± 0.12
50
dM/d(LogDp) [µg m-3 µm-1]
PM2.5 Measured mass Mobile sources Geological (Dust) Ammonium sulphate Ammonium nitrate Incineration Cooking Open burning Oil combustion
40
30
20
10
0 0.01
0.1
1
10
Diameter [µm] March 2001
Nov-Dec 2001
March 2002
Fig. 3. Average particle size distribution (MOUDI) at La Merced collected during March and November 2001, and March 2002. with the highest concentration of particles that probably result from a combination of condensed material and primary emissions. This mode consisted of two sub-modes:
200
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
the condensation sub-mode (size range 0.1–0.2 μm) which contains gas-phase reactions products, and showed only in the March 2001 and 2002 campaigns; and the droplet submode (size range 0.2 to 1.0 μm) that results from reactions that take place in water droplets (Chow, 1995; Pandis, 2003). The third mode in the coarse range (diameters between 2.5 and 10 µm) consisted of particles resulting from grinding activities and dust resuspension (Chow, 1995). Fig. 4 shows that the concentrations of particles greater than 0.32 μm in diameter were highly correlated with the concentration of PM2.5. In turn, the lack of correlations between particle mass concentration with diameters smaller than 0.18 μm and PM2.5 suggests, that these particles have a different source of formation, e.g. smallest particles from nucleation mode growth and the larger particles from primary emissions. The significant decrease of the mass concentration correlation between PM2.5 and particles of
sizes lower than 0.32 μm observed in the MCMA, was also reported for the Pittsburgh, Pennsylvania area (Cabada et al., 2004). The average total mass concentration of the three campaigns comprised the following percentages of the various stages of the MOUDI cut point sizes: 12.3 ± 2.60% of 10 µm, 10.73 ± 1.38% of 5.6 µm, 6.34 ± 0.40% of 2.5 µm, 8.12 ± 1.78% of 1.8 µm, 4.93 ± 0.07% of 1.0 µm, 10.43 ± 0.88% of 0.56 µm, 20.69 ± 5.25% of 0.32 µm, 11.75 ± 2.60% of 0.18 µm, 9.31 ± 2.23% of 0.10 µm, 3.80 ± 1.67% of 0.056 µm and 1.51 ± 0.57 of 18 µm. Fig. 5 shows the daily average particle mass concentrations for the different MOUDI cut sizes, from PM10 to PM0.056, during the three sampling campaigns at La Merced. A seasonal dependence is observed for all sizes of particles, with lower mass concentrations during November 2001 and higher during March 2001 and 2002.
Fig. 4. Average scatter plots of a) PM10, b) PM1.0, c) PM0.56, d) PM0.32, e) PM0.18, f) PM0.1 versus PM2.5 collected with MOUDI sampler in La Merced during March and November 2001 and March 2002.
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
201
Fig. 5. Daily average particle mass concentrations for different MOUDI cut sizes during March and November 2001 and March 2002. Primary and Secondary Organic Carbon Fig. 6 shows spatial distribution of the average PM10 concentration for a) OCT and b) EC in March 1997. Both OCT and EC concentrations were highest at XAL and lowest at CES. Table 6 summarizes the linear regression parameters for OCT vs. EC plots at XAL, MER and CES sites for correlation coefficients above 0.7. In most cases, the highest correlations were observed during the 00:00 to 12:00 period. As expected, slope and intercept coefficients are time- and space- dependent. The lowest (OCT/EC)P ratio was obtained at XAL, mainly due to high soot emissions associated to heavy industrial activities and heavily travelled roads by diesel-powered vehicles in this area. The lowest ratios at MER and CES were 1.2 and 1.3 respectively, as these sites are characterized by mixed residential and commercial activities, as well as traffic of heavy and light duty vehicles. The variability in the intercept values did not allow an accurate estimation of this parameter, however values ranged from 0.5 to 5.6 μg/m3 (5–55% of OCT). In the Los Angeles emission inventory for OCT and EC it is reported that particle emissions from gasoline and LPG combustion are dominated by OC, while diesel particles by EC (Gray and Cass, 1998). This is consistent with the (OCT/EC)P values found for Mexico City as lower ratios were obtained for XAL compared with MER and CES sites.
The (OC/EC)P ratios in this study ranged from 0.7 to 1.9 (see Table 6), while ratios from Los Angeles in 1986 varied from 1.4 to 2.4, where primary organic aerosol is expected to dominate (Turpin et al., 1991). Recently, Yu et al. (2009) applied similar methodology to determine OCP and OCS at two sites downwind Mexico City, reporting ratios of 0.6 and 2.26 for suburban and rural sites, respectively. The (OC/EC)P ratios found in this study fall typically in the range of fossil fuel sources which agree with those reported by Yu et al. (2009). Fig. 7 shows the time series of OCT and OCS in PM2.5 measured during March, 1997. OCS was observed in all sampling periods and sites; at MER OCS reached maximum values of 9.2 μg/m3. XAL and CES showed a scattered secondary organic production for all sampling periods, except during the first week of March 1997 at CES, when ozone reached its highest value (not shown) and OCS formation was present the whole week. The analysis suggested that OCS concentration was due to day time production rather than previous day accumulation, except from March 3 to 5 and March 14 to 15 at MER. Most particles emitted and produced inside the basin were swept away by the wind flow the same day, as suggested by Fast and Zhong (1998). In average, OCS accounts from 2.3 to 3.4 μg/m3 with an average of 25% of the OCT.
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
202
a) Organic Carbon
b) Elemental Carbon
Fig. 6. Average PM10 concentration for a) organic carbon and b) elemental carbon during March, 1997. Table 6. Regression analysis parameters for OCT vs. EC at Xalostoc (XAL), La Merced (MER) and Cerro de la Estrella (CES). b: slope, a: intercept, r: correlation coefficient, n: number of data. Sampling period
b
00:00–06:00 06:00–12:00 00:00–12:00
0.9 ± 0.1 0.7 ± 0.1 0.8 ± 0.07
00:00–06:00 06:00–12:00 00:00–12:00
1.5 ± 0.2 1.2 ± 0.2 1.2 ± 0.1
00:00–06:00 06:00–12:00 12:00–18:00 00:00–18:00
1.9 ± 0.4 1.4 ± 0.2 1.5 ± 0.3 1.3 ± 0.1
a XAL 3.4 ± 1 5.6 ± 2 4.1 ± 1 MER 1.1 ± 0.9 2.6 ± 2.0 2.4 ± 0.7 CES 0.5 ± 1 1.4 ± 2 3.4 ± 1 3.1 ± 0.7
r
n
0.86 0.80 0.88
18 17 35
0.88 0.86 0.92
18 17 35
0.78 0.85 0.81 0.84
18 17 16 51
Fig. 7. Time series of OCT and OCS in PM2.5 measured during March 1997 for Xalostoc, La Merced and Cerro de la Estrella.
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
Calculated Emission Inventory of PM10, PM2.5 and Ammonia Table 7 shows the calculated results for February 1997 in MCMA of the PM10 and PM2.5 emission inventories, being unpaved, paved, and soil the major ones followed by mobile sources, such as diesel vehicles (>3 tons); gasoline vehicles and diesel trailer trucks. The industrial sources, basically mineral non-metal, showed small contribution followed by food-consumption industry, chemical and clothing industry (not shown). The most important difference between calculated results in this work and those from the official 1998 emission inventory (CAM, 2001) was the inclusion of particles from paved and unpaved roads, although the results were highly uncertain. The result calculated in the EI for mobile sources were 11.0% in contrast to 45.0% calculated with CMB model reported in this work (Table 5). Such difference could be due to the fact that the emission factor used in the EI was not determined experimentally and may not represent the Mexico City conditions, therefore to improve the EI estimation this should be taken into account. Table 8 shows the calculated results for February 1997 in Table 7. PM10 and PM2.5 calculated emission inventories for February 1997 in MCMA. PM2.5 PM10 Tons/month % PM2.5 Ton/month % PM10 Unpaved 1,217.0 49.27 8,328.0 75.46 Paved 419.7 16.99 1,217.0 11.03 Soil 266.2 10.77 665.4 6.03 Diesel vehicles >3 ton 91.9 3.72 120.6 1.09 Cars 66.0 2.67 107.2 0.97 Diesel trucks 71.4 2.89 93.7 0.85 Forest fires 54.1 2.19 58.8 0.53 Mineral non-metal 26.8 1.08 57.1 0.52 Diesel Buses 42.1 1.71 55.3 0.50 Food industry 27.4 1.11 42.9 0.39 Other sources 187.5 7.57 290.8 2.63 Total 2,470.1 99.97 11,036.8 100.0
No Emission Source 1 2 3 4 5 6 7 8 9 10 11
Table 8. Ammonia calculated emission inventory for February 1997 in MCMA. No 1 2 3 4 5 6 7 8 9 10
Emission Source Dairy cattle Dogs Sewage Controlled gasoline vehicles Associated animals (rats) Cattle Cats Human transpiration Non-controlled gasoline vehicles Human respiration Other sources Total
Ton/month 151.2 80.4 80.2 59.1 45.4 22.2 19.3 18.5
% Total 30.59 16.26 16.22 11.97 9.19 4.50 3.91 3.74
12.3
2.48
4.0 1.6 494.2
0.80 0.34 100
203
MCMA of the major sources of ammonia. The first eight sources account for 96% of the total emissions. Only two were related to mobile sources combustion, while all the others were associated to biological processes. The monthly ammonia emission to the atmosphere of MCMA was estimated to be nearly 500 tons. Meteorology The synoptic conditions during March 1997 were characterized by weak high-pressure systems with light to moderate winds and near-cloudless skies, these conditions were similar to those reported for the 2006 MILAGRO campaign (Fast and Zhong, 1998; Whiteman et al., 2000; Fast et al., 2007) as expected since both campaigns were at the end of winter and early spring. Weak high-pressure systems and low mixing height (MH), less than 1,000 m were associated to high concentrations of particles, being the highest concentration at NET, XAL and CES. In addition, surface wind speed above 8 m/s was related to high concentrations of PM10. Morning conditions at the southern site (UNAM) were characterized by a stable layer, in its absence the convective boundary layer (CBL) was the deepest of the region. The stable layer is a consequence of the down-slope cold winds that were advected from the mountain range; when these winds were absent the CBL rapidly raised. The greatest CBL growth occurred near mid-day at a rate of 650 m/h. The data suggests that 50% of the warming for the MCMA occurred between 8:00 and 11:00 h, while 40% of the heat in the region was dissipated between 16:00 and 19:00 h. When thermal forcing dominated, the mixing height (MH) rose to 4,600 m AGL, but when mechanical forcing dominated, the MH was close to 2,500 m AGL. The MH depends also on the location and the surface characteristics. In the early morning, at the UNAM site, a stable layer close to the surface was observed in 51% of the cases, due to cooler air advected by down-slope winds. This stable layer, however, was rapidly eroded. As the day progresses, the MH was higher at CUA and TEO sites. An urban effect on MH was observed at the UNAM site, where MH of about 1,000 m AGL was observed after 19:30 h (after sunset). Slightly lower results than those obtained in this paper have been reported (Doran et al., 1998; Whiteman et al., 2000; De Foy, 2006). The mean transport wind speed in the city was below 5 m/s and exhibited variations according with location and synoptic weather conditions. The strongest transport wind speed occurred at CUA and TEO, and the lightest was observed at UNAM, due to mountain blocking effect. A correlation analysis was applied to the spatial average concentrations of PM2.5, calculated from measurements at XAL, MER, CES TLA, NET and PED and the average transport wind speed calculated from observations at CUA, TEO, CHA and UNAM. The results indicated an inverse relationship between transport wind speed and PM2.5 concentrations as shown in Fig. 8. In contrast, no relationship was found with PM10. The ventilation index (VI) of the MCMA increased from 2,500 (bad ventilation) in the morning, to 12,500–15,000 (good ventilation conditions) at 15:00 hr. VI was 30–40% higher in the N and NE of the basin (CUA and TEO), while
204
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
the lowest VI occurred in the SW (UNAM) as a result of the nearby mountain range. Authors are not aware of reports on experimentally determined VI for the MCMA, however, the aforementioned good ventilation conditions indirectly agree with venting timescales modeled results of less than 24 h found by Fast and Zhong (1998) and de Foy et al. (2006). The wind circulation patterns were modeled with RAMS for March 4 and 5 because the highest particle levels were observed on these days. Results showed three principal
patterns: down-slope winds in the E and SW of the basin which shifted up-slope between 08:00 and 17:00 hrs, a N-NE persistent flow through the day with maximum velocities in the afternoon, and an easterly wind current flow developed between 14:00 h and 16:00 h with maximum velocities at 18:00 h (Fig. 9). The high particle concentrations on March 4 could be the result of the conjunction of light winds and high solar radiation that allowed a progressive particle accumulation and secondary aerosol production. On March 5,
Fig. 8. Average transport wind speed in the mixing height measured at Cuautitlan (CUA), Teotihuacan (TEO), Chalco (CHA) and Universidad Nacional Autónoma de México (UNAM) and average PM2.5 concentration at Xalostoc (XAL), La Merced (MER), Cerro de la Estrella (CES), Tlalnepantla (TLA), Netzahualcóyotl (NET) and Pedregal (PED).
Fig. 9. Wind circulation patterns for March 4 and 5, 1997 obtained with the RAMS model.
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
205
Fig. 9. (continued). a cold northern air mass generated intense winds and dry conditions, re-suspended particles and produced high particle concentrations which eventually dissipated. Dispersion Modeling The MCMA basin periodically experiences wind-blown dust events that increase PM10 concentrations above the Mexican ambient air quality standard (120 µg/m3, 24-h average). Wind-blown dust sources include agricultural lands that surround the periphery of the MCMA to the north and northeast sectors of the basin. The climatic conditions which favor dust events in the MCMA usually take place in March (MASNS, 1999) when the land becomes fallow during the dry season and wind-blown dust entrains as a result of moderate or high winds, high temperature and low humidity. The frequency of wind blown dust events used was compiled from a ten-year climate database spanning from 1988 to 1998. In addition, the analysis of the IMADAAVER database of five crustal species (Al, Si, Fe, Mg, and Ca) showed that geological material was the major contributor to PM10 (Edgerton et al., 1999). This section compares observed and modeled PM sources at regional scale in the MCMA which were used to develop an empirical method to estimate dust emissions for this region. The IMADA database provided the observations for model verification in efforts to study relationships between PM10 and agricultural field wind erosion (Villaseñor et al., 2003). Fig. 10 shows the surface wind vectors, windblown dust concentration contours and mixed layer depth for March 5 at 08:00 h as simulated by CALMET and CALPUFF models. Computations at this hour of the day were similar in shape and intensity to the earlier hours of simulation. At 08:00 h light-to-moderate easterly winds blow over the two major dust sources located northeast of the MCMA. Over the eastern mountain range, cool air descended with an easterly component, leading to air streams that cause minor convergence at the center of the MCMA. Two scenarios are observed in this figure in regard to the soil sources. Winds with a northerly component advect dust plumes into the MCMA basin where down-slope flows from the western mountains then advected the dust plumes eastward causing higher concentrations along the slopes of the eastern mountains. Thus, lower PM concentrations occurred at the
PED and MER sites (located in the southwest of the basin), as predicted by CALPUFF. The mixing layer height remained relatively shallow over the domain of interest for most of the morning period, but PM concentrations tended to remained relatively high near the erosion sources. The western portion of the MCMA remained unaffected by windblown dust partly due to westerly winds that counteract the effect of the spreading dust plume. The predictions at 14:00 h represented a transitional period characterized by low concentration of particulate matter measured at all the monitoring locations within the MCMA from 10:00 to 15:00 h (not shown). The height of the mixed layer increased significantly from 500 m to 2,000 m in less than 5 hours. In addition, the wind field displayed high transport wind speeds (about 7m/s) relative to the early hours with a northeasterly component throughout much of the MCMA, which was consistent with lower PM concentrations particularly at the industrial and urban areas. A similar situation occurred at 19:00 h (not shown), where the winds had a well-defined northeasterly component all over the physical domain and were responsible for transporting measurable amounts of geological dust to the densely populated areas of the MCMA. This constituted the second concentration peak observed on March 5. The diurnal pattern of predicted PM concentrations of geological origin for March 6 (not shown) presented a similar spatial behavior as that found for March 5, but with significantly lower values (Villaseñor et al., 2003). It should be noted that a strong northerly basin-to-valley wind was driven by a cold and dry air mass as evidenced from a temperature drop to the north of the MCMA in the evening of March 5. This cold air mass pushed its way through the basin and helped to maintain the strong northerly wind blowing thorough most of the time on March 6, resulting in a cleaner airshed and a better visibility. Fig. 11 shows the relationship between predicted (Cp) and observed (Co) windblown dust concentrations for March 5 and 6 at XAL, NET and MER sites. The model showed a better correlation between Cp and Co compared to March 6. This suggests that much of the suspended PM was associated to agricultural fields located on the northeast sector of the MCMA. In contrast, the poor association between Cp and Co for March 6 suggests that the PM concentrations measured at the receptor sites had a more significant
206
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
Fig. 10. Surface winds and dust concentration contours (a), and mixed layer topography (b) in MCMA during March 5, 1997 at 08:00 hours.
Fig. 11. Scatterplots of predicted (Cp) vs. observed (Co) windblown dust concentrations (µg/m3) for March 5 (a) and March 6 (b), 1997 at Xalostoc, Netzahualcoyotl and La Merced. contribution from local geological sources than from the drier dust of agricultural areas. The disagreement between observations and predictions for March 6 were likely due to omissions of other potential PM sources such as paved and unpaved roads that were not accounted for in the emission inventory. In summary, the simulation of the spatial and temporal evolution of PM concentrations predicted with the CALMET/CALPUFF modeling system showed reasonable agreement with the observed particle measurements of geological origin for March 5 1997. In contrast, the lower correlation between model predictions and observations for March 6 suggests that the PM concentrations measured at the MER and PED sites were from other dust sources within the vicinity of the receptors. Aerosol Optical Properties and Visibility
A variety of diurnal aerosol visible light absorption and scattering patterns was obtained. During the 1997 period of study, most of the days had their prominent absorption peaks early in the morning between 7:00 h and 9:30 h and their prominent scattering peaks later between 9:30 h and 11:00 h. The earlier absorption peaks were attributed to the elevated elemental carbon vehicular emissions during the heavy traffic hours whereas the later scattering peaks were attributed to secondary aerosols formed photochemically in the atmosphere (Eidels-Dubovoi, 2002). The time of the day at which the aerosol absorption and scattering prominent peaks appeared was around that recently reported for urban and suburban sites in Mexico City (Marley et al., 2009; Paredes-Miranda et al., 2009). During the 2002 period of study, the days with prominent bap and bsp peaks appearing in the morning between 7:00 h and 10:30 h prevailed over the days with
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
those peaks occurring in the afternoon between 13:00 h and 16:00 h. Since particle size distribution determines the pattern structure (Eidels-Dubovoi, 2000) that is, the time of the day and frequency at which maximum and minimum values appear, and the same size distribution was used in the calculations of the bap and bsp patterns, both calculated patterns showed similar structures and thus could not reveal the nature of the particles. However, the presence of prominent early morning bap and bsp peaks in the majority of the days of the 2002 studied period suggests that, particles emitted by vehicles prevailed over those formed photochemically after sunrise. The best fit of the theoretical to the observed aerosol visible light absorption patterns was obtained when the values of the imaginary part of the refractive index derived from an original method (Eidels-Dubovoi, 1993) were used in the Mie calculations instead of the values for EC reported in the literature (Horvath, 1993). In Fig. 12, the diurnal visible light absorption patterns calculated with Mie theory and those derived from aethalometer measurements for March 15, 2002 are compared. The fact that bap was slightly underestimated in the first hours of this day and overestimated in the remaining hours could mean that the nature of the particles that absorb visible light changed during the day. The gas contribution to total light extinction relative to the particle contribution was negligible during both studied periods (Table 9). It can also be seen that during the 1997
207
period beg/bext was lower and bep/bext was slightly higher at MER compared to PED. This suggests that particle pollution of the commercial site (MER) did not exceed by far that of the residential site (PED). The MER results given in Table 9 for both studied periods indicate that the gaseous contribution to total visible light extinction diminished and the particle contribution increased slightly on the 2002 period. However, care should be taken in interpreting these small changes as a small increment of particulate matter pollution in 2002, since different methods were used in deriving the particle coefficients for each period. Visibilities averaged over each of the entire studied periods, are given in Table 9. According to the international visibility code (Hulburt, 1941), during the 1997 studied period, the value of visibility obtained for PED corresponded to very clear conditions whereas that obtained for MER corresponded to clear conditions. However, considering the small difference of 6.3 km between PED and MER averaged visibilities over the 1997 period, it may be concluded that the impact of aerosol on visibility was similar at both, the residential and the commercial sites of the MCMA. The MER averaged visibility over the 2002 period was slightly above the lower limit of the interval (4–10 km) classified as light haze in the international visibility code (Hulburt, 1941). The apparent decrease of the visibility during this period compared to the visibility of the 1997 period is not conclusive, since different methods were used in obtaining each result.
Fig. 12. Diurnal visible light absorption pattern obtained from Mie theory and derived from aethalometer measurements for March 15, 2002 at MER. Table 9. Contribution of gases (beg/bext) and particles (bep/bext) to total visible light extinction (bext) and visibility (Vis). Site La Merced Pedregal
Period 28 Feb.–10 March 1997 11–25 March 2002 28 Feb.–10 March 1997
beg/bext 0.057 0.022 0.125
bep/bext 0.943 0.978 0.875
Vis (km) 14.8 4.3 21.1
208
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
CONCLUSIONS Particulate matter collected at different sites, representing urban and rural environments within the MCMA during 1997 field campaign, were used to support modeling for determining the contribution of PM2.5 sources to the atmosphere, the pollutant transport and the estimation of secondary organic carbon. Calculated emission inventories of particles and ammonia were used as inputs to simulate air quality. Additional campaigns were concentrated on optical properties of fine particles to determine their contribution to total light extinction and impact on visibility as well as to obtain the particle mass size distribution. Highest PM2.5 and PM10 concentrations occurred in the NE sector, where industrial and traffic activities are prevalent, and decreased to the SW. The chemical composition of PM2.5 was dominated by carbon, sulfate, nitrate, ammonium and crustal species. Both organic and elemental carbon concentrations were highest in the NE sector, and decreased radically. Receptor model results showed significant spatial variations in PM2.5 source contributions depending on local soil types and land use. NET showed the highest dust contribution, over 38%, while at XAL, MER and CES the mobile sources contributed with an average of 57%, and NET, PED and TLA with 33%. Ammonium sulfate showed the highest percentage at PED (22%). The incineration source, enriched by Zn, Pb and Cl-, varied from 11–2%. Similar contributions were found for cooking and biomass burning (13–5%). Modeling of particle dispersion showed that winds with a northerly component advect dust plumes of crustal and agricultural origin. The high particle concentration on March 5, 1997 showed good correlation between observed and predicted results. It should be mention that more work is still needed considering field measurements of soil dust and land use parameters to improve these results. The analysis of meteorology indicated that surface wind speed above 8 m/s and weak high-pressure systems and low mixing height were associated to high concentrations of PM10 and the latter also to high concentrations of PM2.5. The mean transport wind speed in the MCMA was less than 5 m/s, with the lightest winds in the SW region, and the strongest in the N and NE. An inverse relationship between transport wind speed and average PM2.5 concentrations was observed. The ventilation index was highest in the afternoon period, indicating better pollutant dispersion. From OC and EC data analysis, a 25% average of total organic matter in the PM2.5 fraction was associated to secondary organic aerosol. The calculated EI for ammonia showed that dairy and non-dairy cattle were the dominant sources followed by biological sources and gasoline vehicles. PM2.5 mobile sources derived from the EI were 11% whereas those estimated with the receptor model were 45% The highest particle mass concentrations occurred within the accumulation mode associated mainly to primary emissions. Good correlations were observed between PM2.5
and particles with diameters higher than 0.32 µm. A seasonal dependence of particle mass concentration was observed for all particles sizes, with lower concentrations in November and higher in March. The earlier visible light absorption peaks that appeared in the diurnal patterns were attributed to the elevated elemental carbon vehicular emissions during the heavy traffic hours while the light scattering peaks that occurred near midday were attributed to secondary aerosol formed photochemically in the atmosphere. The presence of prominent early morning aerosol light absorption and scattering peaks in the majority of the days of the 2002 studied period suggests that, during this period, particles emitted by vehicles prevailed over those formed photochemically after sunrise. The contribution of particles to total light extinction was slightly higher at MER compared to PED and the impact of aerosol on visibility was similar at both, the commercial and the residential sites, during the 1997 studied period. ACKNOWLEDGEMENTS The authors gratefully acknowledge the support of PEMEX and DOE’s Office of Biological and Environmental Research and the International institutions for providing the data collected in the 1997 campaign. REFERENCES Adams, K.M., Davis, L.I. Jr., Japar, S.M., Finley, D.R. and Cary, R.A. (1990). Measurement of Atmospheric Elemental Carbon: Real-Time Data for Los Angeles during Summer 1987. Atmos. Environ. 24A: 597–604. Adams, K.M., Davis, L.I. Jr., Japar, S.M. and Finley, D.R. (1990a). Real-time, in Situ Measurements of Atmospheric Optical Absorption in the Visible via Photoacoustic spectroscopy—IV. Visibility degradation and aerosol optical properties in Los Angeles. Atmos. Environ. 24A: 605–610. Aiken, A.C., DeCarlo, P.F., Kroll, J.H., Worsnop, D.R. Huffman, J.A., Docherty, K.S., Ulbrich, I.M., Mohr, C., Kimmel, J.R., Sueper, D., Sun, Y., Zhang, Q., Trimborn, A., Northway, M., Ziemann, P.J., Canagaratna, M.R., Onasch, T.B., Alfarra, M.R., Prevot, A.S.H., Dommen, J., Duplissy, J., Metzger, A., Baltensperger, U. and Jimenez, J.L. (2008). O/C and OM/OC Ratios of Primary, Secondary, and Ambient Organic Aerosols with High-resolution Time-of-flight Aerosol Mass Spectrometry. Environ. Sci. Technol. 42: 4478–4485. Aiken, A.C., Salcedo, D., Cubison, M.J., Huffman, J.A., DeCarlo, P.F., Ulbrich, I.M., Docherty, K.S., Sueper, D., Kimmel, J.R., Worsnop, D.R., Trimborn, A., Northway, M., Stone, E.A., Schauer, J.J., Volkamer, R., Fortner, E., de Foy, B., Wang, J., Laskin, A., Shutthanandan, V., Zheng, J., Zhang, R., Gaffney, J., Marley, N.A., ParedesMiranda, G., Arnott, W.P., Molina, L.T., Sosa, G. and Jimenez, J.L. (2009). Mexico City Aerosol Analysis during MILAGRO Using High Resolution Aerosol Mass Spectrometry at the Urban Supersite (T0) – Part 1: Fine
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
Particle Composition and Organic Source Apportionment. Atmos. Chem. Phys. Discuss. 9: 8377–8427. Aldape, F., Flores, M.J., Flores, A.J., Retama-Hernández, A. and Rivera-Hernández, O. (2005). Elemental Composition and Source Identification of PM2.5 Particles Collected in Downtown Mexico City. Int. J. PIXE 15: 263–270. ASTER Division. (1994). RAMS, The Regional Atmospheric Modeling System. Technical Description (Draft). ASTER Division, Mission Research Corporation, Fort Collins, CO. ASTER Division. (1999). RAMS, User Manual. Cabada, J.C., Rees, S., Takahama, S., Khlystov, A., Pandis, S.N., Davidson, C.I. and Robinson, A.L. (2004). Mass Size Distributions and Size Resolved Chemical Composition of Fine Particulate Matter at the Pittsburgh Supersite. Atmos. Environ. 38: 3127–3141. CAM (Comisión Ambiental Metropolitana). (2001). Inventario de emisiones a la Atmósfera de la Zona Metropolitana de la Ciudad de México 1998. Mexico, D.F. Castro, L.M., Pio, C.A., Harrison, R.M. and Smith, D.J.T. (1999). Carbonaceous Aerosol in Urban and Rural European Atmospheres: Estimation of Secondary Organic Carbon Concentrations. Atmos. Environ. 33: 2771–2781. Chow, J.C. (1995). Measurement Methods to Determine Compliance with Ambient Air Quality Standards for Suspended Particles. J. Air Waste Manage. Assoc. 45: 320–382. Chow, J.C., Watson, J.G., Edgerton, S.A. and Vega, E. (2002). Chemical Composition of PM10 and PM2.5 in Mexico City during Winter 1997. Sci. Total Environ. 287: 177–201. Chow, J.C., Watson, J.G., Edgerton, S.A., Vega, E. and Ortíz, E. (2002). Spatial Differences in Outdoor PM10 Mass and Aerosol Composition in Mexico City. J. Air Waste Manage. Assoc. 52: 423–434. Chu, S.H. (2005). Stable Estimate of Primary OC/EC Ratios in the EC Tracer Method. Atmos. Environ. 39: 1383–1392. DDF, GEM, SEMARNAT, SSA (Departamento del Distrito Federal, Gobierno del Estado de México, Secretaría del Medio Ambiente y Recursos Naturales, Secretaría de Salubridad y Asistencia). (1996). Programa para Mejorar la Calidad del Aire en el Valle de México 1995–2000. Mexico, D.F. de Foy, B., Varela, J.R., Molina, L.T. and Molina, M.J. (2006). Rapid Ventilation of the Mexico City Basin and Regional Fate of the Urban Plume. Atmos. Chem. Phys. 6: 2321–2335. DeCarlo, P.F., Dunlea, E.J., Kimmel, J.R., Aiken, A.C., Sueper, D., Crounse, J., Wennberg, P.O., Emmons, L., Shinozuka, Y., Clarke, A., Zhou, J., Tomlinson, J., Collins, D.R., Knapp, D., Weinheimer, A.J., Montzka, D.D., Campos, T. and Jimenez, J.L. (2008). Fast Airborne Aerosol Size and Chemistry Measurements above Mexico City and Central Mexico during the MILAGRO Campaign. Atmos. Chem. Phys. 8: 4027–4048. Doran, J.C., Bian, X., Edgerton, S.A., Fast, J.D., Hubbe, J.M., Shaw, W.J. and Whiteman, C.D. (1998).The IMADA-AVER Boundary-Layer Experiment in the
209
Mexico City Area. Bull. Am. Meteorol. Soc. 79: 2497– 2508. Edgerton, S.A., Arriaga, J.L., Archuleta, J., Bian, X., Bossert, J.E., Chow, J.C., Coulter, R.L., Doran, J.C., Doskey, P.V., Elliot, S., Fast, J.D., Gaffney, J.S., Guzman, F., Hubbe, J.M., Lee, J.T., Malone, E.L., Marley, N.A., McNair, L.A., Neff, W., E.Ortiz, R. Petty, M., Ruiz, W.J. Shaw, Sosa, G., Vega, E., Watson, J.G., Whiteman, C.D. and Zhong, S. (1999). Particulate Air Pollution in Mexico City: A Collaborative Research Project. J. Air Waste Manage. Assoc. 49: 1221–1229. Eidels-Dubovoi, S. (2002). Aerosol Impacts on Visible Light Extinction in the Atmosphere of Mexico City. Sci. Total Environ. 287: 213–220. Eidels-Dubovoi, S. (Ed.) International Center for Heat and Mass Transfer, UNAM, Instituto de Ingeniería,. 1993. International Symposium on Heat and Mass Transfer in Energy Systems and Environmental Effects. Solar Radiation Attenuation by Atmospheric Aerosol Particles at Different Sites in the Mexico City Valley. p. 140–142. Eidels-Dubovoi, S. (Ed.) WIT PRESS Southampton, Boston 2000. Air Pollution VIII, Sec.12: Aerosols and Particles, Atmospheric Aerosol Contribution to Visible Light Absorption and Scattering in Mexico City and Simple Tools to Determine Refractive Indices and Size Distributions. p. 709–718. Ellsworth-Davis, B. (2001). GIS: A Visual Approach, OnWord Press, 2nd. Edition, New York. EPA (Environmental Protection Agency). (1995). Compilation of Air Pollutant Emission Factors. Vol. I: Stationary Point and Area Sources. AP-42. USA. EPA (Environmental Protection Agency). (1995a). Compilation of Air Pollutant Emission Factors. Vol. II: Mobile Sources. AP-42. USA. ESRI (Environmental Systems Research Institute). (1999). Getting to Know ArcView GIS, Environmental Systems Research Institute Press, USA. Fast, J.D. and Zhong, S. (1998). Meteorological Factors Associated with Inhomogeneous Ozone Concentrations within the Mexico City Basin. J. Geophys. Res. 103: 18927–18946. Fast, J.D., deFoy, B., AcevedoRosas, F., Caetano, E., Carmichael, G., Emmons, L., McKenna,D., Mena, M., Skamarock, W., Tie, X., Coulter, R.L., Barnard, J.C., Wiedinmyer, C. and Madronich, S. (2007). A Meteorological Overview of the MILAGRO Field Campaigns. (2007). Atmos. Chem. Phys. 7: 2233–2257. Forster, P., Ramaswamy, V., Artaxo, P., Berntsen, T., Betts, R., Fahey, D.W., Haywood, J., Lean, J., Lowe, D.C., Myhre, G., Nganga, J., Prinn, R., Raga, G., Schulz, M. and Van Dorland, R. In Solomon, S., Qin, D., Manning, M., Marquis, M. Averyt, K., Tignor, M.M.B., Miller, Jr., H.L. and Chen, Z.L. (Eds.) Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change 2007, Cambridge University. GDF (Gobierno del Distrito Federal) (2001). Inventario de emisiones de la Zona Metropolitana del Valle de México 1998. Gobierno del Distrito Federal, Gobierno del Estado de México, SEMARNAT.
210
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
Gray, H.A. and Cass, G.R. (1998). Source Contributions to Atmospheric Fine Carbon Particle Concentrations. Atmos. Environ. 32: 3805–3825. Hagen, L.J., Wagner, L.E. and Tatarko, J. (1996). Wind Erosion Pediction System (WEPS). Technical Documentation, BETA Release 95-08. USDA-ARS, NPA. Wind Erosion Research Unit, Kansas State University, Manhattan, KS. Hennigan, C.J., Sullivan, A.P., Fountoukis, C.I., Nenes, A., Hecobian, A., Vargas, O., Peltier, R.E., Case Hanks, A.T., Huey, L.G., Lefer, B.L., Russell, A.G. and Weber, R.J. (2008). On the Volatility and Production Mechanisms of Newly Formed Nitrate and Water Soluble Organic Aerosol in Mexico City. Atmos. Chem. Phys. 8: 3761– 3768. Hinds, W.C. (1998). Aerosol Technology: Properties, Behavior and Measurement of Airborne Particles. 2nd. Edition. Wiley, New York. Horvath, H. (1993). Atmospheric Light Absorption—A Review. Atmos. Environ. 27A: 293–317. Hulburt, E.O. (1941). Optics of Atmospheric Haze. J. Opt. Soc. Am. 31: 467–476. Japar, S.M., Brachaczek, W.W., Gorse, Jr. R.A., Norbeck, J.M. and Pierson, W.R. (1986). The Contribution of Elemental Carbon to the Optical Properties of Rural Atmospheric Aerosols. Atmos. Environ. 20: 1281–1289. Johnson, K.S., Zuberi, B., Molina, L.T., Molina, M.J., Iedema, M.J., Cowin, J.P., Gaspar, D.J., Wang, C. and Laskin, A. (2005). Processing of Soot in an Urban Environment: Case Study from the Mexico City Metropolitan Area. Atmos. Chem. Phys. 5: 3033–3043. Kleinman, L.I., Springston, S.R., Daum, P.H., Lee, Y.N., Nunnermacker, L.J., Senum, G.I., Wang, J., WeinsteinLloyd, J., Alexander, M.L., Hubbe, J., Ortega, J., Canagaratna, M.R. and Jayne, J. (2008). The Time Evolution of Aerosol Composition over the Mexico City Plateau. Atmos. Chem. Phys. 8: 1559–1575. Marley, N.A., Gaffney, J.S., Castro, T., Salcido, A. and Frederick, J. (2009). Measurements of Aerosol Absorption and Scattering in the Mexico City Metropolitan Area during the MILAGRO field Campaign: A Comparison of Results from the T0 and T1 Sites. Atmos. Chem. Phys. 9: 189–206. MASNS (Mexican Air Space Navigation Service). (1999). http://www.sct.gob.mx. Miranda, J., Andrade, E., López-Suárez, A., Ledesma, R., Cahill, T.A. and Wakabayashi, P.H. (1996). A Receptor Model for Atmospheric Aerosols from a Southwestern Site in Mexico City. Atmos. Environ. 30: 3471–3479. Moffet, R.C., de Foy, B., Molina, L.T., Molina, M.J. and Prather, K.A. (2008). Measurement of Ambient Aerosols in Northern Mexico City by Single Particle Mass Spectrometry. Atmos. Chem. Phys. 8: 4499–4516. Moffet, R.C., Desyaterik, Y., Hopkins, R.J., Tivanski, A.V., Gilles, M.K., Wang, Y., Shutthanandan, V., Molina, L.T., Abraham, R.G., Johnson, K.S., Mugica, V., Molina, M.J., Laskin, A. and Prather, K.A. (2008b). Characterization of Aerosols Containing Zn, Pb, and Cl from an Industrial Region of Mexico City. Environ. Sci. Technol. 42: 7091–7097.
Osornio-Vargas, Á.R., Bonner, J.C., Alfaro-Moreno, E., Martínez, L., García-Cuellar, C., Ponce-de-León Rosales, S., Miranda, J. and Rosas, I. (2003). Proinflammatory and Cytotoxic Effects of Mexico City Air Pollution Particulate Matter in Vitro Are Dependent on Particle Size and Composition. Environ. Health. Perspect. 111: 1289–1293. Pandis, S. (2003). Atmospheric Aerosol Processes: Particulate Matter Science for Policy Makers. A NARSTO Assessment Part 1, EPRI 1007735, February, 2003. Paredes-Miranda, G., Arnott, W.P., Jimenez, J.L., Aiken, A.C., Gaffney, J.S. and Marley, N.A. (2009). Primary and Secondary Contributions to Aerosol Light Scattering and Absorption in Mexico City during the MILAGRO 2006 Campaign. Atmos. Chem. Phys. 9: 3721–3730. Pielke, R.A., Cotton, W.R., Walko, R.L., Tremback, C.J., Lyons, W.A., Grasso, L.D., Nicholls, M.E., Moran, M.D., Wesly, D.A., Lee, T.J. and Copeland, J.H. (1992). A Comprehensive Meteorological Modeling System— RAMS. Met. Atmos. Phys. 49: 69–91. Querol, X., Pey, J., Minguillón, M.C., Pérez, N., Alastuey, A., Viana, M., Moreno, T., Bernabé, R.M., Blanco, S., Cárdenas, B., Vega, E., Sosa, G., Escalona, S., Ruiz, H. and Artíñano, B. (2008). PM Speciation and Sources in Mexico during the MILAGRO-2006 Campaign Atmos. Chem. Phys. 8: 111–128. Rosen, H., Hansen, A.D.A., Gundel, L. and Novakov, T. (1978). Identification of the Optically Absorbing Component in Urban Aerosols. Appl. Opt. 17: 3859–3861. Schauer, J.J. (2003). Evaluation of Elemental Carbon as a Marker for Diesel Particulate Matter. J. Exposure Sci. Environ. Epidemiol. 13: 443–453. Scire, J.S., Robe, F.R., Fernau, M.E., Insley, E.M. and Yamartino, R.J. (1997). A User's Guide for the CALMET Meteorological Model (Version 5.0). Earth Tech. Inc., Concord, MA. Scire, J.S., Strimaitis, D.G. and Yamartino, R.J. (2000). A User's Guide for the CALPUFF Dispersion Model (Version 5.0). Earth Tech. Inc., Concord, MA. Seibert, P., Beyrich, F., Gryning, S.E., Joffre, S., Rasmussen, A. and Tercier, P. (2000). Review and Intercomparison of Operational Methods for the Determination of the Mixing Height. Atmos. Environ. 34: 1001–1027. Seinfeld, J.H. and Pandis, S.N. (1998). Atmospheric Chemistry and Physics. Wiley, New York. SMA (Secretaría del Medio Ambiente del Gobierno del Distrito Federal). (2007). La calidad del aire en la zona Metropolitana del Valle de México 1986–2006. Informe del estado y tendencias de la contaminación atmosférica. Mexico, DF. Stone, E.A., Zinder, D.C., Sheesley, R.J., Sullivan, A.P., Weber, R.J. and Schauer, J.J. (2008). Source Apportionment of Fine Organic Aerosol in Mexico City during the MILAGRO Experiment 2006. Atmos. Chem. Phys. 8: 1249–1259. Tennekes, H. (1974). The Atmospheric Boundary Layer. Phys. Today. 27: 52–63. Tremback, C.J. and Walko, R.L. (2000). The Regional Atmospheric Modeling System (RAMS): Development
Vega et al., Aerosol and Air Quality Research, 10: 193–211, 2010
for Parallel Processing Computer Architectures. ASTER Division, Mission Research Croporation, Fort Collins, CO. Turpin, B.J. and Huntzicker, J.J. (1995). Identification of Secondary Organic Aerosol Episodes and Quantitation of Primary and Secondary Organic Aerosol Concentrations during SCAQS. Atmos. Environ. 29: 3527–3544. Turpin, B.J., Huntzicker, J.J., Laron, S.M. and Cass, G.R. (1991). Los Angeles Summer Midday Particulate Carbon: Primary and Secondary Aerosol. Environ. Sci. Technol. 25: 1788–1793. Vega, E., García, I., Ruiz, M.E., Barbiaux, M. and Apam, D. (1997). Application of a Chemical Mass Balance (CMB) Receptor Model to Respirable Particulate Matter in Mexico City. J. Air Waste Manage. Assoc. 47: 524–529. Vega, E., Lowenthal, D., Ruiz, H., Reyes, E., Watson, J.G., Chow, J.C., Viana, M., Querol, X., Alastuey, A. (2009). Fine Particle Receptor Modeling in the Atmosphere of Mexico City. J. Air Waste Manage. Assoc. 59: 1417–1428. Vega, E., Mugica, V., Carmona, R. and Valencia, E. (2000). Hydrocarbon Source Apportionment in Mexico City Using the Chemical Mass Balance Receptor Model. Atmos. Environ. 34: 4121–4129. Vega, E., Reyes, E., Mora, V., Sánchez, G., Chow, J.C., Watson, J.G. and Ruiz, M.E. (2003). Análisis de partículas en el invierno de 1997 en la atmósfera de la ciudad de México. Revista del IMIQ. 44: 213–225. Vega, E., Reyes, E., Sanchez, G., Ortiz, E., Ruiz, M., Chow, J.C., Watson, J.G. and Edgerton, S. (2002). Basic statistics of PM2.5 and PM10 in the atmosphere of Mexico City. Sci. Total Environ. 287: 167–176. Villaseñor, R., Lopez-Villegas, T., Eidels-Dubovoi, S., Quintanar, A. and Gallardo, J.C. (2003). A Mesoscale
211
Modeling Study of Wind Blown Dust on the Mexico City Basin. Atmos. Environ. 37: 2451–2462. Volkamer, R., Jiménez, J.L., San Martini, F., Dzepina, K., Zhang, Q., Salcedo, D., Molina, L.T., Worsnop, D.R. and Molina, M.J. (2006). Secondary Organic Aerosol Formation from Anthropogenic Air Pollution: Rapid and Higher than Expected. Geophys. Res. Lett. 33: L17811, doi: 10.1029/2006GL026899. Watson, J.G., Chow, J.C., Lu, Z., Fujita, E.M., Lowenthal, D.H. and Lawson, D.R. (1994). Chemical Mass Balance Source Apportionment of PM10 during the Southern California Air Quality Study. Aerosol Sci. Technol. 21: 1–36. Whiteman, C.D., Zhong, S., Bian, X., Fast, J.D. and Doran, J.C. (2000). Boundary Layer Evolution and RegionalScale Diurnal Circulations over the and Mexican Plateau. J. Geophy. Res. 105: 10081–10102. Yasa, Z., Amer, N.M., Rosen, H., Hansen, A.D.A., Novakov, T. (1979). Photoacoustic Investigation of Urban Aerosol Particles. Applied Optics. 18: 2528–2530. Yokelson, R.J., Urbanski, S.P., Atlas, E.L., Toohey, D.W., Alvarado, E.C., Crounse, J.D., Wennberg, P.O., Fisher, M.E., Wold, C.E., Campos,T.L., Adachi, K., Buseck, P.R. and Hao, W.M. (2007). Emissions from Forest Fires near Mexico City. Atmos. Chem. Phys. 7: 5569–5584. Yu, X.Y., Cary, R.A. and Laulainen, N.S. (2009). Primary and Secondary Organic Carbon Downwind of Mexico City. Atmos. Chem. Phys. 9: 6793–6814. Received for review, June 10, 2009 Accepted, January 5, 2010
