Article

pubs.acs.org/est

Impact of Trash Burning on Air Quality in Mexico City

A. Hodzic,† C. Wiedinmyer,† D. Salcedo,‡,∥ and J. L. Jimenez§
†
National Center for Atmospheric Research, Boulder, Colorado, United States
‡
Centro de Investigaciones Químicas, Universidad Autónoma del Estado de Morelos, Cuernavaca, México
§
Department of Chemistry and Biochemistry, and CIRES, University of Colorado, Boulder, Colorado, United States
*
S Supporting Information

ABSTRACT: Air pollution experienced by expanding urban areas is responsible

for serious health effects and death for millions of people every year. Trash
burning is a common disposal method in poor areas, yet it is uncontrolled in
many countries, and its contribution to air pollution is unclear due to
uncertainties in its emissions. Here we develop a new trash burning emission
inventory for Mexico City based on inverse socioeconomic levels and recently
measured emission factors, and apply a chemistry-transport model to analyze the
effects on pollutant concentrations. Trash burning is estimated to emit 25 tons of
primary organic aerosols (POA) per day, which is comparable to fossil fuel POA
emissions in Mexico City, and causes an increase in average organic aerosol
concentrations of ∼0.3 μg m−3 downtown and up to 2 μg m−3 in highly
populated suburbs near the sources of emission. An evaluation using
submicrometer antimony suggests that our emission estimates are reasonable.
Mitigation of trash burning could reduce the levels of organic aerosols by 2−40% and those of PM2.5 by 1−15% over the
metropolitan area. The trash burning contributions to carbon monoxide, nitrogen oxides, and volatile organic compounds were
found to be very small (<3%), and consequently the contributions to secondary nitrate, sulfate, and secondary organic aerosols
are also very small.

1. INTRODUCTION Mexico City inhabitants have seven times higher concentrations

Air pollution experienced by expanding urban areas is of particles in their lungs than people from less polluted cities.
responsible for serious health effects and premature death for The 2006 MILAGRO field project (Megacity Initiative: Local
millions of people every year.1 Burning of trash is an important and Global Research Observations,7) provided detailed
source in many densely populated areas in developing chemical characterization of the pollution and sources in
countries, which can lead to substantial emissions of Mexico City and its outflow region, including a study of the
atmospheric pollutants including carbon monoxide (CO), emissions from cooking and trash burning.2 Reference 2
nitrogen oxides (NOx), aerosol particles, and volatile organic reported source-specific emission factors for a number of
compounds (VOC). According to Christian et al.2 about 2000 gaseous and particulate species. Fine particle antimony (Sb)
Tg of garbage are generated per year worldwide and about half was put forward as a possible tracer of garbage burning, due to
of that may be burned, making these emissions a potentially the high emission factor from this source. By assuming that all
major health concern. By using the emission factors reported by of the Sb present in Mexico City arises from this source, it was
ref 2 we can estimate the global trash burning emissions of estimated2 that trash burning contributes up to 28% of the fine
primary organic aerosols (POA) as more than 6 Tg per year particulate matter (PM2.5). Trash burning emissions are
which represents about 15% of POA global emissions from all currently not included in the Mexico City anthropogenic
biomass burning sources, and is comparable to the emissions emission inventory,8 and their impact has been ignored in the
from biofuel use.3 The atmospheric fate of these emissions is Mexico City regional modeling studies.
largely undocumented, which makes it difficult to assess their In this study, we provide for the first time an estimate of the
effects on air quality and human health. This assessment is impact of trash burning on air quality in a developing world
particularly needed in developing countries where trash burning megacity. An emission inventory is developed using the
is often uncontrolled and is a potential source of toxic emissions factors measured as part of MILAGRO and the
pollutants such as metals and dioxins4 within and near densely gridded demographic and socioeconomic population distribu-
populated areas.
The characterization of the contribution of different sources Received: November 5, 2011
to air quality was one of the objectives of recent field studies Revised: March 20, 2012
that took place in Mexico City. Air pollution causes serious Accepted: March 29, 2012
health effects for Mexico City inhabitants.5 According to ref 6 Published: March 29, 2012

© 2012 American Chemical Society 4950 dx.doi.org/10.1021/es203954r | Environ. Sci. Technol. 2012, 46, 4950−4957
Environmental Science & Technology Article

Figure 1. Spatial distribution of (a) population density, (b) socioeconomic classes according to Manuel Suárez (personal communication) and (c)
gridded trash burning organic aerosol emissions (kg km−2 day−1) in the Mexico City Metropolitan Area. (a) and (b) show Mexican state boundaries
while (c) shows both the state boundaries and altitude contours. The assumed diurnal profile for trash burning emissions is also shown (d).

tions. The emissions are then used in a 3D chemistry-transport expected to be small, based on low contribution of CO from
model, which has been previously validated for this region, with those sources to the MCMA valley.11
the goal of quantifying the contribution of this source to The spatial distribution of population and socioeconomic
pollutant levels in Mexico City and estimate the likely benefits classifications for Mexico City are shown in Figure 1. For each
of mitigating trash burning in this region. Our analysis has socioeconomic classification, daily per capita trash production
wider implications, given the frequent occurrence of trash
and the percentage of trash that is burned were assigned
burning in developing countries.
according to Table 1. The per capita trash production was
2. DATA AND METHODS taken from ref 12 for Low, Mid, and High socioeconomic
classes. For the Mid-Low and Mid-High classes, the average per
2.1. Creating the Mexico City Trash Burning
Inventory. Emissions from trash burning are not reported in capita trash productions of the Mid and Low, and the Mid and
the Mexico City or the national emissions inventory.8,9 These High classes, respectively, were averaged. There are no
fires are also very small and cannot be detected from satellites,
and therefore are not included in typical biomass burning Table 1. Assigned Per Capita Trash Production Based on
emission inventories (e.g., ref 10). For the purposes of this Ojeda-Benitez et al. (2008), and Assumed Fractions of Trash
study, the Mexico City trash burning emissions were estimated Burned (See Text)
by combining information on population density and socio- socioeconomic kg trash produced/person/ percentage
economic levels with measured emission factors for trash fires. ID class day burned
The inventory was built for the Mexico City Metropolitan Area 1 Low 0.89 60
(MCMA) including the Distrito Federal and the State of 2 Mid-Low 0.96 60
Mexico ranging from 18.95 to 19.99N, and from 99.48 to 3 Mid 1.04 30
99.63W with a spatial resolution of 1 km2. Emissions are 4 Mid-High 1.05 20a
calculated as the product of the amount of trash produced, the 5 High 1.06 20a
fraction burned, and the amount of a compound emitted per
a
unit mass of trash burned. The impact on MCMA air quality of This trash is assumed to be transported and burned in the Low and
trash burning in populated areas outside the MCMA is Mid-Low socioeconomic areas.

4951 dx.doi.org/10.1021/es203954r | Environ. Sci. Technol. 2012, 46, 4950−4957

Environmental Science & Technology Article

published data on the percentage of trash which is burned for diameter) will apply. The model concentrations are summed up
each socioeconomic class, but anecdotal evidence and to 600 nm in physical diameter for comparison with the
conversations with Mexican researchers indicate that this measurements.
fraction may exceed 50% in the poorest areas and is negligible 2.3. WRF/Chem Simulations. The effects of trash burning
in the areas of high socioeconomic level. In the latter areas we emissions on air quality in and around Mexico City are
assumed that a fraction of the trash collected was transported to investigated using the Weather Research and Forecasting model
landfills located in lower socioeconomic areas and burned there. coupled with chemistry (WRF/Chem21). The WRF/Chem
Therefore, we have distributed any assumed emissions from model is a community tool that provides several packages for
trash produced in Mid-High and High socioeconomic areas to gas-phase and aerosol transport and chemistry. In this study the
the Mid-Low and Low areas in equal proportions. The chemistry is simulated using the SAPRC99 gas-phase chemical
uncertainties associated with these estimates in Table 1 are mechanism,22 and the MOSAIC aerosol module.23 The aerosol
large, and the consistency of this data set with ambient data will module solves the chemistry and dynamics of internally mixed
be discussed later in the paper. inorganic and organic aerosol constituents for four size bins
The emission factors (mass of a compound emitted per mass ranging from 40 nm to 10 μm. Details on the model aerosol
of trash burned) for trash burning are taken from refs 2,13, and treatment and application over Mexico City can be found in
14 and are summarized in Supporting Information SI-Table 1. previous model applications for this region;24,25 however, this
Especially relevant emissions factors for this study include those particular study differs from those cited in the treatment of
for primary organic aerosols (6.9 g kg−1 of trash burned, organic aerosol species. POA is considered physically and
calculated from the reported OC emission factors assuming an chemically inert, similar to ref 24. Secondary organic aerosol
OM/OC = 1.3), CO (45 g kg−1 of trash burned), NOx (3 g (SOA) formation from anthropogenic, biomass burning and
kg−1 of trash burned) and antimony (0.011 g kg−1 of PM2.5 trash burning sources is parametrized based on observed ratios
emitted by trash). As there is no available information of organic aerosols to CO as described in ref 11. Due to lack of
concerning the temporal variations of trash burning, we assume information on SOA formation from trash burning emissions,
here that daily emissions are the same every day and that the an emission factor of 0.025 g of surrogate SOA precursor per g
burning is most intense during the afternoon hours as is of CO was used (and converted to SOA upon atmospheric
common for biomass burning activities, due to the higher oxidation), based on recent biomass burning field studies
ambient temperatures and lower humidities (Figure 1d). summarized by ref 26. An alternative method to calculate SOA
2.2. Submicrometer Particulate Antimony (Sb) Meas- would be based on the known SOA precursors in the emissions,
urements. Antimony has been identified as a potential tracer benzene, and toluene. Assuming an SOA yield of 10% for these
of trash burning emissions in Mexico City2 and other species, we obtain a ratio of 0.0031 g SOA/g CO. Accounting
locations.15 To evaluate the trash burning emission estimates for SOA formation from nontraditional precursors (e.g., ref 27)
calculated in this study, we use submicrometer antimony will increase SOA formation several-fold, resulting in predicted
concentrations from the Aerodyne high resolution aerosol mass SOA levels similar to those predicted from the ref 26 approach.
spectrometer (AMS) measurements performed in Mexico City The simulations were carried out from March 18 to 29, 2006.
at the T0 site. The instrument was located in a mixed Two-way nesting was applied between a coarse-scale simulation
residential/industrial neighborhood, and was located about 200 at 36 km horizontal resolution covering all of Mexico and the
m from a major roadway. The AMS has been used before to North of Central America, and a fine-scale simulation with a 4
quantitatively detect lead (Pb) in submicrometer particles in km horizontal grid covering the Mexico City valley. The
Mexico City.16 Peaks corresponding to both Sb isotopes (121Sb updated MCMA 2006 anthropogenic emissions inventory
and 123Sb) in the AMS high resolution spectra were identified based on ref 9 is used. The primary aerosol organic emissions
using their exact mass, and their identity was confirmed by their for anthropogenic sources have been increased by a factor of 2
isotopic ratio. Sb was quantified using a vaporization model as suggested by ref 25. The biomass burning emissions are
similar to that for Pb,16 a collection efficiency of 0.5 that was derived from MODIS satellite retrievals and are similar to those
found to be accurate for other species measured by the AMS used by ref 27. Shrivastava et al.25 used an earlier estimate of
including Pb, and an ionization efficiency relative to nitrate the municipal trash burning emissions in their modeling study
(RIESb) of 0.75. RIESb was derived from the ionization cross- and lumped trash burning emissions together with other
section of Sb by 70 eV electrons (7.4 A2).17,18 Measurement anthropogenic sources; however, a numerical error in the
uncertainties are estimated as ±50%. Sb and its chloride and emission processing produced an estimate that was an order of
oxide have similar or lower melting and boiling temperatures magnitude too low. Trash burning emissions are derived from
than Pb,19 and thus are expected to evaporate and be detected this work. The initial and boundary conditions for the
similarly to Pb in the AMS. The submicrometer Sb data from meteorological variables are taken every 6 h from the National
the AMS might be biased low if some of the ambient Sb was in Center for Environmental Prediction Final Operational Global
the form of species with much lower vapor pressures than the Analysis data (NCEP/FNL) and the model coarse-scale
metal, chloride, or oxide, and did not evaporate in a time scale domain is nudged to the large-scale analyzed meteorology for
of several hours at 600 °C under high vacuum. Antimony has a the predicted wind, temperature, and specific humidity. The
high material density (6.7 g cm−3 for the pure element), but initial and boundary conditions for gas-phase and aerosol
may only represent a small fraction of the mass of the particles variables were obtained from the MOZART4 global chemistry-
in which it is present. If an Sb-containing particle was mostly transport model.28 The ability of state-of-the-art regional
composed of other metals of similarly high densities, the size models including WRF/Chem to simulate the meteorology,
cut of the AMS in physical diameter may be significantly below boundary layer height, and the concentrations of the main gas-
1 μm.20 On the other hand, if the rest of the particle is mostly phase pollutants in the Mexico City region has been evaluated
composed of OA (as expected for trash burning particles), the in detail in previous publications and found to be reasonable
usual AMS size cut around 1 μm (vacuum aerodynamic (e.g., refs 24 and 29).
4952 dx.doi.org/10.1021/es203954r | Environ. Sci. Technol. 2012, 46, 4950−4957
Environmental Science & Technology Article

Table 2. Anthropogenic Emissions Inventory and Estimated Trash Burning Emissions from This Study (Tons Day−1) Integrated
over the Mexico City Metropolitan Area (MCMA)
MCMAa (tons day−1) CO NOx NH3 SO2 volatile organic compoundsd primary organic aerosols (POA)
b e
anthropogenic emissions 2006 5574 332 58 1193 1943 29c
trash burning 166 11 4 2 38 25
a
Emissions are integrated from 19.1 to 19.86N, and 99.42 to 99.69W. bIncludes mobile sources, area sources, and industrial source emissions based
on ref 9. Biomass burning emissions are not included. cPOA emissions reported in the MCMA 2006 inventory have been doubled according to ref
25. dTrash burning emitted VOC includes species reported in SI-Table 1 (propene, benzene, toluene, ethene, acetic acid, formaldehyde, methanol,
formic acid, acetylene, acetaldehyde, acetone). eSO2 anthropogenic emissions also include volcanic and industrial contributions.

Figure 2. Nine-day average diurnal profiles comparing antimony (Sb, ng m−3) submicrometer mass concentrations as measured by the aerosol mass
spectrometer at T0 downtown site, and predicted by the WRF/Chem model at both downtown T0 and suburban T1 locations from 20 to 29 March
2006. Panel (a) shows the modeled Sb from trash burning source only, whereas panel b includes both trash burning and estimated traffic Sb
contributions. The measured CO diurnal profile is also plotted (in green, panel a). The gray shading denotes the variability among observations (1
sigma) and the red bars indicate the variability in the model.

3. RESULTS AND DISCUSSION 3.2. Evaluation of Trash Burning Emission Estimates.

The uncertainties associated with these emission estimates are
3.1. Trash Burning Emission Estimates. The trash
large, and here we examine two ways of evaluating them. First, a
burning emission estimates over the MCMA as derived in
back-of-the-envelope calculation of the trash burning emissions
this work are summarized and compared in Table 2 to the
in Mexico City was performed. If we consider that the Mexico
anthropogenic MCMA emissions inventory for 2006 for four
City population is about 20 million inhabitants, and that each
major pollutant categories. The results suggest that trash
person produces about 1 kg of trash per day, the daily trash
burning contributes a very small fraction to the emissions of production can be estimated to 20 kTons. This number is close
gas-phase pollutants such as CO (3%), NOx (3%) or VOC (2%, to another independent estimate by F. Menendez (Management
species included are listed in SI-Table 1), in comparison to of Solid Waste Commission, personal communication, 2010) where
other anthropogenic sources. VOCs emitted by trash burning 22.1 kTons of trash production per day was estimated for the
include aromatic compounds such as toluene and benzene that MCMA region. According to local authorities, the city has 17
are known SOA precursors, but the increase in their emissions 000 sanitation workers and a fleet of more than 2000 trucks
due to trash is very small, so that their impact on SOA is that can collect 11.8 kTons of trash per day. If we assume that
expected to be negligible. the rest (8.2 kTons) is recycled and burned in equal
More importantly, the trash burning contribution to POA proportions, the burned fraction exceeds 4 kTons of trash
emissions is estimated to be substantial. According to our per day. After applying the emission factor for POA of 6.9 g/kg,
estimates, the trash burning contribution to POA is of the same we can estimate that about 28 Tons of POA are emitted due to
order as all other anthropogenic sources, that is, the trash trash burning every day in the vicinity of Mexico City, which is
burning is responsible for 25 tons day−1 of organic aerosol of the same order as the total emission estimates presented
emissions vs 29 tons day−1 for other anthropogenic sources. above and retrieved using the socio-demographic criteria.
Although the emitted POA amounts are comparable, the trash Second, to verify our estimates we use particulate antimony
burning emissions are expected to be more harmful to local (Sb) that was reported to be a potential tracer of trash burning
populations as they contain significant amounts of toxic emissions.2 High levels of Sb in PM2.5 and PM10 particles were
materials.30 observed in Mexico City and attributed mainly to brake wear
Figure 1c shows the spatial distribution of estimated trash emissions based on factor analysis31 or on comparable ratios of
burning emissions for POA. The emission flux ranges from zero Sb/PM2.5 found in other urban areas where trash burning is
in the downtown area to 30−40 kg km−2 day−1 at the edges of very minor.32 These authors reported the highest Sb
the city. Two potential emission hot spots emerge from our concentrations of 15 ng m−3 in PM2.5 at the urban site, with
inventory. The largest one is located in the southeastern part of lower levels (8 ng m−3) at the T1 site northeast of the city,
the city, whereas the second one is in the northeast close to the where the trash burning might be playing a larger role
T0 measurement site. These areas are densely populated and according to our estimates. The higher concentrations at T0,
are associated with low socioeconomic classes. closer to the areas of higher traffic emissions, are consistent
4953 dx.doi.org/10.1021/es203954r | Environ. Sci. Technol. 2012, 46, 4950−4957
Environmental Science & Technology Article

Figure 3. Predicted changes in POA, OA and PM2.5 due to trash burning emissions. Spatial distribution of (a) surface mass concentrations of POA as
initially predicted by the model in the vicinity of Mexico City between 20 and 29 March 2006. The absolute (b) and percentage (c) increase in POA
due to the introduction of trash burning emissions is shown. The average percentage increase in total organic aerosol (d) and PM2.5 (e) are also
shown.

with the dominance of brake wear for PM2.5 Sb. Most of the antimony was computed using the trash burning emission
brake wear emissions occur in the supermicrometer range, with factors from ref 2 and the chemistry-transport model used here.
a ∼10−20% ratio of PM1/PM2.5 for Sb.33,34 This would suggest The brake wear contribution has been estimated by
an impact of brake wear to submicrometer Sb of the order of multiplying the diurnal cycle of ambient CO measurements
1−3 ng m−3. (shown in Figure 2a, and after subtraction of a background CO
High levels of Sb have also been observed in the emissions concentration of 100 ppb) using an estimated emission ratio of
from garbage burning experiments, suggesting that road traffic SbPM1/CO of 0.85 × 10−6 g/g, based on the measurements of
is not the only source of Sb in PM2.5. Antimony is used as a ref 33. The average measured diurnal cycle of Sb shows values
flame retardant for textiles and in lead alloys in batteries, and ranging from 3 to 5 ng m−3 with an increase in concentrations
antimony trioxide is used as a catalyst in the production of of 1−2 ng m−3 during the morning traffic rush hours, consistent
with road traffic being a source of submicrometer Sb. During
polyethylene terephthalate, found in soft drink bottles and
the day, concentrations of particulate Sb stay steady in the
textile polyester fiber, all of which are very common garbage
growing and well-developed boundary layer showing a much
items.2 Sb from combustion sources will be formed by an flatter diurnal profile than that of CO. This suggests that Sb is
evaporation/condensation process and will be dominantly in being continuously produced by other sources in addition to
the submicrometer mode.34 It is also possible that other road traffic with stronger emissions during the day, as assumed
industrial sources of Sb exist in Mexico City as is the case for here for trash burning.
other metals,35,16 but there is insufficient information to The predicted submicrometer Sb levels from trash burning
estimate their emissions. Their absence from the model may only at T0 are ∼53% of the averaged measured levels. During
result in a low bias for the predicted Sb. the morning rush hour the predicted Sb is 2−3 times lower, as
Figure 2 shows the comparisons between the model expected in the absence of mobile emission sources. When the
predicted levels of submicrometer Sb and the measurements estimated emissions from brake wear are added, the modeled
at the T0 site north of downtown Mexico City. The predicted concentrations are within the uncertainties of the measure-
4954 dx.doi.org/10.1021/es203954r | Environ. Sci. Technol. 2012, 46, 4950−4957
Environmental Science & Technology Article

Figure 4. Average diurnal profiles comparing primary and biomass burning organic aerosols (POA and BBOA) submicrometer mass concentrations
as observed by the aerosol mass spectrometer and predicted by WRF/Chem at T0 and T1 sites from 20 to 29 March 2006. The gray shading denotes
the variability among observations (1 sigma).

ments, adding about 2 ng m−3 of PM1 Sb during the rush hour the modeling was based on the oxidation of VOC and
peak. The highest predicted values of trash buring Sb (∼3.5 ng condensation of their reaction products, the change in
m−3) are found during the afternoon hours at the suburban site concentrations would have also been very small as the
T1 that was located downwind of trash emissions. This emissions of VOC precursors (aromatics) have not significantly
comparison suggest that the trash emission estimates are of the increased due to trash burning. Given the fact that SOA
right order, and that they can be used to evaluate the estimated contributes about half of the total organic mass in Mexico
impact of trash burning on air quality in Mexico City. City,36 the relative contribution of trash burning to the total
3.3. Estimated Contribution of Trash Burning Emis- organic aerosol is smaller as shown in Figure 3d. According to
sions to Organic Aerosol Levels in Mexico City. As our model predictions, the complete mitigation of trash burning
discussed above the contribution of trash burning emissions to from the Mexico City valley could reduce the levels of organic
concentrations of main gaseous pollutants was found to be aerosols by 2−5% downtown, and 15−40% in the most
relatively small (<3%) in this region, so we have focused our impacted areas. The highest reduction is expected to occur near
discussion below exclusively on the aerosol fraction. The largest densely populated areas of the city, and therefore the health
increase due to trash burning was found for POA as expected benefits might be considerable.
from the estimated doubling of POA emissions due to trash Figure 3e shows the contribution of trash burning to the
burning, compared to the initial inventory. Figure 3 shows the modeled total PM2.5 concentrations in the vicinity of Mexico
predicted enhancement in 9-day average surface concentrations City. The model results suggest that trash burning represents
of POA resulting from the impact of trash burning in the less than 1% of the PM2.5 mass in the city center where the
Mexico City valley, while Figure 4 shows the diurnal profiles of PM2.5 concentrations are the highest, from 2 to 7% in the
POA for the urban (T0) and suburban (T1) sites. In
prevailing outflow region north and northeast of the city, and
comparison to POA concentrations generated by other
about 15% in the strongest emission hotspot which experiences
anthropogenic sources (Figure 3a), the predicted POA levels
average PM2.5 levels of 5−8 μg m−3. The contribution of trash
from trash burning have a lower peak, but are also more widely
burning to secondary inorganic aerosol compounds ranged
spread over the Mexico City valley. Simulated POA
from 1% downtown to 7% in the emission hotspots (not shown
concentrations increase with the inclusion of trash burning by
0.2−0.6 μg m−3 in the downtown area, and up to 1.5−2 μg m−3 here). Although NOx, NH3, and SO2 are emitted from trash
in the southeast and east suburbs of the city. The locations of burning, the amounts emitted are relatively small compared to
maximum increases are unsurprisingly the trash burning other sources in this region.
emission hotspots (Figure 1c), resulting in higher exposures Our inventory and model-based estimates of the trash
for areas of high population density of low socioeconomic level. burning effect on aerosols are more than five times lower than
Some of the emitted material is advected downwind to the the estimates by ref 2. These authors estimated a contribution
northern edges of the plateau (north of T1 site) under the of trash burning of ∼28% of PM2.5 in downtown Mexico City,
influence of southerly and southwesterly winds. The model or about 10 μg m−3, based on attributing 100% of the measured
results indicate that the ∼2/3 of the MCMA located to the Sb in PM2.5 to this source. However as discussed above there
southeast and east parts of the valley that generally experience are other important sources of Sb, especially in the PM1 to
low levels of POA emissions see their average levels increase by PM2.5 size range, and a lower predicted impact than the one
40−80% during trash burning. reported by ref 2 is consistent with the fact that other sources
The change in SOA concentrations was found to be also contribute to submicrometer Sb concentrations in the city.
negligible (<1%) when the trash burning emissions are Finally, our results show a strong spatial variability in the role
included in the simulations. This result is expected since that trash burning plays on air quality in Mexico City. The
SOA formation is parametrized in our study based on CO model suggests that trash burning impacts are highest close to
emissions (based on the results summarized by ref 26 for their source region, while the regional impact is more limited.
biomass burning sources), and the increase in CO emissions Mitigation of trash burning could lead to reduced human health
due to trash burning is very limited (3%). As discussed above, if effects, especially in the poorer areas of Mexico City, while the
4955 dx.doi.org/10.1021/es203954r | Environ. Sci. Technol. 2012, 46, 4950−4957
Environmental Science & Technology Article

regional climate impact of these emissions is expected to be NCAR (FINN): A high resolution global model to estimate the
limited in this case. emissions from open burning. Geosci. Model Dev. 2011, 4, 625−641,

■
DOI: 10.5194/gmd-4-625-2011.
ASSOCIATED CONTENT (11) Hodzic, A.; Jimenez, J. L. Modeling anthropogenically
controlled secondary organic aerosols in a megacity: A simplified
*
S Supporting Information
framework for global and climate models. Geosci. Model Dev. 2011, 4,
SI-Table 1. This material is available free of charge via the 901−917, DOI: 10.5194/gmd-4-901-2011.
Internet at http://pubs.acs.org.

■
(12) Ojeda-Benitez, S.; Armijo-de Vega, C.; Ysabel Marquez-
Montenegro, M. Household solid waste characterization by family
AUTHOR INFORMATION socioeconomic profile as unit of analysis. Resour. Conserv. Recyl. 2008,
Corresponding Author 52, 992−999.
(13) Akagi, S. K.; Yokelson, R. J.; Wiedinmyer, C.; Alvarado, M. J.;
*E-mail: [email protected].
Reid, J. S.; Karl, T.; Crounse, J. D.; Wennberg, P. O. Emission factors
Present Address for open and domestic biomass burning for use in atmospheric models.
∥
Now at: Facultad de Ciencias, Universidad Nacional Atmos. Chem. Phys. 2011, 11, 4039−4072, DOI: 10.5194/acp-11-4039-
Autónoma de México, Querétaro, México. 2011.
Notes (14) Lemieux, P. M.; Lutes, C. C.; Santoianni, D. A. Emissions of
The authors declare no competing financial interest. organic air toxics from open burning: A comprehensive review. Prog.

■
Energy Combust. 2004, 30, 1−32.
(15) García, J. H.; Lia, W.-W.; Cárdenasa, N.; Arimoto, R.; Walton, J.;
ACKNOWLEDGMENTS Trujillo, D. Determination of PM2.5 sources using time-resolved
We gratefully acknowledge T.J. Christian (University of integrated source and receptor models. Chemosphere 2006, 65, 2018−
Montana) for providing the TOC picture of trash burning. 2027.
This research has been supported by the National Center for (16) Salcedo, D; Onasch, T. B.; Aiken, A. C.; Williams, L. R.; Bd,
Atmospheric Research which is operated by the University Foy; Cubison, M. J.; et al. Determination of particulate lead using
Corporation for Atmospheric Research on behalf of the aerosol mass spectrometry: MILAGRO/MCMA-2006 observations.
National Science Foundation, and by US DOE grants Atmos. Chem. Phys. 2010, 10, 5371−5389.
(Atmospheric Systems Research program, BER) DE- (17) Freund, R. S.; Wetzel, R. C.; Shul, R. J.; Hayes, T. R. Cross-
section measurements for electron-impact ionization of atoms. Phys.
SC0006711 and DE-SC0006035.

■
Rev. A 1990, 41, 3575−3595.
(18) Jimenez, J. L.; Jayne, J. T.; Shi, Q; Kolb, C. E.; Worsnop, D. R.;
REFERENCES Yourshaw, I; et al. Ambient aerosol sampling using the Aerodyne
(1) Shindell, D.; Kuylenstierna, J.C. I.; Vignati, E.; et al. aerosol mass spectrometer. J. Geophys. Res. 2003, 108, 8425.
Simultaneously mitigating near-term climate change and improving (19) CRC Handbook of Chemistry and Physics, 92nd ed.; William, M.
human health and food security. Science 2012, 335, 183−189. H., Ed.; CRC Press: Boca Raton, FL, 2011.
(2) Christian, T. J.; Yokelson, R. J.; Cárdenas, B.; Molina, L. T.; (20) DeCarlo, P.; Slowik, J. G.; Worsnop, D. R.; Davidovits, P.;
Engling, G.; Hsu, S. C. Trace gas and particle emissions from domestic Jimenez, J. L. Particle morphology and density characterization by
and industrial biofuel use and garbage burning in central Mexico. combined mobility and aerodynamic diameter measurements. Part 1.
Atmos. Chem. Phys. 2010, 10, 565−584. Theory. Aerosol Sci. Technol. 2004, 38, 1185−1205, DOI: 10.1080/
(3) Andreae, M. O.; Merlet, P. Emission of trace gases and aerosols 027868290903907.
from biomass burning. Global Biogeochem. Cycles 2001, 15 (4), 955− (21) Grell, G. A.; Peckham, S. E.; Schmitz, R.; McKeen, S. A.; Frost,
966, DOI: 10.1029/2000GB001382. G.; Skamarock, W. C.; Eder, B. Fully coupled “online” chemistry
(4) Lemieux, P. M.; Lutes, C. C.; Abbott, J. A.; et al. Emissions of within the WRF model. Atmos. Environ. 2005, 39, 6957−6975,
polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans DOI: 10.1016/j.atmosenv.2005.04.027.
from the open burning of household waste in barrels. Environ. Sci. (22) Carter, W. P. L.: Documentation of the SAPRC-99 Chemical
Technol. 2000, 34, 377−384. Mechanism for VOC Reactivity Assessment, Draft report to the
(5) Loomis; Dana; Castillejos, M.; Gold, D.; McDonnell, W.; Borja- California Air Resources Board, Contracts 92-329 and 95−308, 8 May,
Aburto, V. H. Air pollution and infant mortality in Mexico City.
2000. http://www.cert.ucr.edu/∼carter/absts.htm (accessed Aug 15,
Epidemiology 1999, 10 (No 2), 118−123.
2011).
(6) Brauer, M.; Avila-Casado, C.; Fortoul, T. I.; Vedal, S.; Stevens, B.;
(23) Zaveri, R. A.; Easter, R. C.; Fast, J. D.; Peters, L. K. Model for
Churg, A. Air pollution and retained particles in the lung. Environ.
Simulating Aerosol Interactions and Chemistry (MOSAIC). J.
Health Perspect. 2001, 109, 1039−1043.
(7) Molina, L. T.; Madronich, S.; Gaffney, J. S.; Apel, E.; de Foy, B.; Geophys. Res., [Atmos.] 2008, 113, D13204 DOI: 10.1029/
Fast, J.; Ferrare, R.; Herndon, S.; Jimenez, J. L.; Lamb, B.; Osornio- 2007jd008782.
Vargas, A. R.; Russell, P.; Schauer, J. J.; Volkamer, R.; Zavala., M. An (24) Fast, J.; Aiken, A. C.; Allan, J.; Alexander, L.; Campos, T.;
overview of MILAGRO 2006 campaign: Mexico City emissions and Canagaratna, M. R.; Chapman, E.; DeCarlo, P. F.; de Foy, B.; Gaffney,
their transport and transformation. Atmos. Chem. Phys. 2010, 10, J.; de Gouw, J.; Doran, J. C.; Emmons, L.; Hodzic, A.; Herndon, S. C.;
8697−8760, DOI: 10.5194/acp-10-8697-2010. Huey, G.; Jayne, J. T.; Jimenez, J. L.; Kleinman, L.; Kuster, W.; Marley,
(8) Secretaria de Medio Ambiente, Gobierno del Distrito Federal. N.; Russell, L.; Ochoa, C.; Onasch, T. B.; Pekour, M.; Song, C.;
Inventario de Emisiones de Contaminantes Criterio para 2006. http:// Ulbrich, I. M.; Warneke, C.; WelshBon, D.; Wiedinmyer, C.; Worsnop,
www.sma.df.gob.mx/sma/index.php?opcion=26&id=501 (accessed D. R.; Yu, X. Y.; Zaveri, R. Evaluating simulated primary
Sep 1, 2011). anthropogenic and biomass burning organic aerosols during
(9) Song, J.; Lei, W.; Bei, N.; Zavala, M.; de Foy, B.; Volkamer, R.; MILAGRO: Implications for assessing treatments of secondary organic
Cardenas, B.; Zheng, J.; Zhang, R.; Molina, L. T. Ozone response to aerosols. Atmos. Chem. Phys. 2009, 9, 6191−6215, http://www.atmos-
emission changes: A modeling study during the MCMA-2006/ chem-phys.net/9/6191/2009/.
MILAGRO Campaign. Atmos. Chem. Phys. 2010, 10, 3827−3846, (25) Shrivastava, M.; Fast, J.; Easter, R.; Gustafson, W. I., Jr.; Zaveri,
DOI: 10.5194/acp-10-3827-2010. R. A.; Jimenez, J. L.; Saide, P.; Hodzic, A. Modeling organic aerosols in
(10) Wiedinmyer, C.; Akagi, S. K.; Yokelson, R. J.; Emmons, L. K.; a megacity: Comparison of simple and complex representations of the
Al-Saadi, J. A.; Orlando, J. J.; Soja, A. J. The Fire INventory from volatility basis set approach. Atmos. Chem. Phys. 2011, 11, 6639−6662.

4956 dx.doi.org/10.1021/es203954r | Environ. Sci. Technol. 2012, 46, 4950−4957

Environmental Science & Technology Article

(26) Cubison, M. J.; Ortega, A. M.; Hayes, P. L.; Farmer, D. K.; Day,
D.; Lechner, M. J.; Brune, W. H.; Apel, E.; Diskin, G. S.; Fisher, J. A.;
Fuelberg, H. E.; Hecobian, A.; Knapp, D. J.; Mikoviny, T.; Riemer, D.;
Sachse, G. W.; Sessions, W.; Weber, R. J.; Weinheimer, A. J.;
Wisthaler, A.; Jimenez, J. L. Effects of aging on organic aerosol from
open biomass burning smoke in aircraft and lab studies. Atmos. Chem.
Phys. 2011, 11, 12049−12064.
(27) Hodzic, A.; Jimenez, J. L.; Madronich, S.; Canagaratna, M. R.;
DeCarlo, P. F.; Kleinman, L.; Fast, J. Modeling organic aerosols in a
megacity: Potential contribution of semi-volatile and intermediate
volatility primary organic compounds to secondary organic aerosol
formation. Atmos. Chem. Phys. 2010, 10, 5491−5514, DOI: 10.5194/
acp-10-5491-2010.
(28) Emmons, L. K.; Walters, S.; Hess, P. G.; Lamarque, J.-F.; Pfister,
G. G.; Fillmore, D.; Granier, C.; Guenther, A.; Kinnison, D.; Laepple,
T.; Orlando, J.; Tie, X.; Tyndall, G.; Wiedinmyer, C.; Baughcum, S. L.;
Kloster, S. Description and evaluation of the Model for Ozone and
Related chemical Tracers, version 4 (MOZART-4). Geosci. Model Dev.
2010, 3, 43−67.
(29) Hodzic, A.; Jimenez, J. L.; Madronich, S.; Aiken, A. C.;
Bessagnet, B.; Curci, G.; Fast, J.; Lamarque, J.-F.; Onasch, T. B.; Roux,
G.; Schauer, J. J.; Stone, E. A.; Ulbrich, I. M. Modeling organic aerosols
during MILAGRO: Importance of biogenic secondary organic
aerosols. Atmos. Chem. Phys. 2009, 9, 6949−6981, DOI: 10.5194/
acp-9-6949-2009.
(30) Institute of Medicine, Long-term health consequences of
exposure to burn pits in Iraq and Afghanistan, The National Academis
Press, 2011. http://www.nap.edu/catalog.php?record_id=13209 (ac-
cessed Sep 1, 2011).
(31) 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.; Artíñano, B. PM speciation and
sources in Mexico during the MILAGRO-2006 campaign. Atmos.
Chem. Phys. 2008, 8, 111−128.
(32) Querol, X. Interactive comment on “Trace gas and particle
emissions from domestic and industrial biofuel use and garbage
burning in central Mexico” by T. J. Christian et al. Atmos. Chem. Phys.
Discuss. 2009, 9, C1312−C1314.
(33) Bukowiecki, N.; Lienemann, P.; Hill, M.; Figi, R.; Richard, A.;
Furger, M.; Rickers, K.; Falkenberg, G.; Zhao, Y.; Cliff, S. S.; Prevot, A.
S. H.; Baltensperger, U.; Buchmann, B.; Gehrig, R. Real-world
emission factors for antimony and other brake wear related trace
elements: Size-segregated values for light and heavy duty vehicles.
Environ. Sci. Technol. 2009, 43, 8072−8078.
(34) Iijima, A.; Sato, K.; Fujitani, Y.; Fujimori, E.; Saito, Y.; Tanabe,
K.; Ohara, T.; Kozawa, K.; Furuta, N. Clarification of the predominant
emission sources of antimony in airborne particulate matter and
estimation of their effects on the atmosphere in Japan. Environ. Chem.
2009, 6, 122−132, DOI: 10.1071/EN08107.
(35) Moffet, R. C.; Desyaterik, Y.; Hopkins, R. J.; et al. Character-
ization of aerosols containing Zn, Pb, and Cl from an industrial region
of Mexico City. Environ. Sci. Technol. 2008, 42 (19), 7091−7097.
(36) Aiken, A. C.; Salcedo, D.; Cubison, M. J.; Huffman, J. A.;
DeCarlo, P. F.; et al. Mexico City aerosol analysis during MILAGRO
using high resolution aerosol mass spectrometry at the urban supersite
(T0)Part 1: Fine particle composition and organic source
apportionment. Atmos. Chem. Phys. 2009, 9, 6633−6653.

4957 dx.doi.org/10.1021/es203954r | Environ. Sci. Technol. 2012, 46, 4950−4957