Quantification of Global Primary Emissions of PM2.5, PM10, and TSP

Oct 27, 2014 - Recently, a global fuel consumption data product (PKU-FUEL-2007) with ... (44) The EFs for coal combustion in energy and industrial sec...
0 downloads 0 Views 4MB Size
Download PDF
Article pubs.acs.org/est

Quantification of Global Primary Emissions of PM2.5, PM10, and TSP from Combustion and Industrial Process Sources Ye Huang, Huizhong Shen, Han Chen, Rong Wang, Yanyan Zhang, Shu Su, Yuanchen Chen, Nan Lin, Shaojie Zhuo, Qirui Zhong, Xilong Wang, Junfeng Liu, Bengang Li, Wenxin Liu, and Shu Tao* Laboratory for Earth Surface Processes, College of Urban and Environmental Sciences, Peking University, Beijing 100871, China S Supporting Information *

ABSTRACT: Emission quantification of primary particulate matter (PM) is essential for assessment of its related climate and health impacts. To reduce uncertainty associated with global emissions of PM2.5, PM10, and TSP, we compiled data with high spatial (0.1° × 0.1°) and sectorial (77 primary sources) resolutions for 2007 based on a newly released global fuel data product (PKU-FUEL-2007) and an emission factor database. Our estimates for developing countries are higher than those previously reported. Spatial bias associated with large countries could be reduced by using subnational fuel consumption data. Additionally, we looked at temporal trends from 1960 to 2009 at country-scale resolution. Although total emissions are still increasing in developing countries, their intensities in terms of gross domestic production or energy consumption have decreased. PM emitted in developed countries is finer owing to a larger contribution from nonindustrial sources and use of abatement technologies. In contrast, countries like China, with strong industry emissions and limited abatement facilities, emit coarser PM. The health impacts of PM are intensified in hotspots and cities owing to covariance of sources and receptors. Although urbanization reduces the per person emission, overall health impacts related to these emissions are heightened because of aggregation effects.

■

INTRODUCTION Particulate matter (PM) in air affects both climate1 and health.2 Frequent occurrences of thick haze in eastern China have raised extensive health concerns in recent years.3,4 Based on the data collected by the American Cancer Society as a part of the Cancer Prevention Study II, Pope et al. estimated that a 10 μg/ m3 increase in PM2.5 (particles with aerodynamic size less than 2.5 μm) within a range from 5 to 30 μg/m3 in air could lead to a 6 and 8% increase in cardiopulmonary and lung cancer mortalities, respectively, for people aged 30 or older;2 while in 2005, 376000 premature deaths were attributed to exposure to PM2.5 in China.5 Additionally, PM in air may cause changes in surface temperature,6 solar radiation,7 cloud properties,8 and precipitation,9 leading to major effects on climate. It was estimated that 32−73% of ambient PM were from primary sources.10,11 Wildfires, dust, and sea salt were major natural sources. Both health and climate impacts of PM are size dependent. Evidence from both epidemiological and in vitro studies suggest that fine, especially untrafine, particles are more toxic.12,13 For climate impact, finer particles have longer residence time and are more effective in scattering solar radiation.14,15 A series of efforts were made to quantify primary emissions of PM from various sources on global,16,17 regional,18,19 and country20−25 scales. These inventories provide essential input for modeling fate, spatiotemporal distributions, and effects of © XXXX American Chemical Society

PM. Much effort has been invested to reduce uncertainties in these PM emission inventories; e.g., emissions of PM are compiled for developing and developed countries, using different emission factors (EFs), defined as mass of pollutant emitted per unit fuel consumed or material produced.26 Country- and period-specific EFs for given sources have been adopted to account for variation in fuels, facility type, operational status, environmental conditions, and other factors through time.20,27−29 The technology division method was developed and applied to describe the impact of abatement technologies.19,24,30−32 Still, source complexity and lack of data lead to large uncertainties in inventories. As one of many examples, primary organic aerosol may evaporated after emitted into the atmosphere,33 causing difficulty in emission estimation. Recently, a global fuel consumption data product (PKUFUEL-2007) with high spatial (0.1° × 0.1°) and source (64 fuel types) resolutions was compiled on the basis of subnational fuel consumption data for many large countries. Using subnational instead of national fuel data reduces the spatial bias due to uneven per-person fuel consumption within countries.34 Using this new data product and an EF database with recently Received: August 2, 2014 Revised: October 22, 2014 Accepted: October 27, 2014

A

dx.doi.org/10.1021/es503696k | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

published data, 0.1° × 0.1° gridded global emission inventories for TSP (total suspended particles), PM10 (particles with aerodynamic size less than 10 μm), and PM2.5 from combustion and industrial process sources were compiled for 2007 in this study (PKU-PM-2007, http://inventory.pku.edu.cn/). Dust and sea salt were excluded in this study because totally different approaches have to be used.35,36 Although there is growing concern on ultrafine particles with sizes less than 100 nm, the emission of them is difficult to quantify at this stage due to data availability. Historical PM emissions from 1960 to 2009 were compiled at country-scale resolution. The uncertainty in these inventories was characterized. The spatial and temporal distributions, changes in emission intensities and fine to coarse PM ratios, and their potential health impacts are discussed.

briquette, each with a different EF.28,47,50−52 For all other countries, use of honeycomb briquette was negligible.54 The fractions of biomass fuels consumed in woodstoves and fireplaces were taken from Bond et al.,31 while ratios of improved woodstove were from Shen et al.40 The fractions of shaft, precalciner, and other kilns were fitted using an S-curve (Table S6, Supporting Information).24,54 A large uncertainty (50%) was set to technologies with assumed uniform distributions. Noncompliance Rate. The results of a reliable survey conducted in Jiangsu, China, suggest that SO2 scrubbers perform well over 90% of the time.55 Considering that PM control facilities are often better maintained than SO2 scrubbers, we assumed a noncompliance rate of 5% (0−10%, uniformly distributed) for all developing countries. Spatial Distribution Mapping. The spatial distribution mapping of the emissions was conducted on the basis of a 0.1° × 0.1° grid map of fuel consumption database (PKU-FUEL2007)34 and corresponding EFs. In brief, the fuel inventory was generated on the basis of subnational fuel data for some countries (county data for China, the United States, and Mexico, 0.5° × 0.5° grid data for 36 Europe countries, and provincial/state data for India, Brazil, Canada, Australia, Turkey, and South Africa) and national data for others for most fuels. These data were disaggregated into 0.1° × 0.1° grids using total population, rural population, and gross domestic production as proxies, depending fuel categories. Data for wildfire, aviation, shipping, power stations, natural gas flaring, and agricultural waste were from the literature.34 Detailed descriptions can be found in a previous paper.34 For production of cement, lime, coke, and aluminum, the geolocations of the sources were directly from the literature38 except for cement production in China, which was disaggregated based on provincial production56 and industrial coal consumption as a proxy.34 For other noncombustion industrial processes, activity strengths for all countries from the literature were disaggregated using industrial coal consumptions as proxy.34 Uncertainty Analysis. A Monte Carlo simulation was conducted to characterize overall uncertainty. Variations in source strengths, EFs, coal ash contents, fractions, and efficiencies of control technologies, as well as the noncompliance rate, were modeled. Values of these parameters were randomly selected from predefined distributions based on literature values25,40 or those as mentioned in previous parts. Our results are presented as interquartile ranges below. To quantify relative contributions of various parameters to the overall uncertainty, simulations were conducted for each parameters in turn with other parameters fixed. Calculation of Relative Potential Health Effect. As a normalized dimensionless indicator, the relative potential health effect (RPHE) was proposed by Shen et al. for polycyclic aromatic hydrocarbons.40 A slightly modified method was applied in this study for PM2.5. In brief, the RPHE in a grid was proportional to the total population of the grid and the sum of the emission contributions of the grid and 120 surrounding grids (11 × 11 grid in total). The contributions of the 121 grids were weighted inversely to distances to the central grid, and the global RPHE were normalized.

■

METHODOLOGY Emission Estimation. Emissions of TSP, PM10, and PM2.5 from specific sources were calculated as products of the source strength (quantities of fuel consumed or material produced) and EFs for that particular source; 77 sources were included and are listed in Table S1 (Supporting Information). In addition to the 65 combustion sources,34 12 material production sources were incorporated, including five processes involved in the iron and steel industry (sintering, pig iron production, open hearth furnace, convertor, and arc furnace)37 as well as production of lime, glass, ferroalloy, lead, magnesium, zinc,38 and fertilizers.39 Detailed source information is given in Table S1 (Supporting Information). For years other than 2007, annual fuel consumption in individual countries are from Shen et al.,40 except for coal consumption in energy, industrial, and residential sectors in China41 and several sources associated with industrial processes.38,39 EFs. EFs reported in the literature are associated with large uncertainty due to differences among facilities and fuels, environmental and operation conditions, burning stages and so on. Our strategy was to collect as many data as possible, from which distributions of the EFs can be derived for Monte Carlo simulation to characterize uncertainty associated (Table S1, Supporting Information). More than 2500 EFs collected from the literature were compiled into a database. For fuels used in transportation, country- and time-specific EFs were projected using a series of regression models.42,43 The models for TSP are provided in Table S2 (Supporting Information), while fixed relationships between PM2.5 and TSP (0.85) and between PM10 and TSP (0.91) were applied.44 The EFs for coal combustion in energy and industrial sectors were ash content corrected20,25 based on coal ash contents (log-normal distributed with a standard deviation (SD) of 0.18 for log transformed ash content) from the literature.30,28,45−49 For all other sources, arithmetic means were used. For the sources with abatement facilities, the EFs for specific technologies (cyclones, wet scrubbers, electrostatic precipitators, or fabric filters) were calculated using their unabated EFs and removal efficiencies (Table S3, Supporting Information).18,20,24,25 Technology Splits. The overall EFs for specific country categories and specific year were different and were derived using technology split method.31 For the energy and industry sectors, penetration rates for pulverized coal-fired boilers, stokers, and various control technologies were taken from Bond et al.31 with slight modification (Tables S4 and S5, Supporting Information). Bituminous coal consumed in the residential sector of China was subdivided into chunk coal and honeycomb

■

RESULTS AND DISCUSSION Global Emission and Source Profiles. Annual total global emissions of TSP, PM10, and PM2.5 for 2007 were estimated to be 162 (123−224), 99 (80−130), and 78 (64−101) Tg, B

dx.doi.org/10.1021/es503696k | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

Figure 1. (A) Global distributions of total anthropogenic emissions of PM2.5. (B) Pie charts showing the relative contributions from major sectors for various regions. (C) Global distributions of per person anthropogenic emissions of PM2.5.

respectively, of which, anthropogenic sources accounted for 105 (74−158), 55 (42−78), and 40 (31−56) Tg. Although our results were broadly similar to those from EDGARv4.216 and HATP_V2,17 relatively large differences were found for individual countries and sectors, reflecting differences in EFs, noncompliance rates, and coal ash contents. For instance, anthropogenic PM10 (2007) and PM2.5 (2008) in developing countries were estimated to be 49.4 and 32.4 Tg in this study, respectively. These values were 4.7 and 12.9% higher than those reported in EDGARv4.2 and HTAP V2, respectively, even though some minor (less than 2% of the total) sources such as solvent and other product use and direct soil emissions in DEGARv4.2 and HTAP V2 were not included in our inventory. The calculated relative variations of the distributions from Monte Carlo simulation, defined as the differences between the 75th and the 25th percentiles divided by medians, were 41.8, 14.6, 3.1, 1.2, and 2.1% for EFs, fuel quantities, noncompliance rate, technology split rate, remove efficiency, and ash content,

respectively, indicating dominant contribution of EFs. Among the various sources shown in source profiles (Figure S1, Supporting Information), wildfires were the largest single source globally, followed by residential biomass burning, power generation, and cement production. In total, biomass burning, including wildfires, residential fuels, and agriculture wastes, accounted for 45.2 (TSP), 59.9 (PM10), and 67.6% (PM2.5) of the total global emission. Power generation contributed 14.0, 9.5, and 6.4% of TSP, PM10, and PM2.5, respectively. Unlike incomplete combustion products, such as black carbon and polycyclic aromatic hydrocarbons,40,57 rather small fractions (0.8−1.4%) of primary PM were from motor vehicles. Per person average annual global emissions of primary TSP, PM10, and PM2.5 from all sources for 2007 were 25.0, 15.6, and 12.4 kg, respectively. Even though per person fuel consumption in developed countries was higher,34 their per person PM emissions were significantly lower (6.6, 5.4, and 4.4 kg) than those in developing countries (18.8, 9.6, and 6.9 kg). This is C

dx.doi.org/10.1021/es503696k | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

developing countries brought millions of rural residents to cities. In China alone, more than 260 million people have moved to urban areas over the past three decades. This trend will continue for years to come.63 Thus, urbanization in developing countries will lead to a significant reduction in PM emission in the future. Our estimates suggest that annual emission of PM2.5 will be reduced by around 3800 ton per million people as rural residents migrate to cities, climbing up the energy ladder. Temporal Variation. Since spatially resolved fuel consumption data is only available for 2007, annual PM emissions from 1960 to 2009 could only be derived at country-scale resolution in this study. Temporal trends for total PM2.5 emissions from six sectors (left panels) and eight source types (right panels) are shown in Figure S5 (Supporting Information). We show estimates for the global total (with uncertainty ranges), developed countries, developing countries, and specifically for China, and the United States. Globally, emission of PM2.5 increased continuously from 1960 through the 1980s, and then it was relatively constant until 1996, before decreasing slowly, with strong fluctuations owing to fast changes in economic and energy demands.60 In comparison with EDGARv4.2 (Figure S6, Supporting Information), our temporal trends in total anthropogenic PM10 emissions were similar, with significant differences for countries with high emissions, reflecting differences in our technology division and EFs. The overall trends were very different between developed and developing countries (Figure S7, Supporting Information). In 1960, anthropogenic PM2.5 emissions were mainly from Europe (32.2%), east Asia (22.6%), North America (18.5%), and south and southeast Asia (12.9%). Post 1960, emissions decreased in Europe and North America but increased in Asia, making this continent the predominant emission region (66.7%). In developed countries, emissions were reduced in the transportation (gasoline) and industry (coal and processes) sectors. For example, since the 1990s, emissions in the United States decreased owing to the implementation of Clean Air Act and other regulations.64,65 In developing countries, increased emissions over the past 50 years were related to coal-based industrial activities. In China, for example, coal consumption has increased from 60 MT in 196066 to 2118 MT in 2009,60 resulting in enhanced PM emission. Although the number of motor vehicles has risen in recent years in China and other developing countries,67,68 PM2.5 emission from this source did not show a significant increase, reflecting the rapid implementation of emission regulations and controls. For instance, the China IV standard, equivalent to Euro IV, was introduced in Beijing (2008),69 Shanghai (2009),70 Guangzhou (2010),71 and across China (2011) (GB17691-2005 and GB18352.3-2005)72,73 much earlier than in developed countries with respect to its developing status (i.e., per person GDP). Emission histories in other developed and developing countries followed trends similar to those above (Figure S8, Supporting Information). Figure S9 (Supporting Information) shows a time series of annual emissions of PM10 and TSP for global, developed countries, and developing countries. Slight differences in the trends for PM2.5 were found. In developed countries, emissions of both PM2.5 and PM10 increased before 1970, partly owing to the increase in number of motor vehicles; there was no such trend for TSP, which is primarily of industrial origin.18,24

primarily because of differences in source profiles and EFs (Figure S2, Supporting Information). Strong dependences on biomass fuels, active cement production, and relatively high EFs for anthropogenic sources led to high per person PM emission in developing countries. In China, for example, more than 130 million rural households still use wood and crop residues for cooking and heating, contributing greatly to primary PM emission.58 Likewise, 78.5% of cement is produced in developing countries, where emission control facilities are often not installed.25,59 In contrast, major PM emission sources in developed countries, including industrial boilers and motor vehicles, are well controlled. Spatial Distribution. Geographical distributions of total and per person anthropogenic emissions of PM2.5 for 2007 are shown in Figure 1. The relative contributions from major sectors for various regions are shown as pie charts. Of the total emission of anthropogenic PM2.5, excluding shipping and aviation, 87.7% was from developing countries, with major contributions from China (14.3 Tg), India (5.5 Tg), Nigeria (1.6 Tg), Indonesia (1.3 Tg), and Russia (0.99 Tg). In contrast, among-country differences and spatial variations of per person emissions were much less marked. This reflects a correlation between densely populated and hot emission regions. The average per person anthropogenic PM2.5 emission for developed countries (4.4 kg) was also much lower than that for developing countries (6.9 kg). Among other reasons, biomass fuels with very high EFs were extensively used in residential areas of developing countries.60 The penetration rate of fabric filters in coal-fired power stations in China25 was much lower than that in the United States.61 Spatial distributions of total and per person anthropogenic emissions of primary PM10 and TSP are shown in Figure S3 (Supporting Information). Although their general patterns are similar to those for PM2.5, differences between developed and developing countries are larger than those for PM2.5 because coal-fired power stations, cement production, and heavy industry in developing countries all emit coarser PM. Spatial bias caused by significant differences in per person fuel consumption within large countries can be reduced substantially by using subnational, instead of national, fuel consumption data, when population is used as a proxy for disaggregation.34 We confirmed the advantage of this approach by comparing our PM2.5 emissions based on PKU-FUEL-2007 with results derived using national fuel consumption data. Relative differences for the countries with subnational fuel data are shown in Figure S4 (Supporting Information); values ranged from 6 (Brazil) to 96% (Mexico) difference in spatial distribution. Our fine spatial resolution allowed us to compare emissions between urban and rural areas for anthropogenic sources in all countries (Table S7, Supporting Information). Despite the fact that most industrial and transport sources locate in urban areas, residential fuel consumptions are quite different between rural and urban areas, especially in developing countries. Rural residents rely more on solid fuels, like crop residue, firewood, and coal, emitting large quantities of PM. In contrast, urban residents use cleaner fuels, like natural gas, liquefied petroleum gas, and electricity.62 As a result, per person annual primary PM emission in rural areas are much higher than those in urban areas. Further, this difference in developed countries (12 and 2.8 kg PM2.5 for rural and urban areas) is larger than that in developing countries (8.4 and 4.6 kg PM2.5 for rural and urban areas). Recent rapid economic growth and urbanization in D

dx.doi.org/10.1021/es503696k | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

Temporal changes in per person annual PM2.5 emissions from 1960 to 2009 in eight representative countries are shown in Figure S10 (Supporting Information); we plotted these as totals from all sources, as well as separately for industrial fuel use, industrial processes, and residential fuel consumption. Per person PM2.5 emissions in residential sectors of all eight countries decreased over time, with different rates primarily reflecting replacement of solid fuels (coal and biomass fuels) with cleaner ones (e.g., electricity, liquified petroleum gas, natural gas, and biogas). Additionally, decreases occurred on implementation of centralized heating systems in cities.74 For example, coal consumption in the UK dropped from 48.6 to 24.5 Tg from 1960 to 1970,60 while coal used in urban households in China decreased from 65.6 to 37.5 Tg from 1995 to 2007.62 For the industrial sector, per person annual PM2.5 emissions in developed countries peaked around 1960−1970, having declined continuously since this time because of enforcement of stricter regulations and improvements in abatement technologies.31 A drop in energy demand associated with the shock therapy period60 in the former Soviet Union led to a sharp decrease in PM2.5 emission. Meanwhile, increasing emissions are seen in China and India, the two largest transition economies, because of rapid industralization. Emission Intensity. We calculated emission intensities, i.e., anthropogenic emission (excluding shipping and aviation) per unit of fuel consumption (EIfuel) or emission per unit of GDP (EIGDP), to characterize how well emission was controlled and their respective fuel efficiencies. Globally, the average EIGDP for PM2.5, PM10, and TSP values for 2007 data were 0.61, 0.84, and1.59 g/USD, while their average EIfuel values were 90.9, 125, and 236 kg/TJ, respectively. Spatial distributions for EIfuel and EIGDP for PM2.5, PM10, and TSP are shown in Figure S11 (Supporting Information). Similar spatial distribution patterns were observed for all three PM size categories. In developing countries, both EIfuel and EIGDP values were significantly higher than those in developed countries. Using PM2.5 as an example, 2007 EIfuel values for low, lower middle, upper middle, and high income countries75 were 279, 197, 126, and 27.8 kg/TJ, while EIGDP values were 3.5, 1.7, 1.1, and 0.15 g/USD, respectively. The differences in EIfuel reflect extensive use of biomass fuel and coal in residential sectors, as well as the lack of effective abatement facilities in energy and industry sectors in developing countries.24,25,30 Dominant secondary industries with relatively low energy efficiency and high energy consumption led to relatively low EIGDP values in developing countries. For example, 86.6% of bricks53 and 78.5% of cement38 are produced in developing countries. Temporal trends for EIfuel and EIGDP for the three PM size fractions for the global total, the four economic categories listed above, as well as China, and the United States are shown in Figure 2. Although emission intensities are significantly different for developing and developed countries, they have all decreased at different rates since the 1960s. EIGDP decreased faster than EIfuel for the majority of countries, reflecting the global reduction in energy consumption per unit GDP. In general, emission intensities were negatively correlated with income. There were a few exceptions, e.g., EIfuel for PM10 and TSP for upper middle income countries, including China, are higher than those for lower middle and low income countries. This likely reflects emissions from heavy industry in China. In contrast, for PM2.5, even though Elfuel for China was among the highest before 1986, it decreased at an accelerated pace from 1986 to 2009 from 324

Figure 2. Time series of EIfuel (right panel) and EIGDP (left panel) values for PM2.5, PM10, and TSP. Results are shown for the global total, the four economic categories low, lower middle, upper middle, and high income countries, as well as China and the United States.

to 152, while values for low and lower middle income countries decreased at a much slower rate. These changes reflect a rapid increase in energy efficiency, penetration of abatement technologies in the industrial sector,25,45 and replacement of solid fuels with cleaner ones in the residential sector.60,62 Size Distribution. It is well documented that PM2.5 penetrates more deeply into the lungs compared with coarse particles.76,77 Epidemiological surveys also suggest that PM2.5 is more strongly associated with health impacts.78 Using the available TSP, PM10, and PM2.5 data, we looked at the size distributions of PM from various sources and countries. Figure 3 shows the relative contributions of the three PM size fractions (PM2.5, PM2.5−10, and PM>10,) to primary PM emission from major sources (A) and from a number of representative countries (B) in 2007. The ratio of these three fractions was 0.485:0.129:0.386 (shown as a red triangle; Figure 3) for all sources/countries. PM2.5 shows the most marked differences, with circles spread from left to right with little vertical change (Figure 3A). The finest PM was from agricultural and residential biomass fuel burning; their contributions to the total PM were above 90%. Contributions of PM2.5 from wildfires, transportation, and residential coal combustion varied from 65 to 88%, while industrial and energy sectors emitted relatively coarse PM, with a PM2.5 fraction below 30%. Such differences should be taken into consideration in formulating abatement strategies to protect human health. Differences among countries were also most pronounced for PM2.5 (Figure 3B). Of the various representative countries, contributions of PM2.5 to total PM were higher than 55% in the United States (67.7%), Germany (67.0%), France (76.9%), Australia (69.5%), and less than 50% in China (31.0%), India (37.7%), South Africa (33.9%), and Egypt (26.9%). Relative contributions of PM2.5 ranged between 32.2 and 82.8% in lower middle income countries like Central African Republic (67.7%), Democratic Republic of the Congo (68.3%), Ethiopia (74.6%), and Mozambique (68.2%). These differences were primarily related to different source profiles. The countries with highest PM2.5 E

dx.doi.org/10.1021/es503696k | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

Figure 3. Relative contributions of PM2.5, PM2.5−10, and PM>10 to total PM in 2007. (A) with respect to various emission sources and (B) with respect to selected countries. The circle areas are proportional to annual emissions. The left and right panels are various sources and various countries or country categories, respectively.

fractions were either those dominated by wildfires and deforestation, like Australia and Central African Republic, or those with better emission abatement, like the United States and Germany. Relatively low PM2.5 fractions were found in countries dominated by industrial emissions, such as China. To examine geographical distribution patterns for PM size, fine particle ratios (Rfine, defined as the ratios of PM2.5 to PM2.5−10) for all sources (except shipping and aviation) were calculated for all grids (Figure S12, Supporting Information); a similar map was compiled for the PM10/PM>10 ratio (Figure S13, Supporting Information). Although the spatial distributions of PM2.5 (Figure 1) and PM10 (Figure S3, Supporting Information) are similar, they are not identical (Figure S12, Supporting Information). Relatively high Rfine values in North America and Western Europe reflect the fact that coarse particles are largely removed by abatement facilities.61 While relatively high Rfine values in South America, as well as central and southern Africa are largely due to biomass combustion in both residential60 and wildfire/deforestation79 sectors. Rfine values from southeast Asia were higher than those in China because of biomass burning in both residential and agricultural sectors,34,60 while significant contribution of heavy industry activity resulted relatively coarse PM in China in comparison with other developing countries.24 Figure 4 shows the dynamic change in Rfine from 1960 to 2009 for total anthropogenic emissions, as well as those from three other sectors in selected countries. Rfine values showed an increase in most developed countries, since emission of coarser PM is easier to be removed in abatement facilities. It also reflect that the relative contribution of power stations, which emit more coarser PM than industrial processes, had decreased. While Rfine values for developing countries decreased slowly, driven by increasing coal usage in the residential sector from 1960 to 1990. A slow increase in Rfine in China and other developing countries also reflected increasing abatement efforts.

Figure 4. Temporal trends of Rfine for total anthropogenic emission in selected countries from 1960 to 2009.

We expect that the relative contribution of ultrafine particles will follow a same trend. Thus, this indicator is an important parameter in health impact assessments, in addition to total emission and exposure. We explored these differences by plotting Rfine against per person GDP for the United States, Germany, China, and India as representative countries (Figure S14, Supporting Information). For developed countries, (United States and Germany), Rfine values have continually increased because of advances in abatement technology and the shift toward cleaner energy. While Rfine values in developing countries (China and India) have decreased owing to the rapid growth of heavy industry during recent economic expansion. Potential Health Impact. The spatial covariance of PM emission and population densities intensifies the potential health impact of emissions at the population level in regions F

dx.doi.org/10.1021/es503696k | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

linked to both their strong emissions and proximity to receptors. Differences in RPHE between developed and developing countries were magnified by higher population density. For anthropogenic sources, the average RPHE of developing countries (0.92) was 12.5 times that of developed countries (0.07), while the difference in average emission was only 7.8 times. Similarly, the average RPHE/e value in developing countries (2.6 × 10−14/g) was higher than that in developed countries (1.6 × 10−14/g), because of the differences in source types and population proximity to these sources (e.g., the relatively high contributions from residential solid fuel consumption in developing countries). On the basis of the high-resolution maps generated in this study, RPHE and RPHE/e were determined separately for urban and rural areas.34 Although average per person emission of PM2.5 in rural areas was higher than that for urban areas, both RPHE and RPHE/e in urban areas (0.73 and 5.7 × 10−14/ g) were higher than for rural areas (0.26 and 9.6 × 10−15/g). Again, this reflects highly concentrated sources and receptors in cities. This is especially true for developed countries because of their higher urbanization rates. Thus, aggregation effects related to rapid urbanization in the past decades and in the near future, has caused and will continue to cause high health risks, despite reductions in the per person PM2.5 emission. Studies to quantify the influence of urbanization on emissions and their associated health risks using atmospheric transport and population exposure models are urgently needed. Subnationally disaggregated fuel consumption data (PKUFUEL-2007) was used for PM inventories in this study;34 hence, spatial bias in emission data, and their associated health risks (RPHEs) could be calculated with a reduced uncertainty by taking into account uneven spatial distributions of per person fuel consumption and emission. Figure S15 (Supporting Information) compares spatial distributions of RPHEs derived from both subnationally and nationally disaggregated activity data for selected countries: China, the United States, India, Canada, and Australia. The reduction in spatial bias of RPHEs was similar to that for emissions, but was enhanced because emissions from adjacent grids are also taken into account in RPHE calculations. With nationally disaggregated results, the overestimation or underestimation of RPHE were obvious. In the case of China, RPHEs were overestimated in Tibet, Qinhai, northern Inner Mongolia, and Guangdong, while they were underestimated in most other regions, especially in the three northeastern provinces. It appears that the subnational aggregation method greatly reduced spatial bias not only for emissions but also for estimations of their health-related effects.

with high anthropogenic emissions. By modifying the method proposed by Shen et al.,40 we calculated RPHEs for PM2.5 (Figure 5A) in 2007. The indicator is only a rough estimation

Figure 5. (A) Geographical distribution of the relative potential health effect (RPHE) indicator based on PM2.5 and population. The frequency distributions of normalized grid emission and RPHE are shown in the inserted panel. (B) Relative contributions of various sources to total PM2.5 emission, RPHE, and RPHE/e (RPHE/ emission). Sources are ranked from left to right, with decreasing values of RPHE/e.

of potential impact without taking emission height, source type, PM size, meteorological conditions, and transformation processes into consideration. Compared with the global PM2.5 emission map (Figure 1), the RPHEs in hot emission areas, especially east, south, and southeast Asian regions, were elevated. This is due to the fact that high population density led to strong anthropogenic emissions, and subsequently to heavy exposure of large populations to these emissions. Frequency distributions of normalized grid emissions and RPHEs also showed this trend and the central tendency of RPHE shifted to right, reflecting elevated risk (Figure 5A). In contrast, RPHEs were weak in South America, central/southern Africa, and Australia, where strong emissions occurred in low population density regions related to wildfires and deforestation. Different sources contributed differently to RPHE as seen in Figure 5B, comparing total emissions, RPHEs, and RPHE/e (RPHE per unit emission) of various sources. Motor vehicles had highest values (6.06 × 10−14/g) because of their inherent link to people, while the RPHE/e of navigation (2.55 × 10−16/ g), which occurs in remote locations, was extremely low. Similarly, although open burning of biomass (wildfire, deforestation, and agricultural waste) contributed to approximately half of the total PM2.5 emission globally, RPHE/e values for these sources were rather small, since they occurred relatively far away from populated areas. Activities with relatively high RPHE/e values included power stations (3.85 × 10−14/g), industry (2.75 × 10−14/g), and residential sources (1.76 × 10−14/g). In terms of their health effects, highest RPHEs were found for emissions from power stations, cement production, residential crop residue and firewood burning,

■

ASSOCIATED CONTENT

* Supporting Information S

Materials, including our source list, EFs for various sources, regression models for EFs, PM removal efficiency and parameters for technology split, PM source profiles, geographical distributions for total and per person anthropogenic PM emissions, relative differences between subnationally and nationally disaggregated emissions, rural and urban anthropogenic PM2.5 emissions for 222 countries/regions, time trends for total PM2.5 emissions from six sectors and eight source types, comparison with other inventories, relative contributions of 12 global regions to PM2.5 emissions, time trends for PM2.5 emission from motor vehicles, time trends for PM10 and TSP from all sources, time trends for per person PM2.5 emissions in G

dx.doi.org/10.1021/es503696k | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

(15) Seinfeld, J. H.; Pandis, S. N. Atmospheric chemistry and physics: from air pollution to climate change; John Wiley & Sons: New York, 2006. (16) European Commission, Joint Research Centre (JRC)/Netherlands Environmental Assessment Agency (PBL). Emission Database for Global Atmospheric Research (EDGAR), release version 4.2.2011; http://edgar.jrc.ec.europa.eu. (17) Janssens-Maenhout, G.; Dentener, F.; Van Aardenne, J.; Monni, S.; Pagliari, V.; Orlandini, L.; Klimont, Z.; Kurokawa, J.; Akimoto, H.; Ohara, T.; Wankmueller, R.; Battye, B.; Grano, D.; Zuber, A.; Keating, T. EDGAR-HTAP: a Harmonized Gridded Air Pollution Emission Dataset Based on National Inventories; European Commission Publications Office: Ispra, Italy, 2012; JRC68434, EUR Report No. EUR 25 299-2012, ISBN 978-92-79-23122-0, ISSN 0013-936X. (18) Klimont, Z.; Cofala, J.; Bertok, I.; Amann, M.; Heyes, C.; Gyarfas, F. Modelling particulate emissions in Europe. A Framework to Estimate Reduction Potential and Control Costs. IR-02-076; IIASA: Austria, 2002; http://www.iiasa.ac.at/rains/reports.html. (19) Zhang, Q.; Streets, D. G.; Carmichael, G. R.; He, K. B.; Huo, H.; Kannari, A.; Klimont, Z.; Park, I. S.; Reddy, S.; Fu, J. S.; Chen, D.; Duan, L.; Lei, Y.; Wang, L. T.; Yao, Z. L. Asian emissions in 2006 for the NASA INTEX-B mission. Atmos. Chem. Phys. 2009, 9, 5131−5153. (20) U.S. Environmental Protection Agency. Clearinghouse for Inventories & Emissions factors; http://www.epa.gov/ttn/chief/index. html (accessed Oct 11, 2012). (21) National Atmospheric Emissions Inventory (NAEI); http:// naei.defra.gov.uk/data_warehouse.php (accessed Sep 27, 2012). (22) Reddy, M. S.; Venkataraman, C. Inventory of aerosol and sulphur dioxide emissions from India: IFossil fuel combustion. Atmos. Environ. 2002, 36 (4), 677−697. (23) Zhang, Q.; Streets, D. G.; He, K.; Klimont, Z. Major components of China’s anthropogenic primary particulate emissions. Environ. Res. Lett. 2007, 2 (4), 045027. (24) Lei, Y.; Zhang, Q.; He, K. B.; Streets, D. G. Primary anthropogenic aerosol emission trends for China, 1990−2005. Atmos. Chem. Phys. 2011, 11 (3), 931−954. (25) Zhao, Y.; Nielsen, C. P.; Lei, Y.; McElroy, M. B.; Hao, J. Quantifying the uncertainties of a bottom-up emission inventory of anthropogenic atmospheric pollutants in China. Atmos. Chem. Phys. 2011, 11 (5), 2295−2308. (26) Junker, C.; Liousse, C. A global emission inventory of carbonaceous aerosol from historic records of fossil fuel and biofuel consumption for the period 1860−1997. Atmos. Chem. Phys. 2008, 8 (5), 1195−1207. (27) Zhang, J. J.; Morawska, L. Combustion sources of particles: 2. Emission factors and measurement methods. Chemosphere 2002, 49 (9), 1059−1074. (28) Chen, Y.; Zhi, G.; Feng, Y.; Fu, J.; Feng, J.; Sheng, G.; Simoneit, B. R. Measurements of emission factors for primary carbonaceous particles from residential raw - coal combustion in China. Geophys. Res. Lett. 2006, 33(20). (29) Shen, G. F.; Wei, S. Y.; Wei, W.; Zhang, Y. Y.; Min, Y. J.; Wang, B.; Wang, R.; Li, W.; Shen, H. Z.; Huang, Y.; Yang, Y. F.; Wang, W.; Wang, X. L.; Wang, X. J.; Tao, S. Emission factors, size distributions, and emission inventories of carbonaceous particulate matter from residential wood combustion in rural China. Environ. Sci. Technol. 2012, 46 (7), 4207−4214. (30) Bond, T. C.; Streets, D. G.; Yarber, K. F.; Nelson, S. M.; Woo, J. H.; Klimont, Z. A technology-based global inventory of black and organic carbon emissions from combustion. J. Geophys. Res. 2004, 109, D14. (31) Bond, T. C.; Bhardwaj, E.; Dong, R.; Jogani, R.; Jung, S. K.; Roden, C.; Streets, D. G.; Trautmann, N. M. Historical emissions of black and organic carbon aerosol from energy-related combustion, 1850−2000. Glob. Biogeochem. Cycles 2007, 21, GB2018. (32) Klimont, Z.; Cofala, J.; Xing, J.; Wei, W.; Zhang, C.; Wang, S.; Kejun, J.; Bhandari, P.; Mathur, R.; Purohit, P.; Rafaj, P.; Chambers, A.; Amann, M. Projections of SO2, NOx and carbonaceous aerosols emissions in Asia. Tellus B 2009, 61 (4), 602−617.

eight representative countries, geographical distributions of emission intensity, Rfine and PM10/PM>10 ratios, relationship between Rfine and per person GDP, spatial distributions of RPHEs derived from the subnationally, and nationally disaggregated activity data. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone and Fax: +86-10-62751938. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Funding for this study was provided by the National Natural Science Foundation of China (41390240 and 41130754).

■

REFERENCES

(1) Quaas, J.; Boucher, O.; Bellouin, N.; Kinne, S. Satellite-based estimate of the direct and indirect aerosol climate forcing. J. Geophys. Res. 2008, 112, D05204. (2) Pope, C. A., III; Burnett, R. T.; Thun, M. J.; Calle, E. E.; Krewski, D.; Ito, K.; Thurston, G. D. Lung cancer, cardiopulmonary mortality, and long-term exposure to fine particulate air pollution. JAMA 2002, 287 (9), 1132−1141. (3) Tao, M.; Chen, L.; Su, L.; Tao, J. Satellite observation of regional haze pollution over the North China Plain. J. Geophys. Res. 2012, 117, D12203. (4) Wang, L. T.; Wei, Z.; Yang, J.; Zhang, Y.; Zhang, F. F.; Su, J.; Meng, C. C.; Zhang, Q. The 2013 severe haze over southern Hebei, China: model evaluation, source apportionment, and policy implications. Atmos. Chem. Phys. 2014, 14 (6), 3151−3173. (5) Nawahda, A.; Yamashita, K.; Ohara, T.; Kurokawa, J.; Yamaji, K. Evaluation of premature mortality caused by exposure to PM2. 5 and ozone in East Asia: 2000, 2005, 2020. Water Air Soil Pollut. 2012, 223 (6), 3445−3459. (6) Giorgi, F.; Bi, X. Q.; Qian, Y. Indirect vs. direct effects of anthropogenic sulfate on the climate of east Asia as simulated with a regional coupled climate-chemistry/aerosol model. Climatic Change. 2003, 58, 345−376. (7) Qian, Y.; Kaiser, D. P.; Leung, L. R.; Xu, M. More frequent cloudfree sky and less surface solar radiation in China from 1955 to 2000. Geophys. Res. Lett. 2006, 33, L01812. (8) Kawamoto, K.; Hayasaka, T.; Uno, I.; Ohara, T. A correlative study on the relationship between modeled anthropogenic aerosol concentration and satellite-observed cloud properties over east Asia. J. Geophys. Res. 2006, 111, D19201. (9) Huang, Y.; Chameides, W. L.; Dickinson, R. E. Direct and indirect effects of anthropogenic aerosols on regional precipitation over East Asia. J. Geophys. Res. 2007, 112, D03212. (10) Wang, X. H.; Bi, X. H.; Sheng, G. Y.; Fu, J. M. Chemical composition and sources of PM10 and PM2.5 aerosols in Guangzhou, China. Environ. Monit. Assess. 2006, 119, 425−439. (11) Hang, N. T.; Kim Oanh, N. T. Chemical characterization and sources apportionment of fine particulate pollution in a mining town of Vietnam. Atmos. Res. 2014, 145−146, 214−225. (12) Schwartz, J.; Neas, L. M. Fine particles are more strongly associated than coarse particles with acute respiratory health effects in schoolchildren. Epidemiology 2000, 11, 6−10. (13) Li, N.; Sioutas, C.; Cho, A.; Schmitz, D.; Misra, C.; Sempf, J.; Wang, M.; Oberley, T.; Froines, J.; Nel, A. Ultrafine particulate pollutants induce oxidative stress and mitochondrial damage. Environ. Health Persp. 2003, 111, 455−460. (14) Perrino, C. Atmospheric particulate matter, Proceedings of a C.I.S.B. Minisymposium, 2010. H

dx.doi.org/10.1021/es503696k | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

(33) Robinson, A. L.; Donahue, N. M.; Shrivastava, M. K.; Weitkamp, E. A.; Sage, A. M.; Grieshop, A. P.; Lane, T. E.; Pierce, J. R.; Pandis, S. N. Rethinking organic aerosols: semivolatile emissions and photochemical aging. Science 2007, 315, 1259−1262. (34) Wang, R.; Tao, S.; Ciais, P.; Shen, H. Z.; Huang, Y.; Chen, H.; Shen, G. F.; Wang, B.; Li, W.; Zhang, Y. Y.; Lu, Y.; Zhu, D.; Chen, Y. C.; Liu, X. P.; Wang, W. T.; Wang, X. L.; Liu, W. X.; Li, B. G.; Piao, S. L. High-resolution mapping of combustion processes and implications for CO2 emissions. Atmos. Chem. Phys. 2013, 13 (10), 5189−5203. (35) Ginoux, P.; Prospero, J. M.; Torres, O.; Chin, M. Long-term simulation of global dust distribution with the GOCART model: correlation with North Atlantic Oscillation. Environ. Modell. Softw. 2004, 19, 113−128. (36) Mahowald, N.; Lamarque, J. F.; Tie, X.; Wolff, E. Sea salt aerosol response to climate change: last glacial maximum, preindustrial and doubled carbon dioxide climates. J. Geophys. Res. 2006, 111, D05303. (37) World Steel Association. http://www.worldsteel.org (accessed Mar 8, 2012). (38) U.S. Geological Survey. (USGS) Commodity Statistics and Information. http://minerals.usgs.gov/minerals/pubs/commodity (accessed Mar 8, 2012). (39) International Fertilizer Industry Association (IFA); http://www. fertilizer.org/ (accessed Mar 28, 2013). (40) Shen, H. Z.; Huang, Y.; Wang, R.; Zhu, D.; Li, W.; Shen, G. F.; Wang, B.; Zhang, Y. Y.; Chen, Y. C.; Lu, Y.; Chen, H.; Li, T. C.; Sun, K.; Li, B. G.; Liu, W. X.; Liu, J. F.; Tao, S. Global Atmospheric Emissions of Polycyclic Aromatic Hydrocarbons from 1960 to 2008 and Future Predictions. Environ. Sci. Technol. 2013, 47 (12), 6415− 6424. (41) Wang, R.; Tao, S.; Wang, W. T.; Liu, J. F.; Shen, H. Z.; Shen, G. F.; Wang, B.; Liu, X. P.; Li, W.; Huang, Y.; Zhang, Y. Y.; Lu, Y.; Chen, H.; Chen, Y. C.; Wang, C.; Zhu, D.; Wang, X. L.; Li, B. G.; Liu, W. X.; Ma, J. M. Black Carbon Emissions in China from 1949 to 2050. Environ. Sci. Technol. 2012, 46 (14), 7595−7603. (42) Shen, H. Z.; Tao, S.; Wang, R.; Wang, B.; Shen, G. F.; Li, W.; Su, S. S.; Huang, Y.; Wang, X. L.; Liu, W. X.; Li, B. G.; Sun, K. Global time trends in PAH emissions from motor vehicles. Atmos. Environ. 2011, 45 (12), 2067−2073. (43) Wang, R.; Tao, S.; Shen, H. Z.; Wang, X. L.; Li, B. G.; Shen, G. F.; Wang, B.; Li, W.; Liu, X. P.; Huang, Y.; Zhang, Y. Y.; Lu, Y.; Ouyang, H. L. Global emission of black carbon from motor vehicles from 1960 to 2006. Environ. Sci. technol. 2012, 46 (2), 1278−1284. (44) Sánchez-Ccoyllo, O. R.; Ynoue, R. Y.; Martins, L. D.; Astolfo, R.; Miranda, R. M.; Freitas, E. D.; Borges, A. S.; Fornaro, A.; Freitas, H.; Moreira, A.; Andrade, M. F. Vehicular particulate matter emissions in road tunnels in Sao Paulo, Brazil. Environ. Monit. Assess. 2009, 149, 241−249. (45) State Environmental Protection Administration of China. Handbook of Industrial Pollution Emission Rates; China Environmental Science Press: Beijing, China, 1996. (46) Yi, H. H.; Hao, J. M.; Duan, L.; Tang, X. L.; Ning, P.; Li, X. H. Fine particle and trace element emissions from an anthracite coal-fired power plant equipped with a bag-house in China. Fuel 2008, 87, 2050−2057. (47) Chen, Y. J.; Sheng, G. Y.; Bi, X. H.; Feng, Y. L.; Mai, B. X.; Fu, J. M. Emission factors for carbonaceous particles and polycyclic aromatic hydrocarbons from residential coal combustion in China. Environ. Sci. Technol. 2005, 39 (6), 1861−1867. (48) Oanh, N. T. K.; Albina, D. O.; Ping, L.; Wang, X. K. Emission of particulate matter and polycyclic aromatic hydrocarbons from select cookstove-fuel systems in Asia. Biomass Bioenergy 2005, 28 (6), 579− 590. (49) Tewalt, S. J.; Belkin, H. E.; SanFilipo, J. R.; Merrill, M. D.; Palmer, C. A.; Warwick, P. D.; Karlsen, A. W.; Finkelman, R. B.; Park, A. J., comp. Chemical analyses in the World Coal Quality Inventory, version 1: U.S. Geological Survey Open-File Report 2010-1196, 2010; http://pubs.usgs.gov/of/2010/1196/pdf/of2010-1196.pdf.

(50) Liu, Y.; Zhang, Y. X.; Wei, Y. J.; Dou, H.; Gu, D. S.; Zeng, L. M.; Shao, M. Measurement of emission factors of carbonaceous aerosols from residential coal combustion. Acta Scien. Circumst. 2007, 27 (9), 1409−1416. (51) Zhi, G. R.; Chen, Y. J.; Feng, Y. L.; Xiong, S. C.; Li, J.; Zhang, G.; Sheng, G. Y.; Fu, J. Emission characteristics of carbonaceous particles from various residential coal-stoves in China. Environ. Sci. technol. 2008, 42 (9), 3310−3315. (52) Zhi, G. R.; Peng, C. H.; Chen, Y. J.; Liu, D. Y.; Sheng, G. Y.; Fu, J. M. Deployment of coal briquettes and improved stoves: Possibly an option for both environment and climate. Environ. Sci. Technol. 2009, 43 (15), 5586−5591. (53) United Nations Industrial Development Organization (UNIDO). International Yearbook of Industrial Statistics 2008; Edward Elgar Publishing: Cheltenham, U.K., 2008. (54) Worrell, E.; Price, L.; Martin, N.; Hendriks, C.; Meida, L. O. Carbon dioxide emissions from the global cement industry. Annu. Rev. Energy Env. 2001, 26, 303−329. (55) Monthly report on the operation of SO2 scrubbers at coal power plants; Jiangsu Department of Environmental Protection: Nanjing, China, 2007−2009. (56) Cementchina.net; http://www.cementchina.net/ (accessed Mar 8, 2012). (57) Wang, R.; Tao, S.; Balkanski, Y.; Ciais, P.; Boucher, O.; Liu, J. F.; Piao, S. L.; Shen, H. Z.; Vuolo, M. R.; Valari, M.; Chen, H.; Chen, Y. C.; Cozic, A.; Huang, Y.; Li, B. G.; Li, W.; Shen, G. F.; Wang, B.; Zhang, Y. Y. Exposure to ambient black carbon derived from a unique inventory and high-resolution model. Proc. Natl. Acad. Sci. U.S.A. 2014, 111 (7), 2459−2463. (58) China Health and Nutrition Survey (CHNS); http://www.cpc. unc.edu/projects/china. (59) Xu, Y. A. Improvements in the operation of SO2 scrubbers in China’s coal power plants. Environ. Sci. Technol. 2011, 45 (2), 380− 385. (60) International Energy Agency (IEA). IEA World Energy Statistics and Balances; http://www.oecd-ilibrary.org/statistics (accessed May 25, 2012). (61) U. S. Energy Information Administration. Form EIA-767 and EIA-860; http://www.eia.gov/electricity/data/detail-data.html. (62) National Bureau of Statistics and National Energy Administration. China Energy Statistical Yearbook, 1989−2008; China Statistics Press: Beijing, 2009. (63) The World Bank and the Development Research Center of China’s State Council. Urban China: Toward Efficient, Inclusive and Sustainable Urbanization. 2014; http://www.worldbank.org/en/ country/china/publication/urban-china-toward-efficient-inclusivesustainable-urbanization. (64) U. S. Environmental Protection Agency. History of the Clean Air Act; 2010; http://www.epa.gov/air/caa/requirements.html. (65) Domenici, P. Clean Air Act Amendments of 1977. Nat. Res. J. 1979, 19, 475. (66) United Nations Statistics Division. United Nations Energy Statistics, 1950−1995; United Nations: New York, 1995. (67) He, K. B.; Huo, H.; Zhang, Q.; He, D. Q.; An, F.; Wang, M.; Walsh, M. P. Oil consumption and CO2 emissions in China’s road transport: current status, future trends, and policy implications. Energy Policy 2005, 33 (12), 1499−1507. (68) State Statistical Bureau of China. Statistical Yearbook of China; China Statistics Press: Beijing, 2009. (69) China National Radio.http://www.cnr.cn/gundong/201201/ t20120125_509092713.shtml (accessed Oct 11, 2013). (70) China Securities Journal. http://www.cs.com.cn/xwzx/05/ 200910/t20091027_2244716.htm (accessed Oct 11, 2013). (71) Guangzhou China. http://www.guangzhou.gov.cn/node_2190/ node_2207/2010/06/12/1276310797336552.shtml (accessed Oct 11, 2013). (72) Limits and measurement methods for exhaust pollutants from compression ignition and gas fuelled positive ignition engines of I

dx.doi.org/10.1021/es503696k | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

vehicles (III, IV, V). GB17691-2005; http://www.zhb.gov.cn/ image20010518/3882.pdf (accessed Oct 11, 2013). (73) Limits and measurement methods for emissions from light-duty vehicles (III, IV). GB18352.3-2005; http://kjs.mep.gov.cn/hjbhbz/ bzwb/dqhjbh/dqydywrwpfbz/200707/W020110127402088768173. pdf (accessed Oct 11, 2013). (74) District Heating and Cooling: Environmental Technology for the 21st Century; International Energy Agency: Paris, 2002. (75) World Bank list of economies. http://data.worldbank.org/ about/country-classifications/country-and-lending-groups (accessed Jul 19, 2014). (76) Pope, C. A.; Dockery, D. W. Health effects of fine particulate air pollution: Lines that connect. J. Air Waste Manage. Assoc. 2006, 56 (6), 709−742. (77) Oberdorster, G.; Oberdorster, E.; Oberdorster, J. Nanotoxicology: An emerging discipline evolving from studies of ultrafine particles. Environ. Health Perspect. 2005, 113 (7), 823−839. (78) Chen, R. J.; Li, Y.; Ma, Y. J.; Pan, G. W.; Zeng, G.; Xu, X. H.; Chen, B. H.; Kan, H. D. Coarse particles and mortality in three Chinese cities: The China Air Pollution and Health Effects Study (CAPES). Sci. Total Environ. 2011, 409 (23), 4934−4938. (79) van der Werf, G. R.; Randerson, J. T.; Giglio, L.; Collatz, G. J.; Mu, M.; Kasibhatla, P. S.; Morton, D. C.; DeFries, R. S.; Jin, Y.; van Leeuwen, T. T. Global fire emissions and the contribution of deforestation, savanna, forest, agricultural, and peat fires (1997−2009). Atmos. Chem. Phys. 2010, 10 (23), 11707−11735.

J

dx.doi.org/10.1021/es503696k | Environ. Sci. Technol. XXXX, XXX, XXX−XXX