Emission trends and mitigation options for air pollutants in East Asia

Abstract. Emissions of air pollutants in East Asia play an important role in the regional and global atmospheric environment. In this study we evaluated the recent emission trends of sulfur dioxide (SO2), nitrogen oxides (NOx), particulate matter (PM), and non-methane volatile organic compounds (NMVOC) in East Asia, and projected their future emissions up until 2030 with six emission scenarios. The results will provide future emission projections for the modeling community of the model inter-comparison program for Asia (MICS-Asia). During 2005–2010, the emissions of SO2 and PM2.5 in East Asia decreased by 15 and 12%, respectively, mainly attributable to the large-scale deployment of flue gas desulfurization (FGD) at China's power plants, and the promotion of highly efficient PM removal technologies in China's power plants and cement industry. During this period, the emissions of NOx and NMVOC increased by 25 and 15%, driven by rapid increase in the emissions from China due to inadequate control strategies. In contrast, the NOx and NMVOC emissions in East Asia except China decreased by 13–17%, mainly due to the implementation of stringent vehicle emission standards in Japan and South Korea. Under current regulations and current levels of implementation, NOx, SO2, and NMVOC emissions in East Asia are projected to increase by about one-quarter over 2010 levels by 2030, while PM2.5 emissions are expected to decrease by 7%. Assuming enforcement of new energy-saving policies, emissions of NOx, SO2, PM2.5 and NMVOC in East Asia are expected to decrease by 28, 36, 28, and 15%, respectively, compared with the baseline case. The implementation of "progressive" end-of-pipe control measures would lead to another one-third reduction of the baseline emissions of NOx, and about one-quarter reduction of SO2, PM2.5, and NMVOC. Assuming the full application of technically feasible energy-saving policies and end-of-pipe control technologies, the emissions of NOx, SO2, and PM2.5 in East Asia would account for only about one-quarter, and NMVOC for one-third, of the levels of the baseline projection. Compared with previous projections, this study projects larger reductions in NOx and SO2 emissions by considering aggressive governmental plans and standards scheduled to be implemented in the next decade, and quantifies the significant effects of detailed progressive control measures on NMVOC emissions up until 2030.


2000
; Klimont et al., 2001Klimont et al., , 2009Cofala et al., 2007Cofala et al., , 2012Xing et al., 2011;Zhao et al., 2013c). However, most of these projections were based on the emissions for the year 2005 or earlier and did not consider the dramatic recent changes. Latest projections include Cofala et al. (2012) and Zhao et al. (2013c). Cofala et al. (2012), one of the latest projections, projected global emissions of SO 2 , NO x 15 and PM 2.5 for four energy scenarios developed by IEA (2012a), but did not envisage further end-of-pipe mitigation measures in the future. Zhao et al. (2013c) developed six NO x emission scenarios up to 2030 based on the 2010 emission inventory, and quantified the effects of various control policies, but did not include the analysis of other air pollutants. 20 Although there have been a number of studies on recent and future emission trends in East Asia, they are proved inadequate when serving for the development of broadly effective air quality and climate policies. Firstly, future control measures must be developed by taking full account into the latest policies. Hence a comprehensive review of recent mitigation measures in the entire region is important but has not been presented.
not provide full insight into the future trends of major air pollutants. Secondly, the attainment of stringent ambient air quality standard (e.g. China's standard of 35 µg m −3 for annual average PM 2.5 concentration, released in 2012) requires simultaneous reduction of multiple pollutants including SO 2 , NO x , PM 2.5 , and non-methane volatile organic compounds (NMVOC) to a large extent (Wang and Hao, 2012). Therefore, it is essential that a full range of relevant pollutants is considered, and scenarios at different stringency levels from the business-as-usual case to the maximum feasible reduction case are developed, so that cost-effective emission controls can balance measures over all pollutants and over a wide range of stringency levels. Thirdly, most studies focused on either end-of-pipe control measures, or energy saving measures; their roles in inte-15 grated control policies tackling multiple pollutants and global warming simultaneously have been insufficiently studied. Considering the above, a comprehensive projection of multiple pollutants' emissions incorporating up-to-date base-year data, control measures scheduled to be implemented, and other potential energy saving and end-of-pipe measures at different stringency levels, will contribute to both air pollution research and Japanese government set a long-term target to improve energy intensity of GDP by an additional 30 % by 2030 (IEA, 2008). The government of South Korea has made a commitment to reduce its GHG emissions by 30 % compared to its business as usual projection by 2020 (IEA, 2012b). Chinese government has set a target to reduce CO 2 emissions per unit GDP by 40-45 % in 2020 compared with the 2005 levels (Wang and 10 Hao, 2012). Total energy consumption in East Asia increased by 31 % during [2005][2006][2007][2008][2009][2010]. China experienced the fastest increase of 43 % driven by rapid GDP growth rate, while Japan's energy consumption decreased during the five years because of lower GDP growth rate and stringent energy saving policies. The growth rate of South Korea is medium (19 %).

Power plants
The energy consumption in China's power sector increased remarkably by 35 % during 2005-2010 due to rapid increase in the demand of electricity (NBS, 2007(NBS, , 2011a, while those of Japan and South Korea remained relatively stable (http://www.iea.org/ statistics/). Introduction

Tables Figures
Back Close

Full Screen / Esc
Printer-friendly Version

Interactive Discussion
Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | the share of nuclear power generation in Japan dropped dramatically to less than 10 %, owing to the Fukushima nuclear power plant accident on 11 March 2011 (http://www.iea.org/statistics/). This made the future of nuclear power in Japan quite uncertain. In South Korea, nuclear power generation is expected to keep increasing in the next decade, as five reactors are under construction and six more have been 5 announced (IEA, 2012b). Considering the coal-intensive power generation mix, Chinese government has also been promoting the development of clean energy power through subsidy policies. By 2010, the capacities of hydro power, natural gas power, wind power, and solar power have increased dramatically to 1.82, 2. 25, 23.8

Industrial sector
During 2005-2010, China's energy consumption of industrial sector increased dramatically at an annual average rate of 9.0 % (cf. 7.4 % for total energy consumption) due 25 largely to the rapid increase of energy-intensive products, e.g. cement and steel (NBS, 2007(NBS, , 2011a. However, with the target to reduce energy intensity per GDP by 20 % during 2005-2010, China put much effort to replace out-of-date production technologies 2608 first released in 1980 and strengthened in 1992 and 1999, were all voluntary. As of 2005, 30 % of newly-built houses and 85 % of buildings larger than 2 000 m 2 complied with the voluntary standards (IEA, 2008). In Korea, the building energy codes have been at a relatively low level for a long time, until a performance based strong building design code applied to large commercial buildings in 2011 (IEA, 2006(IEA, , 2012b. 15 Japan is a world leader for the energy efficiencies of appliances in residential and commercial buildings. "Top Runner program", which set energy efficiency targets for appliances based on the most energy-efficient products on the market, has been successfully enforced. For example, the efficiency of air conditioners and refrigerators increased by 68 % (1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004) and 55 % (1998)(1999)(2000)(2001)(2002)(2003)(2004), both exceeding the targets of 20 66 % and 31 % (IEA, 2008; Energy Conservation Center of Japan, 2011). Similar programs have been promoted in South Korea and China recently (UNEP, 2010).
Due to the coal-intensive energy structure, China has been promoting clean energy in residential sector. Direct combustion of biomass has been gradually replaced with commercial fuels in the last decade, and its share in rural cooking decreased from 38 %

Transportation sector
During 2005-2010, the energy consumption in China's transportation sector has been growing at an annual average rate of 10 % attributed to the explosive growth of vehicle population (NBS, 2007(NBS, , 2011a. In contrast, the transportation energy consumption in South Korea was stable and that of Japan has been dwindling (http://www.iea.org/ 5 statistics/). The decline in Japan's vehicle energy consumption is largely due to its fuel efficiency standards, which are among the most aggressive ones in the world. For passenger vehicles, there has been a consistent improvement in the average fuel economy from 13.5 km L Korea has been systematically promoting compressed natural gas (CNG) buses since 2000. As of 2008, 19 thousand intra-city buses and 429 garbage trucks have utilized CNG. China has also launched several initiatives to promote electric vehicles, and their population reached 12 000 by 2010 (Yang, 2012). The most recent develop-25 ment plan for new energy vehicles (issued in 2012) aimed to increase the population of electric vehicles to 0.5 million and 5 million in 2015 and 2020 respectively through a series of subsidy policies.

2611
In 2006, China set targets to reduce the national SO 2 emissions by 10 % by 2010 (Wang and Hao, 2012). By the year 2010, over 83 % of coal-fired power plants (about 88 % of pulverized coal combustion plants) or up to 560 gigawatts (GW) installed flue gas desulfurization (FGD) (MEP, 2011). The recently released 12th Five-Year Plan aims at another 8 % reduction in total SO 2 emissions, which requires nearly all coal- 10 fired power plants to be equipped with high efficiency FGD facilities (95 % removal efficiency).
Low NO x combustion technology (mainly Low NO x Burner, LNB) was the major NO x control technology in China's coal-fired power plants by 2010. The penetration of flue gas denitrification (Selective Catalytic Reduction, SCR; and Selective Non-Catalytic 15 Reduction, SNCR) was only 1.1 % in 2005and 12.8 % in 2010(MEP, 2011. In the 12th Five-Year Plan, Chinese government aims to reduce the national 2010 NO x emissions by 10 % by the year 2015, and the key measures to fulfill this target is large scale deployment of SCR/SNCR facilities. The NO x emission control policies are described in more details in our previous paper (Zhao et al., 2013c). 20 The emission control of particulate matter in China's power sector has achieved noticeable progress in the last decade. Since 2003, all new and rebuilt units have to attain the PM in-stack concentration standard of 50 mg m −3 (GB13223-2003). As a result, over 92 % of pulverized coal units installed electrostatic precipitators (ESP) by 2005. In addition, fabric filters (FF) have been put into commercial use in the past five years, and 25 its penetration increased to 7 % by 2010 (Zhao et al., 2013a). Furthermore, the rapid deployment of wet-FGD also helped to reduce PM emissions owing to its ancillary benefit on PM removal (Zhao et al., 2010) In Japan, application of best available technologies to control SO 2 , NO x , and PM is required for most power generation units across the country. The penetrations of wet-FGD, LNB+SCR and high efficiency deduster (HED, e.g., FF, and electrostatic-5 fabric integrated precipitator) are all as high as 90-100 %, and increased slightly during 2005(Klimont et al., 2009. In South Korea, FGD systems have been installed for most power generation units; the penetration increased slightly from 95 % to 97 % during 2005-2010. For NO x , SCR has been the dominant control technology, with its share increasing from 56 % 10 in 2005 to 68 % in 2010. About one third of coal-fired power generation units have been equipped with HED by 2010, and the rest was equipped with ESP (NIER, 2010(NIER, , 2013; Clean Air Policy Supporting System, CAPSS, http://airemiss.nier.go.kr/).

Industrial sector
The penetrations of control technologies for industrial boilers and industrial processes 15 are presented in Table 3, Table 4 and Table S1, respectively.
In China, SO 2 and NO x control technologies have been rarely installed in the industry sector. In recent years, FGD units to control SO 2 have been installed for small number of coal-fired boilers and sintering plants in selected regions. The application of NO x control technologies is described in more details in our previous paper (Zhao 20 et al., 2013c). In contrast with SO 2 and NO x , China has been controlling PM emissions from industrial sources since late 1980s. However, the emission standards for industrial sources have been updated slowly before 2010 (see details in Lei et al., 2011). The 11th Five-Year Plan requested to promote high efficiency FF in some high-emission industries. Most industrial boilers were historically equipped with wet scrubbers (WET) and cyclone dust collectors (CYC), while high efficiency FF began to penetrate recently (Lei et al., 2011;Zhao et al., 2013a WET (Lei et al., 2011;Zhao et al., 2013a).

5
The only control measures for NMVOC emissions in China's industry sector are associated with fossil fuel exploitation and distribution. Emission standards for gasoline distribution released in 2007 requires: (1) installation of vapor recovery systems and modified loading techniques (Stage IA control) for loading and unloading operations; (2) improvement in the service station tank (Stage IB control) and installation of vapor 10 balancing system between a vehicle and service station tank (Stage II control); (3) installation of internal floating covers (IFC) or secondary seals for new-built or retrofitted storage tanks. These standards were scheduled to be implemented in relatively large cities of "key regions" from 2008-2010 onwards, and in relatively large cities in other provinces from 2012-2015 onwards. We estimated vapor recycling systems have been 15 installed for about 15 % of all the gasoline storage and distribution operations by 2010 (see Table 4 for details).
In Japan, industrial emissions are limited strictly by the Air Pollution Control Act. The thresholds changed very slightly since 1995, but they are still among the most stringent in the world (Ministry of the Environment of Japan, 2013 Emission standards for industrial sources in South Korea are generally less stringent than those of Japan and more stringent than those of China (Ministry of Environment of South Korea, 2013). In contrast with Japan, the control measure portfolio for cement kilns is an equal mix of ESP and HED; ESPs still dominated the PM removal technologies for industrial boilers and sintering machines, and HEDs have not been widely applied. FGD system was widely applied for some high-emitting sources like industrial boilers and sintering plants, while penetrations of 85 % and 100 % respectively by 2010 (NIER, 2010(NIER, , 2013. Similar to Japan, dominant control measures for NO x emissions 10 are low NO x combustion technologies by 2010.

Residential sector
There are only limited legislations addressing residential sources. In Japan, about half of residential and commercial boilers are equipped with HED, driven by the stringent regulation of local government. In South Korea and China, dominant control technolo-15 gies are CYC and WET (Table 3).
Compared with boilers, emissions from small stoves are more difficult to control. In Japan, small incinerators dwindled rapidly in the last decade due to a regulation (released in 2000) with the purpose of mitigating dioxin pollution (Ministry of the Environment of Japan, 2013; Wakamatsu et al., 2013). Previous research found briquette 20 stoves have lower emission factors for SO 2 and PM (Lei et al., 2011). We estimate briquette accounted for 6-7 % of total residential coal consumption in China during 2005(NBS, 2007, 2008a, b, 2009, 2011a. Emissions from small stoves could be further reduced by switching to a newer type of installation, e.g., installing a catalyst or non-catalyst insert, using primary and secondary air deflectors, etc. Such kinds of Introduction   Table 6.

Non-energy related sectors
Chinese government have released standards to limit the solvent content in some prod-15 ucts, including wood paint, interior wall paints, adhesives for shoes production, decorative adhesives, and printing inks. Driven by these standards, the solvent contents of some products decreased, and the penetration of low-solvent products increased during 2005-2010. Tables 5 and S2 show the penetrations of major control measures for solvent use; Table S3 shows the changes in the emission factors of typical sources was estimated that the actually achieved reductions were higher, but the O 3 and PM concentrations have not declined as expected (Wakamatsu et al., 2013)

Effect of control measures on recent emission trends
The historical emissions of China are estimated with a model structure developed in our previous paper (Zhao et al., 2013c). The emissions from each sector in each province were calculated from the activity data (energy consumption, industrial products, etc.), technology-based uncontrolled emission factors, and penetrations of control technologies. The data sources for China are described in our previous paper (Zhao et al., 2013b).
The historical emissions of Japan are consistent with the JATOP Emission Inventory-Data Base (JEI-DB), developed by Japan Petroleum Energy Center (JPEC) (JPEC, 2012a-c). Special attention was paid to the road vehicle emissions. The basic con-15 cept of estimation is multiplying the traffic volume (considering vehicle type mix ratio by regulations), and the emission factor. JPEC adjusts that value with correction factors: a deterioration correction factor depending on accumulated mileage, a temperature correction factor and a humidity correction factor. It also includes original research data including start emission factor, evaporative emission factors, high emission vehi-20 cle ratio, and vehicle usage profile from a questionnaire survey (JPEC, 2012c). The emissions from other sources was calculated using local statistical information and emission factors, and allocated to area (1 km × 1 km) and time (hourly) (JPEC, 2012a).
The historical emissions of South Korea were calculated by the National Institute of Environmental Research (NIER) in South Korea, and the data sources are described 25 in some research reports and a web-based database (NIER, 2010(NIER, , 2013 The emission for those stacks have been estimated using CEMSs for years 2007-2010 (pre-2007, the emission factor method). There are "not-so-small" emission gaps before and after 2007 due to this methodological change. We, therefore, substituted pre-2007 emissions for those stacks using extrapolation of 2007-2010 "CEMS-based" estimation considering the changes of control measures.

5
The emissions for North Korea, Mongolia, Hong Kong & Macao, and Taiwan are adopted from the Gains-Asia model of IIASA (http://gains.iiasa.ac.at/models/). National air pollutant emissions in East Asia are summarized in Table 8. The sectoral emissions in China are given in Fig. 2, and those in Japan and South Korea are shown in Fig. 3.  The decline in China's SO 2 emissions is mainly attributable to the large scale deployment of FGD for power plants. In comparison, SO 2 emissions from China's industrial sector kept increasing during this period, slowing down the declining rate of total SO 2 emissions; this is consistent with the recent estimates by Zhang et al. (2012b), Lu et al. (2011), Klimont et al. (2013. 10 SO 2 emissions of Japan decreased by 20 %, mainly attributed to the increasing penetration of high-efficient desulfurization technologies in the industrial sector, and the replacement of coal and oil with clean and renewable energy. South Korea's SO 2 emissions roughly remained constant, because the reduction of the emissions from power plants (owing to the deployment of FGDs) was offset by the increasing emissions from 15 industrial sources.

PM 10 and PM 2.5
In 2010, the total PM 10 and PM 2.5 emissions in East Asia were 15.8 Mt and 11.8 Mt respectively. During 2005-2010, the PM 10 and PM 2.5 emissions decreased by 15 % and 11 % respectively. This trend was also donimated by the trend in emissions from 20 China, as China's PM 10 /PM 2.5 emissions represent 94 % of those of East Asia.
China's PM 10 and PM 2.5 emissions decreased by 15 % and 12 % respectively during the five years. We estimate that emissions of power plants and cement industry experienced the fastest decrease (43-47 % reduction from 2005-2010), as a result of the rapid evolution of end-of-pipe removal equipments (see Tables 2 and 4). The emissions 25 of industrial combustion and steel industry increased by 14-32 %, while the emissions of other sectors kept relatively stable. 14,2014 Emission trends and mitigation options for air pollutants in East Asia Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | PM 10 and PM 2.5 emissions decreased by 7 % and 8 % in East Asia except China. The declining rate is as large as 19-28 % in Japan, and transportation sector contributes 70 % of this decline. Emissions from South Korea increased somewhat due to the increase in industrial fuel consumption, which is further a result of the relatively stable energy intensity of industrial sector (see Sect. 2.1.2).  In China, the NMVOC emissions from transportation and residential combustion decreased owing to improving vehicle emission standards and the replacement of biomass with cleaner energy sources. However, these reductions were offset by the dramatic increase of emissions from industrial process (+46 %) and solvent use (+102 %), leading to a 21 % increase of the total NMVOC emissions.

15
Japan's NMVOC emissions decreased by 30 % mainly because of the implementation of stringent vehicle emission standards. In South Korea, although the enhancement of vehicle emission standards lowered NMVOC emissions from transportations, the emissions from solvent use increased even more rapidly, leading to a 15 % increase in total NMVOC emissions.

Future emission scenarios for air pollutants
To quantify the effects of various measures on future air pollutant emissions, in this study we developed emission scenarios for SO 2 , NO x , PM, and NMVOC based on the energy saving policies and end-of-pipe control strategies. The scenarios are developed with the same model structure as that for the estimation of historical emissions ACPD 14,2014 Emission trends and mitigation options for air pollutants in East Asia We developed two energy scenarios, a business as usual scenario (BAU) and an alternative policy scenario (PC). The BAU scenario is based on current legislations and implementation status (until the end of 2010). In the PC scenario, we assume the introduction and strict enforcement of new energy-saving policies, including life style changes, structural adjustment, and energy efficiency improvement. Life style changes imply slower growth of energy service demand, including energy-intensive industrial products, building area and residential service demand, vehicle population, electricity production, and heat supply, due to more conservative life styles. Structural adjustment includes promotion of clean and renewable fuels and energy-efficient technologies, 15 such as renewable energy power and CHP for power plants and heat supply sector respectively, arc furnace and large precalcined kilns for industrial sector, biogas stoves and heat pumps for residential sector, electric vehicles and bio-fuel vehicles for transportation sector, etc. Assumed energy efficiency improvement includes the improvement of the energy efficiencies of single technologies in each sector. 20 We developed three end-of-pipe control strategies for each energy scenario, including baseline (abbr.  Table 1. In this paper we focus on the development of energy scenarios and emission scenarios for China. The scenarios for other countries are adapted from those developed by IIASA in a project funded by United Nations Environment Programme (UNEP) and World Meteorological Organization (WMO) (UNEP and WMO, 2011). Both the energy consumption and air pollutant emissions were calculated with a 5-year resolution, though the parameters and results are presented for selected years only. Detailed assumptions of the energy scenarios and emission scenarios are documented below. The total electricity production is projected to be 10-12 % lower in PC scenarios that of BAU senario. The PC scenario considers aggressive development plans for clean and renewable energy power generation, therefore, the proportion of electricity production from coal-fired power plants is expected to decrease to 57 % in 2030 in PC scenarios, contrasted by 73 % in BAU scenario. 5 We projected lower yields of energy-intensive industrial products in PC scenario than those of BAU scenario because of a more conservative life style. The shares of less energy-intensive technologies are assumed to be higher in PC scenario than BAU scenario.
For residential sector, China's building area per capita in PC scenario is expected 10 to be 3-4 m 2 lower than that of BAU scenario in both urban and rural area. The energy demand for heating per unit area is somewhat lower in our PC scenario because of implementation of new energy conservation standards for the design of buildings. Replacement of coal and direct biomass burning with clean fuels are assumed in both urban and rural areas, with faster progress in the PC scenario.

15
The vehicle population per 1000 persons is projected at 380 and 325 in BAU and PC scenarios, respectively. The PC scenario assumes an aggressive plan to promote electric vehicles, and a progressive implementation of new fuel efficiency standards, resulting in 33 % and 57 % improvement in the fuel economy of new passenger cars and new heavy duty vehicles by 2030. 20 Table 7 shows current and future energy consumption in East Asia. Total energy consumption in East Asia was 123 EJ in 2005 and 161 EJ in 2010. The energy consumption of China accounts for 69-76 % of the total energy amount during 2005-2010, followed by 13-18 % for Japan, and about 7 % for South Korea. By 2030, the total energy consumption is projected to increase to 243 EJ under the BAU scenario and to 25 195 EJ under the PC scenario, 51 % and 21 % larger than that of 2010.
Of all the countries, China is expected to experience the fastest growth rate in energy consumption. By 2030, China's energy consumption is projected to increase by 64 % and 27 % from the 2010 level in BAU and PC senarios, respectively. Industry fuel ACPD 14,2014

Emission trends and mitigation options for air pollutants in East Asia
S. X. Wang et al. consumption is expected to increase notably slower than the total fuel use in both scenarios resulting from the economic structure adjustment. In contrast, the energy consumption of transportation is projected to increase dramatically by 200 % and 101 % in BAU and PC scenarios respectively, measured in 2030 against the 2010 levels, driven by the swift increase in vehicle population. The growth rate of energy consumption 5 in other sectors is close to that of the total amount. Because of the energy saving measures, the energy consumption of power plants, industry, residential, and transportation sectors in PC scenario are 18 %, 19 %, 27 %, and 33 % lower than the BAU scenario, respectively. Coal continues to dominate China's energy mix, but the proportion decreases from 68 % in 2010 to 60 % and 52 % in 2030 under the BAU and PC 10 scenarios, respectively. In contrast, the shares of natural gas and "other renewable energy and nuclear energy" are estimated to increase from 3.4 % and 7.5 % in 2010 to 5.5 % and 8.9 % in 2030 under the BAU scenario, and 9.3 % and 15.8 % under the PC scenario, respectively. By 2030, the energy consumption of East Asia except China is projected to increase 15 slightly by 12 % and 2 % from the 2010 level in BAU and PC scenarios, respectively. Japan and South Korea are two major energy consumers except China. Under current policies, Japan's energy consumption are projected to increase very slightly by 2 % from 2010 to 2030, because of slow economic growth rate and a tendency towards higher energy efficiency resulting from current legislaion. Due to the implementation of 20 low carbon policies intended to limit CO 2 concentrations to 450 ppm, Japan's energy consumption would be reduced by 6 % by 2030 from the 2010 level. This reduction is mainly attributed to the decline in energy consumption of transportation sector, resulting from improved fuel economy and reduced mileage travelled. to 12 % and 29 % under PC scenario. In contrast, the proportion of renewable energy would increase from 16 % in 2010 to 23 % and 33 % in 2030 under BAU and PC scenarios, respectively.  More ambitious application of measures is required in the "key regions". For PM, HED 15 would spread much more rapidly, with its share approaching 35 % and 50 % in 2020 and 2030, respectively. In the BAU[2]/PC[2] scenario, the best available technologies, i.e. FGD, LNB+SCR, and HED for PM, are assumed to be fully applied by 2030. Table 2 gives the national average penetration of control technologies. Note that the penetrations in the "key regions" are usually larger than those of other regions.

Industrial sector
The latest national emission standard for industrial boilers was released in 2001 (GB13271-2001 The emissions from "industry process" were mainly regulated by the "emission standard for industrial kiln and furnace" before 2010. Standards for specific industry were only issued for cement plants ( GB4915-2004), and coking oven (GB16171-1996). However, new emission standards for various industries were issued explosively during 2010-2012, which might significantly alter the emission pathways in the future. A series of new emission standards for iron and steel industry were released in 2012, 15 including the standards for sintering, iron production, steel production, steel rolling and so on. Sintering machine acts as the main source for SO 2 and NO x emissions in iron and steel industry, and also an important source for PM emissions. Wet-FGDs are required to be installed in order to attain the SO 2 concentration threshold, and the 12th Five-Year Plan also requires large-scale deployment of FGD. absorb most SO 2 produced due to its basic nature, even the strenghthend SO 2 limit could be attained under good technical conditions. The attainment of the NO x limit calls for the update with low NO x combustion technology (or installation of SNCR if the update is not applicable) for existing kilns, and simutaneous utilization of low NO x combustion technology and SNCR/SCR for new kilns. To attain the new emission standard for the nitric acid industry (GB26131-2010), the dual-pressure process should be equipped with absorption method (ABSP) or SCR, while other processes need to adopt both ABSP and SCR. and Stage IB+Stage II control measures in gasoline storage and distribution would approach 75 % and 100 % by 2020 and 2030, respectively. The application rate in crude oil distribution would be 25 % and 50 % by 2020 and 2030, respectively (see Table 4). on specific industry) will be released and implemented in key provinces as of 2015, and in other provinces as of 2020. Afterwards, the emission standards will become more stringent gradually. In terms of technologies, 10 we would prefer basic management techniques (e.g., leakage detection and repair system for refinery, improved solvent management for paint production) when they are applicable. End-of-pipe measures (condensation, adsorption, absorption, incineration etc.) are adopted when high removal rate is required. The penetration of selected control measures assumed for key sources are summarized in Table 4.  timeline of the emission standards is given in Fig. 1. The removal efficiencies of the future emission standards are from the GAINS-Asia model of IIASA

Future emission trends and effects of control measures
The air pollutant emissions in each scenario are estimated based on the assumptions 5 in Sects. 3.1 and 3.2. Table 8 shows the national air pollutant emissions in East Asia under each scenario. Figure 2 shows the emissions by sector in China, and Fig. 3 shows the emissions by sector in Japan and South Korea.

NO x
Under current legislation and current implementation status, NO x emissions in East

SO 2
The SO 2 emissions in East Asia are predicted to have a 24 % growth from 2010 to 2030 if we stick to current legislation and current implementation status. The enforement of advanced energy-saving measures could lead to substantial reduction in SO 2 emissions (36 % reduction) from the baseline projection, exceeding the effect of pro-  We also note that the SO 2 emissions are projected to be 21.7 Mt in 2015 under the BAU[1] scenario, 11.1 % lower than that of 2010. This implies that if the control policies in the 12th Five-Year Plan could be implemented successfully (as assumed in the BAU[1] scenario), the national target to reduce the SO 2 emissions by 8 % during 2011-2015 would be achieved. 20 The SO 2 emissions in East Asia except China, and two major energy consumers therein (Japan and South Korea) are expected to stay relatively stable until 2030 under current legislations. The implementation of new energy saving polices could lead to 9-18 % reduction in SO 2 emissions from the levels of baseline projection. The reduction is mainly achieved thorough the promotion of nuclear and renewable power generation 25 and replacement with cleaner fuel type in the industrial sector. Under the maximum feasible reduction measures, the SO 2 emissions in East Asia except China, Japan and South Korea would be reduced to 33 %, 52 % and 39 % of the baseline projection, respectively.

Emission trends and mitigation options for air pollutants in East Asia
S. X. Wang et al. technologies, the PM 2.5 emissions in East Asia except China, Japan and South Korea would account for about one quarter, half and 40 % of the levels of the baseline projection.

NMVOC
Under current legislation and current implementation status, NMVOC emissions in East

5
Asia are projected to increase by 24 % by 2030 from the 2010 levels. The implementation of assumed energy saving measures and "progressive" end-of-pipe control measures are expected to reduce NO x emissions by 15 % and 23 % respectively from the baseline projection. Up to 62 % of the total NMVOC emissions are expected to remain even the assumed energy saving measures and progressive end-of-pipe control The emissions in East Asia except China are expected to increase by 5 % from 2010 to 2030 under current legislation. The growth rates in Japan and South Korea are 4 % and 9 % respectively. This slight upward trend is an integrated effect of the reduction in transportation emissions owing to increased share of low-emission vehicles, and the increase of emissions from solvent use owing to inadequate control policies. By 5 2030, solvent use contributes about 80 % of total NMVOC emissions in both Japan and Korea under the baseline projection. As solvent use has little to do with fuel consumption, the implementation of energy saving policies has very limited effects on the reduction of NMVOC emissions. In contrast, the full application of end-of-pipe control measures would reduce the emissions from solvent use dramatically, thereby reducing 10 about three quarters of the total NMVOC emissions from the baseline projection.

Comparison with other studies
In 2010, China contributes 88 %, 94 %, 94 %, 94 %, and 88 % of the total NO x , SO 2 , PM 10 , PM 2.5 , and NMVOC emissions in East Asia, respectively. As a developing coun-15 try, China has substantial potential to reduce air pollutant emissions with the implementation of aggressive control policies, and is therefore believed to dominate the emission trends of East Asia in the next 20 yr. In addition, many previous studies on emission projection have focused on China. Some Asian or global studies did incorporated Japan, South Korea, and other countries, but they seldom presented emissions 20 of these countries seperately, making it difficult to review their emission projections. Given the reasons above, we would just compared the previous studies with China's emission trends in this section.
The studies estimating historical emissions of China are numerous. Since this study focuses on the temporal trends, we included in our comparison only the studies which Introduction  Aardenne et al., 1999;Streets and Waldhoff, 2000;Klimont et al., 2001;Klimont et al., 2002) were based on the emissions in 1995 or before. They usually substantially underestimated the rapid economic growth during 2000-2010. In addition, none of them anticipated the aggressive control policies in China since 2005. Therefore, these projections have deviated greatly from the actual status. In this study, we just compared recent emission projections reported since 2005 (or using the base year of 2000 or later) with our projection, which is shown in Fig. 4.  nario was a least cost optimization scenario that would achieve the same health benefit as the advanced control technology scenario. Xing et al. (2011) projected NO x emissions for 2020 with four scenarios based on the emissions of 2005, including a scenario assuming current legislation and implementation status, a scenario assuming improvement of energy efficiencies and current environmental legislation, a scenario assuming improvement of energy efficiencies and better implementation of environmental legislation, and a scenario assuming improvement of energy efficiencies and strict environmental legislation. These two studies were actually conducted in cooperation. Similar to , their projections for 2010 were also significantly lower than our estimation. As for the growth rates until 2020 or 2030, all the scenarios in these scenarios, which projected the growth rates for the same period at 36 % and −3 %, respectively. This study predicted a stronger growth potential of China's energy consumption in the future, leading to larger growth rate or 20 smaller declining rate above.

SO 2 emissions
A number of studies have evaluated China's recent SO 2 emission trends (Klimont et al., 2013;Lu et al., 2011;Kurokawa et al., 2013). China has been implementing PM control policies for several decades. Therefore, 25 all of the projections reviewed here have more or less assumed PM control policies in the future. The PM 10 emissions growth rate until 2020 of the least aggressive scenario in Xing et al. (2011)  projects much lower emissions than any previously developed scenario. Kurokawa et al. (2013) have estimated the recent trends in China's NMVOC emissions, which showed a slightly stronger upward trend (16 % growth during 2005-2008) than this study (9 % growth for the same period).

15
Only three studies have projected China's NMVOC emissions since 2005. Compared with our study,  made similar estimation of NMVOC emissions in 2010, but predicted much higher growth rates for the period 2010-2020 in all its three scenarios, as  hardly assumed any effective control measures in these scenarios. Xing et al. (2011) andWei et al. (2011) have considered the effect of 20 recent vehicle emission standards on NMVOC emissions, and assumed pretty simple but progressively emerging control polices until 2020, and therefore achieved similar growth rates to ours for both baseline and progressive strategies. Given China is still in the starting stage of NMVOC emission controls, and new policies could only emerge slowly in the next 5-10 yr, so the emission trends should not deviate greatly from the The comparison of SO 2 VCDs derived by Lu et al. (2011) andFioletov et al. (2013) with the estimated SO 2 emissions are shown in Fig. 5a and b, respectively. It can be seen that the temporal trends of SO 2 VCD retrieved by Lu et al. (2011) Lu et al. (2011) in the following discussion. As shown in Fig. 5a operation status has been questioned before by both the government and research community (Xu et al., 2009). In response to this situation, Chinese government began to request the installation of continuous emission monitoring systems (CEMSs) together with FGDs since July 2007 (SEPA, 2007). Therefore, the average removal efficiency should have improved ever since, contributing to the rapid decline in SO 2 Under current legislation and current implementation status (BAU[0] scenario), NO x , SO 2 , and NMVOC emissions in East Asia are estimated to increase by about one quarter by 2030 from the 2010 levels, while PM 2.5 emissions are expected to decrease by 5 7 %. Assuming enforcement of new energy-saving policies, emissions of NO x , SO 2 , PM 2.5 and NMVOC in East Asia are expected to decrease by 28 %, 36 %, 28 %, and 15 % respectively compared with the baseline case. The implementation of the "progressive" end-of-pipe control measures is expected to lead to another one third reduction of the baseline emissions of NO x , and about one quarter reduction for SO 2 , PM 2.5 , 10 and NMVOC. Exploring the potential of currently known best available technologies, their full implementation could reduce the emissions of NO x , SO 2 , and PM 2.5 in East Asia to only about one quarter and NMVOC to one third of the levels of the baseline projection.
Comparison with emission projections in the literature indicates that this study (1) re-15 produces the recent emission trends until 2010; (2) projects larger reduction in NO x and SO 2 emissions by considering aggressive govermental plans and standards scheduled to be implemented in the next decade; (3) quantifies the significant effects of detailed progressive control measures on NMVOC emissions up to 2030; (4) quantifies the technologically feasible reduction potentials. The results of this study provide future emis-20 sion projections for the modeling community of the MICS-Asia program. The modelers could assess the impact of emission changes on future air quality with the projected emission trends in this study. In addition, the emission projections at various stringency levels from business-as-usual case to maximum feasible reduction case provide a basis for further studies on cost-effective emission control strategies, which can balance 25 control measures over all pollutants and over a wide range of stringency levels. and slowly strengthened control measures after 2015 (as assumed in the "progressive" end-of-pipe control strategy) could reduce China's emissions of NO x , SO 2 , and PM 2.5 significantly. The resulted NO x , SO 2 , and PM 2.5 emissions would be 16-26 % lower than the 2010 levels by 2020, and even lower by 2030, demonstrating a high mitigation potential when these legislations are enforeced efficiently. Therefore we believe it is 5 essential to support and monitor the progress of implementation of these legislations. Secondly, the contributions of advanced energy saving measures to the reduction of SO 2 and PM 2.5 emissions exceeds those of progressive end-of-pipe control measures by 2030. Since end-of-pipe control technologies, e.g., FGD facilities and high-efficient dedustors, have already been widely applied in typical sources in the base year, their 10 reduction potential would become smaller and smaller in the future. The energy saving measures would play an irreplacable role for further reduction of air pollutant emissions. Thirdly, control policies for NMVOC emissions are sadly lacked in China and South Korea at present, this study demonstrate that the simultaneous enforcement of energy saving measures and progressive end-of-pipe control measures (mainly assuming en-15 forement of European standards) could reduce 38 % of the total NMVOC emissions from the levels of baseline projection. Even though, large reduction potential still remains. Relative policies should be carefully optimized to reduce NMVOC emissions efficiently and effectively.  sions in China and India, 1996, Atmos. Chem. Phys., 11, 9839-9864, doi:10.5194/acp-11-9839-2011, 2011            (2013), in which a filtering procedure was applied to remove local biases, in particular 6 volcanic signals. (c) Average NO2 VCD and total NOX emissions in Eastern Central China. 7

Supplementary material related to this article is available online at
NO2 VCD was retrieved from OMI and SCIAMACHY in this study. 8

Fig. 5.
Inter-annual relative changes of SO 2 and NO 2 VCD from satellite observations and emission estimation in this study. All data are normalized to 2005. (a) Average SO 2 VCD and total SO 2 emissions in Eastern Central China (latitude < 45 • N, longitude > 100 • E). SO 2 VCD was derived by Lu et al. (2011). (b) Average SO 2 VCD and total SO 2 emissions over an area of Eastern China (34 • N-38 • N, 112 • E-118 • E). SO 2 VCD was derived by Fioletov et al. (2013), in which a filtering procedure was applied to remove local biases, in particular volcanic signals. (c) Average NO 2 VCD and total NO x emissions in Eastern Central China. NO 2 VCD was retrieved from OMI and SCIAMACHY in this study.