Atmospheric Chemistry and Physics Open Access Atmospheric Measurement Techniques

Abstract. We investigate projected 2000–2050 changes in concentrations of aerosols in China and the associated transboundary aerosol transport by using the chemical transport model GEOS-Chem driven by the Goddard Institute for Space Studies (GISS) general circulation model (GCM) 3 at 4° × 5° resolution. Future changes in climate and emissions projected by the IPCC A1B scenario are imposed separately and together through sensitivity simulations. Accounting for sulfate, nitrate, ammonium, black carbon (BC), and organic carbon (OC) aerosols, concentrations of individual aerosol species change by −1.5 to +0.8 μg m−3, and PM2.5 levels are projected to change by about 10–20% in eastern China as a result of 2000–2050 change in climate alone. With future changes in anthropogenic emissions alone, concentrations of sulfate, BC, and OC are simulated to decrease because of assumed reductions in emissions, and those of nitrate are predicted to increase because of higher NOx emissions combined with decreases in sulfate. The net result is a predicted reduction of seasonal mean PM2.5 concentrations in eastern China by 1–8 μg m−3 (or 10–40%) over 2000–2050. It is noted that current emission inventories for BC and OC over China are judged to be inadequate at present. Transboundary fluxes of different aerosol species show different sensitivities to future changes in climate and emissions. The annual outflow of PM2.5 from eastern China to the western Pacific is estimated to change by −7.0%, −0.7%, and −9.0% over 2000–2050 owing to climate change alone, changes in emissions alone, and changes in both climate and emissions, respectively. The fluxes of nitrate and ammonium aerosols from Europe and Central Asia into western China increase over 2000–2050 in response to projected changes in emissions, leading to a 10.5% increase in annual inflow of PM2.5 to western China with future changes in both emissions and climate. Fluxes of BC and OC from South Asia to China in spring contribute a large fraction of the annual inflow of PM2.5. The annual inflow of PM2.5 from South Asia and Southeast Asia to China is estimated to change by −8%, +281%, and +227% over 2000–2050 owing to climate change alone, changes in emissions alone, and changes in both climate and emissions, respectively. While the 4° × 5° spatial resolution is a limitation of the present study, the direction of predicted changes in aerosol levels and transboundary fluxes still provides valuable insight into future air quality.


Introduction
Aerosols are important air pollutants that lead to negative health impacts, reductions in visibility, and changes in climate (Intergovernmental Panel on Climate Change (IPCC), 2007).Concentrations of major atmospheric aerosol species (sulfate, nitrate, ammonium, black carbon, organic carbon, and mineral dust) are especially high in China (Matsui et al., 2009;Tie and Cao, 2010;Qu et al., 2010;Cao et al., 2012), driven by a combination of direct and precursor emissions (Streets et al., 2003) and regional meteorological conditions (Zhang et al., 2010a;Zhu et al., 2012).Estimating future aerosol levels in China is essential in considerations of air quality both over China itself and in the Northern Hemisphere.
In the absence of changes in emissions of primary aerosols as well as aerosol precursors, climate change itself will influence future aerosol levels.For example, coupled climate-chemical transport modeling studies show that climate change alone can lead to increased surface ozone in anthropogenically impacted regions by 1-10 ppbv in summertime over the coming decades, based on the IPCC future scenarios (Jacob and Winner, 2009).This increase is a result of slower transport, enhanced biogenic hydrocarbon emissions, and accelerated decomposition of peroxyacetyl nitrate (PAN) at higher temperatures (Hogrefe et al., 2004;Liao et al., 2006;Murazaki and Hess, 2006;Steiner et al., 2006;Racherla and Adams, 2008;Wu et al., 2008;Jacob and Winner, 2009;Andersson and Engardt, 2010;Chang et al., 2010;Lam et al., 2011;Katragkou et al., 2011;Langner et al., 2012;Reuten et al., 2012;Wang et al., 2013a).A warmer future climate is also predicted to influence aerosol levels over the United States and Europe by as much as 1 µg m −3 through altered concentrations of atmospheric oxidants, by changed precipitation and boundary layer height, and by shifting gas-particle equilibria (Liao et al., 2006;Unger et al., 2006;Bauer et al., 2007;Jacob and Winner, 2009;Pye et al., 2009;Day and Pandis, 2011;Lam et al., 2011;Tai et al., 2012a, b;Juda-Rezler et al., 2012).Because of the enormous importance of China as a source of aerosols, a study that addresses how aerosol levels in China may change over the coming decades is called for.
Another issue that is associated with the aerosol levels in China is the future transboundary aerosol transport.A number of observational and modeling analyses have demonstrated the importance of present-day long-range transport of aerosols from East Asia.Surface observations at island sites (Huebert et al., 2001;Prospero et al., 2003) and aircraft observations in Asian outflow over the Northwest Pacific (Jordan et al., 2003;Maxwell-Meier et al., 2004) and the Northeast Pacific (Clarke et al., 2001;Price et al., 2003) have documented the spring maximum in transpacific transport.By using the global atmospheric chemical transport model GEOS-Chem, Park et al. (2003) predicted that transpacific transport contributes about 10 % of the annual mean natural background surface-layer concentrations of black carbon over the United States.By using satellite measurements of aerosol optical depth over the North Pacific together with GEOS-Chem simulation, Heald et al. (2006) showed that transport from Asia led to a seasonal mean increase of surface-layer sulfate concentration of 0.16 µg m −3 (with 50 % uncertainty) in the northwestern United States in spring of 2001.Chin et al. (2007) predicted an enhancement of similar magnitude in surface-layer sulfate aerosol in the western United States in 2001 by long-range aerosol transport; the annual mean contribution to sulfate concentration was estimated to be 0.1-0.2µg m −3 using the global model GO-CART.Yu et al. (2008) performed a satellite-based assessment of transpacific transport of anthropogenic and biomass burning aerosols based on 2002-2005 aerosol optical depths from the Moderate Resolution Imaging Spectroradiometer (MODIS).They estimated that about 25 % of aerosol mass exported from East Asia to the northwestern Pacific Ocean can reach the west coast of North America.Two studies examined aerosol transport from Europe and South Asia to China.Chin et al. (2007) estimated that European emissions can increase the surface ammonium sulfate concentrations over eastern Asia by 0.2-0.5 µg m −3 .Using the GEOS-Chem model, Zhang et al. (2010b) estimated that organic carbon aerosol from South Asia contributed 50-70 % of organic carbon (OC) mass over southern China and 20-50 % of OC over the western North Pacific in the middle troposphere in summer of 1998.
We present here a study to estimate the following: (1) the changes of aerosol levels in China over the years 2000-2050 as a result of projected changes in emissions and climate, (2) the changes of transboundary fluxes of aerosols into or out of China over this time period.This study builds on two previous ones.Liao et al. (2007) used the Goddard Institute for Space Studies (GISS) general circulation model (GCM) 3 to drive GEOS-Chem to simulate climatological presentday aerosol levels in the United States, and Pye et al. (2009) investigated the effects of projected climate and emissions changes on 2000-2050 sulfate-nitrate-ammonium aerosols in the United States using the same GISS Model 3/GEOS-Chem combination.As a basis for the present study, the IPCC emission scenario A1B (Nakicenovic and Swart, 2000) is adopted; this scenario represents a future world with rapid economic growth and introduction of new and more energyefficient technologies.GISS Model 3 global meteorological fields are used to drive the atmospheric chemical transport model GEOS-Chem for both present day (1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005) and years 2046-2055.Effects of climate change alone, emission changes alone, and both climate and emissions changes together on aerosol levels and transboundary fluxes are simulated.The models in the present study have a relatively coarse spatial resolution of 4 • latitude by 5 • longitude, which are not expected to capture the characteristically high concentrations of aerosols in China's major urban areas.Nevertheless, the direction of projected changes in aerosol levels should be correctly predicted.
The methods and model setup used to simulate presentday and year 2050 aerosols are described in Sect. 2. Section 3 evaluates simulated present-day concentrations of aerosols in China.Section 4 shows predictions of future aerosol levels over China due to changes in climate alone, emissions alone, and combined climate and emissions changes, and Sect. 5 estimates future changes in transboundary transport of aerosols to examine inflow to and outflow from China.
We perform simulations for four cases:

Emissions
Present-day and year 2050 assumed anthropogenic emissions of aerosol precursor and aerosols are listed in Table 1.Emissions of O 3 precursors (including NO x , CO, and non-methane volatile organic compounds (NMVOCs)) follow those in Wu et al. (2008), and those of NH 3 and SO 2 are taken from Pye et al. (2009).The base year for present-day anthropogenic emissions is 1999 for the United States (Wu et al., 2008) and 1998elsewhere (Pye et al., 2009).Year 2050 anthropogenic emissions of ozone precursors, aerosol precursors, and aerosols from the IPCC A1B scenario were generated by the Integrated Model to Assess the Greenhouse Effect (IMAGE) socioeconomic model using prescribed growth factors for different regions, species, and sources (Streets et al., 2004).Ammonia emissions have an imposed seasonality that was determined as a function of temperature for one base year in this model.In present day, anthropogenic emissions of NO x , CO, NMVOCs, SO 2 , NH 3 , OC, and BC in eastern China are estimated to account for 12 %, 16 %, 12 %, 20 %, 18 %, 10 %, and 19 %, respectively, of the total global emissions.Relative to the assumed present day, year 2050 anthropogenic emissions of NO x , CO, NMVOCs, SO 2 , NH 3 , OC, and BC in eastern China (20 • -55 • N, 98 • -125 • E) are estimated to change by +72 %, −7 %, +86 %, −29 %, −12 %, −40 %, and −59 %, respectively.
Present-day and year 2050 natural emissions of ozone and aerosol precursors include NO x from lightning and soil, and biogenic hydrocarbons (Table 2), which are calculated based on the GISS Model 3 meteorological parameters.Lightning NO x emissions are parameterized based on convective cloudtop height (Price and Rind, 1992;Wang et al., 1998).Soil NO x emissions are calculated as a function of temperature, wind speed, and precipitation (Yienger and Levy, 1995).Representation of biogenic emissions follows the algorithm of Guenther et al. (1995), which considers light and temperature dependence but does not account for the suppression of isoprene emissions under elevated ambient CO 2 concentrations (Rosenstiel et al., 2003) and climate-induced changes in land cover.Simulated natural emissions of NO x and biogenic hydrocarbons in eastern China are estimated to increase, respectively, by +20 % and +22 % over 2000-2050, with the increases in biogenic emissions resulting mainly from the future increases in temperature.
For both radiative forcing in climate simulation and chemical reactions, present-day methane levels in the model are based on observations and set to 1750 ppb with a 5 % interhemispheric gradient (Wu et al., 2008).The future (2046)(2047)(2048)(2049)(2050)(2051)(2052)(2053)(2054)(2055) methane level in GEOS-Chem follows the IPCC A1B scenario and is set to 2400 ppb for simulations in which changes in anthropogenic emissions are considered (Pye et al., 2009).Projected changes are all statistically significant at the 95% level, as determined by the student's two sample t-test.predicted to be more pronounced than under scenario B1 and less pronounced than under A2 (IPCC, 2007).
Projected changes in precipitation in China from present day to year 2050 are shown in     maximum concentrations of 7-9 µg m −3 in JJA and of 5-7 µg m −3 in DJF, MAM, and SON in eastern China.Although strong photochemistry facilitates maximum sulfate formation in JJA, prevalent precipitation in southern China in JJA (Fig. 2) leads to enhanced wet removal of sulfate in that region.Simulated concentrations of NO − 3 are generally higher than those of SO 2− 4 in eastern China, which is likely caused by the overestimate of NH 3 emissions (Wang et al., 2013b).The highest NO − 3 concentrations of 9-12 µg m −3 are simulated in DJF because of the relatively low temperatures and precipitation.In contrast, high temperatures and large rainfall in JJA lead to the lowest NO − 3 concentrations.Because of the excess amount of NH 3 in eastern China (Wang et al., 2013b), ammonium aerosol exists predominantly as ammonium sulfate or ammonium nitrate; its concentrations are simulated to be in the range of 3-7 µg m −3 in eastern China in all seasons.Predicted BC and OC concentrations are high in DJF and SON and low in MAM and JJA, owing to the seasonal variation of precipitation (Fig. 2).Simulated PM 2.5 concentrations show the highest values of 18-32 µg m −3 in DJF and of 18-28 µg m −3 in JJA.In the surface layer, NO − 3 is predicted to have been the most abundant aerosol species over eastern China in 1996-2005, followed by SO 2− 4 , NH + 4 , OC, and BC.

Comparisons of simulated concentrations with measurements
Three previous studies have compared the simulated aerosol concentrations in GEOS-Chem with measurements taken        1.5 µg m −3 in large-scale simulations and up to 3.5 µg m −3 in fine-scale simulations under the IPCC A1B scenario.

Effect of changes in anthropogenic emissions alone
In the IPCC A1B scenario, year 2050 annual anthropogenic emissions of SO 2 , BC, and OC in eastern China are predicted to decrease by 29 %, 59 %, and 40 %, respectively, relative to the present-day values.Future concentrations of sulfate, BC, and OC in China are hence predicted to decrease (Fig. 8).The predicted reductions in sulfate aerosol in eastern China lie within the range of 0.5-3.5 µg m −3 throughout the year.
Although the simulated present-day concentrations of BC and OC are lower than those of sulfate, the decreases in BC and OC are 0.5-2 µg m −3 and 0.5-3.5 µg m −3 , respectively, in eastern China.Unlike sulfate, BC, and OC, based on changes in emissions alone, nitrate concentrations are predicted to increase in the future due to changing anthropogenic emissions, with the largest increases of 1-2 µg m −3 in DJF, 3.5-4.5 µg m −3 in JJA, and 2-3.5 µg m −3 in MAM and SON in eastern China.These increases can be explained by the assumed 2000-2050 (Table 1) as well as the fact that the future reductions in SO 2 favor the formation of ammonium nitrate.As a result, ammonium concentrations show increases of about 0.5 µg m −3 over those locations with large increases in nitrate.
The net effect of the changes in all aerosol species owing to changes in emissions alone is an overall decrease in PM 2.5 concentrations in China except for the Tibet Plateau.Over eastern China, the projected largest decreases in PM 2.5 as a result of the future changes in emissions are 7.5-8 µg m −3 (or 20-40 % relative to the present-day values) in DJF, 1-5 µg m −3 (or 10-30 %) in MAM and SON, and 1-3 µg m −3 (or 10-40 %) in JJA.As noted, the increase in PM 2.5 over or near the Tibet Plateau is caused by long-range transport from South Asia.

Effect of changes in both climate and anthropogenic emissions
Predicted future changes in aerosols as a consequence of future changes in both climate and anthropogenic emissions are summarized in Fig. 9. Concentrations of sulfate, BC, and OC in China are predicted to decrease in all seasons under future changes in both climate and anthropogenic emissions, although climate change can, to some extent, offset the effect of reductions in anthropogenic emissions on these species (Fig. 7).In eastern China, concentrations of sulfate are predicted to decrease 0.5-3.5 µg m −3 in SON and 0.5-2 µg m −3 in other seasons, while those of BC and OC to decrease by 0.5-2 µg m −3 and 0.5-3.5 µg m −3 , respectively, in all seasons.
For nitrate aerosol in eastern China, while climate change exerts an effect opposite to that of changes in anthropogenic emissions in DJF, JJA, and SON, the effect of climate change enhances the changes due to emissions in MAM (Fig. 7).As a result of changes in both climate and anthropogenic emissions, concentrations of nitrate in eastern China are predicted to exhibit reductions of 0.5-1 µg m −3 in DJF and increases of 0.5-4.5 µg m −3 in other seasons.
The simulated 2000-2050 changes in PM 2.5 are dominated by future changes in emissions; concentrations of PM 2.5 in eastern China are simulated to decrease by 3-7.5 µg m −3 (or 10-30 % relative to the present-day values) in DJF, 1-5 µg m −3 (or 30-40 %) in JJA, and 3-7.5 µg m −3 (or 20-40 %) in SON.In MAM, climate change has a dominant role in influencing the future changes in aerosols in the lower reaches of Yangtze River, where increases of up to 1 µg m −3 in PM 2.5 are simulated.
With reductions in SO 2 , BC, and OC in eastern China over 2000-2050 by 28.5 %, 58.6 %, and 39.8 % (Table 1), respectively, annual mean concentrations of sulfate, BC, and OC over eastern China are simulated to be reduced, respectively, by 19.2 %, 56.4 %, and 35.9 % with future changes in emissions alone whereas by 18.5 %, 56.6 %, and 36.5 % with future changes in both climate and emissions.These results

Estimated present-day outflow
Simulated present-day seasonal and annual total fluxes of SO 2− 4 , NO − 3 , NH + 4 , BC, OC, and PM 2.5 across the meridional plane along 135 • E from 20 • to 55 • N are listed in Table 3.Among all aerosol species, the estimated present-day outflow of SO 2− 4 of 5.2 Tg yr −1 across this plane is the largest, contributing 48 % of the annual outflow of PM 2.5 .The estimated fluxes of NO − 3 , NH + 4 , BC, and OC account for 19 %, 15 %, 5 %, and 13 % of annual outflow of PM 2.5 , respectively.Aerosol fluxes show strong seasonal variations.The  4 is simulated to peak in DJF, which results from the simulated high concentrations of sulfate (Fig. 5) and the strong westerlies (Fig. 11) in that season.All other aerosol species exhibit maximum outflow in MAM, a result that agrees with the conclusions from previous studies (Holzer et al., 2005).Heald et al. (2006) and Chin et al. (2007) reported that the export from Asia is most efficient in spring with nearly all East Asian air involved in transpacific transport.The outflow of aerosols is generally the weakest in JJA, as reported by Holzer et al. (2005) and Yu et al. (2008).
The pressure-latitude cross sections of the simulated fluxes of PM 2.5 at 135 • E are shown in Fig. 11.The maximum fluxes of aerosols are found at 500-700 hPa in DJF and SON and at 400-600 hPa in MAM and JJA.The locations of these maximum fluxes shift from 20 • -35 • N in DJF and MAM to 30 • -50 • N in JJA and SON, which are consistent with the changes in westerlies (Fig. 11).Our simulated vertical distributions of the strongest fluxes of aerosols agree with those reported by Wang et al. (2009), who showed that the transport to North America occurs mainly in the mid-to upper troposphere.

Effect of changes in climate alone
Simulated future changes in seasonal and annual fluxes of SO 2− 4 , NO − 3 , NH + 4 , BC, OC, and PM 2.5 through the meridional plane along 135 • E are also listed in Table 3.As a result of the climate change alone, the annual fluxes of SO   −4.2 %, −15.6 %, −10.1 %, 0.0 %, −4.5 %, and −7.0 %, respectively, relative to the present-day values.Nitrate aerosol outflow is projected to exhibit the largest reduction, because nitrate aerosol concentrations in China decrease significantly in a warmer climate (Fig. 7).The fluxes of all aerosol species show reductions in all seasons except a small increase (0.7 %) in flux of SO 2− 4 in MAM.

Effect of changes in anthropogenic emissions alone
Annual fluxes of SO 2− 4 , NO − 3 , NH + 4 , BC, OC, and PM 2.5 are simulated to change by −4.6 %, +24.6 %, +8.3 %, −39.6 %, −22.4 %, and −0.7 %, respectively (Table 3), as a result of the changes in anthropogenic emissions alone.While nitrate and ammonium outflow fluxes exhibit large increases owing to the projected increases in concentrations in China (Fig. 8), the fluxes of SO 2− 4 , BC, and OC decrease, corresponding to the future decreases in concentrations of these species (Fig. 8).Consequently, the annual outflow of PM 2.5 is estimated to exhibit a modest change of −0.7 % over 2000-2050 as a result of changes in emissions alone.

Effect of changes in both climate and anthropogenic emissions
Simulated year 2050 outflow fluxes of aerosols through the meridional plane along 135   upper troposphere over 100 • -125 • E (Fig. 13).In MAM and JJA, the southerlies associated with the summer monsoon generally favor northward transport (Fig. 13).The fluxes of aerosols across the southern boundary of China are generally much smaller than those through the eastern and western boundaries.One exception is that the fluxes of BC and OC to China in MAM are amplified as a result of the predominant biomass burning in South Asia in this season (Zhang et al., 2010b).

Effect of changes in climate alone
As a result of the future climate change alone, the changes in annual fluxes of aerosol species are generally small except that the annual southward flux of NO − 3 is simulated to increase by 26.7 % relative to the present-day value.The percentage changes in seasonal fluxes are usually much larger.
In DJF, the northerlies in the lower troposphere east of 100 • E are predicted to be stronger and the southerlies west of 100 • E become weaker (Fig. 13), which lead to increases in transport of SO 2− 4 , NO − 3 , and PM 2.5 from China to South Asia by 33 %, 11 %, and 67 %, respectively.In JJA, the southerlies associated with the summer monsoon are simulated to strengthen in the future climate and hence favor the northward transport of aerosols.The fluxes of all aerosol species to South Asia show reductions in SON, which can be explained by the weaker northerlies in the future atmosphere (Fig. 13).

Effect of changes in anthropogenic emissions alone
The annual fluxes of SO 2− 4 and NH + 4 are simulated to change from a net southward outflow to a net northward inflow, as a result of the changes in anthropogenic emissions over 2000-2050 (Table 5).The annual southward flux of NO − 3 increases by 60 %, and the annual northward fluxes of BC, OC, and PM 2.5 increase by 48 %, 14 %, and 281 %, respectively.The annual fluxes of SO 2− 4 and NH + 4 exhibit large increases in northward transport because of the large increases in aerosol concentrations in South Asia and Southeast Asia (Fig. 8), and the annual inflow of BC and OC also increases because the reductions in carbonaceous aerosol concentrations in eastern China are projected to be larger than those in South Asia and Southeast Asia (Fig. 8).In contrast, the annual flux of NO − 3 exhibits southward outflow because of the large increases in NO − 3 levels in eastern China (Fig. 8).5).As a result of the future changes in anthropogenic emissions alone, concentrations of sulfate, BC, and OC are simulated to decrease by 0.5-3.5 µg m −3 , 0.5-2 µg m −3 , and 0.5-3.5 µg m −3 , respectively, in eastern China.On the contrary,   negative change in annual outflow of PM 2.5 as a result of changes in emissions alone.The present-day inflow of PM 2.5 aerosols from Europe and Central Asia to western China (the fluxes through the meridional plane along 75 • E from 35 • to 55 • N) is calculated to be 3.9 Tg yr −1 .Over 2000-2050, the fluxes of nitrate and ammonium aerosols are estimated to increase largely as a result of future changes in emissions, leading to an overall 10.5 % estimated increase in annual inflow of PM 2.5 to western China as a result of future changes in both emissions and climate.

Effect of changes in both climate
The present-day net transport of PM 2.5 aerosols across the southern boundary (the latitudinal plane along 21.7 • N from 90 • to 125 • E) is estimated as 0.26 Tg yr −1 , in which the fluxes of sulfate, nitrate, and ammonium are southward out of China and those of BC and OC are northward into China.Fluxes of BC and OC to China in MAM contribute to a large fraction of the annual inflow of PM 2.5 , as a result of biomass burning in South Asia in this season.The annual net inflow of PM 2.5 across the southern boundary of China is estimated to change by −8 %, +281 %, and +227 % over 2000-2050 owing to climate change alone, changes in emissions alone, and changes in both climate and emissions, respectively.
Results from the present study indicate that climate change is important for domestic air quality in China as well as longrange transport of aerosols.
(1) year 2000 climate and emissions, (2) 2050 climate and 2000 anthropogenic emissions of aerosol precursor and aerosols, (3) 2000 climate and 2050 anthropogenic emissions of aerosol precursor and aerosols, and (4) 2050 climate and emissions.Each case is integrated for 10 yr (driven by 1996-2005 meteorological fields to represent year 2000 climate or by 2046-2055 meteorology to represent year 2050 climate) following 1 yr of model spin-up.All the results presented in this paper are 10 yr averages.Statistical analysis based on Student's two-sample t test is applied to the meteorological fields and concentrations of the 10 yr simulations.

Fig. 1 .
Fig. 1.Projected changes in surface air temperature (K) in China from the present day (1996-2005) to future (2046-2055) under the IPCC A1B scenario.Projected changes are all statistically significant at the 95% level, as determined by the student's two sample t-test.

Fig. 1 .
Fig. 1.Projected changes in surface air temperature (K) in China from the present day (1996-2005) to future (2046-2055) under the IPCC A1B scenario.Projected changes are all statistically significant at the 95 % level, as determined by Student's two-sample t test.

Fig. 2 .
Fig. 2. (a) Simulated precipitation (mm day -1 ) in China in present day; (b) Projected changes in precipitation (mm day -1 ) in China from the present day (1996-2005) to future (2046-2055) under the IPCC A1B scenario; (c) The percentage changes in precipitation relative to present day.The dotted areas in (c) are statistically significant at the 95% level, as determined by the student's two sample t-test.

Fig. 2 .
Fig. 2. (a) Simulated precipitation (mm day −1 ) in China in present day; (b) projected changes in precipitation (mm day −1 ) in China from the present day (1996-2005) to future (2046-2055) under the IPCC A1B scenario; (c) the percentage changes in precipitation relative to present day.The dotted areas in (c) are statistically significant at the 95 % level, as determined by Student's two-sample t test.

Fig. 3 .
Fig. 3. (a) Simulated cloud fraction (true fraction, 1.0= total overcast cloud) in China in present day; (b) Projected changes in cloud fraction in China from the present day (1996-2005) to future (2046-2055) under the IPCC A1B scenario.The dotted areas in (b) are statistically significant at the 95% level, as determined by the student's two sample t-test.

Fig. 3 .
Fig. 3. (a) Simulated cloud fraction (true fraction, 1.0 = total overcast cloud) in China in present day and (b) projected changes in cloud fraction in China from the present day (1996-2005) to future (2046-2055) under the IPCC A1B scenario.The dotted areas in (b) are statistically significant at the 95 % level, as determined by Student's two-sample t test.
Fig. 2. Relative to present day, year 2050 precipitation in DJF is estimated to increase by 50-80 % in northern China and decrease by about 20 % in southern China; note that the present-day precipitation in DJF is the smallest among all seasons.In March-April-May (MAM), year 2050 precipitation is estimated to increase by 20-40 % in southern China where the present-day seasonal precipitation is greatest.Precipitation is predicted to increase generally in eastern China in June-July-August (JJA) and September-October-November (SON).Note that the projected patterns of precipitation changes (the increases in precipitation in northern China in DJF and the increases in precipitation in eastern China in JJA) from the GISS Model 3

Fig. 4 .
Fig. 4. (a) Simulated planetary boundary layer (PBL) depth (km) in China in present day; (b) Projected changes in PBL depth (km) in China from the present day (1996-2005) to future (2046-2055) under the IPCC A1B scenario; (c) The percentage changes in PBL depth relative to present day.The dotted areas in (c) are statistically significant at the 95% level, as determined by the student's two sample t-test.

Fig. 4 .
Fig. 4. (a) Simulated planetary boundary layer (PBL) depth (km) in China in present day; (b) projected changes in PBL depth (km) in China from the present day (1996-2005) to future (2046-2055) under the IPCC A1B scenario; (c) the percentage changes in PBL depth relative to present day.The dotted areas in (c) are statistically significant at the 95 % level, as determined by Student's two-sample t test.

Fig. 6 .
Fig. 6.Comparisons of simulated present-day concentrations of sulfate, nitrate, BC, and OC aerosols with measurements.Simulated values are seasonal averages over 1996-2005.Also shown is the 1:1 line (dashed) and linear fit (solid line and equation).R is the correlation coefficient between simulated and measured concentrations.

Fig. 6 .
Fig. 6.Comparisons of simulated present-day concentrations of sulfate, nitrate, BC, and OC aerosols with measurements.Simulated values are seasonal averages over 1996-2005.Also shown is the 1 : 1 line (dashed) and linear fit (solid line and equation).R is the correlation coefficient between simulated and measured concentrations.

Fig. 7 .
Fig. 7. Predicted changes in surface-layer concentrations of aerosols (g m -3 ) due to changes in climate alone from the present day (1996-2005) to the future (2046-2055).Greenhouse gases follow the IPCC scenario A1B.Anthropogenic emissions are held at present-day values, but natural emissions may change in response to climate.The dotted areas are statistically significant at the 95% level, as determined by the student's two sample t-test.

Fig. 7 .Figure 7
Fig. 7. Predicted changes in surface-layer concentrations of aerosols (µg m −3 ) due to changes in climate alone from the present day (1996-2005) to the future (2046-2055).Greenhouse gases follow the IPCC scenario A1B.Anthropogenic emissions are held at present-day values, but natural emissions may change in response to climate.The dotted areas are statistically significant at the 95 % level, as determined by Student's two-sample t test.

Fig. 8 .
Fig. 8. Predicted changes in surface-layer concentrations of aerosols (g m -3 ) due to changes in anthropogenic emissions alone from the present day (1996-2005) to the future (2046-2055).Almost all the changes over China are statistically significant at the 95% level, as determined by the student's two sample t-test.

Fig. 8 .
Fig. 8. Predicted changes in surface-layer concentrations of aerosols (µg m −3 ) due to changes in anthropogenic emissions alone from the present day (1996-2005) to the future (2046-2055).Almost all the changes over China are statistically significant at the 95 % level, as determined by Student's two-sample t test.

Fig. 9 .
Fig. 9. Predicted changes in surface-layer concentrations of aerosols (g m -3 ) due to changes in both climate and anthropogenic emissions from the present day (1996-2005) to the future (2046-2055).Almost all the changes over China are statistically significant at the 95% level, as determined by the student's two sample t-test.

Fig. 9 .
Fig. 9. Predicted changes in surface-layer concentrations of aerosols (µg m −3 ) due to changes in both climate and anthropogenic emissions from the present day (1996-2005) to the future (2046-2055).Almost all the changes over China are statistically significant at the 95 % level, as determined by Student's two-sample t test.

Fig. 10 .
Fig. 10.The locations of the 3 vertical planes through which fluxes of transboundary aerosols are calculated: the meridional plane along 135E from 20 to 55N to show the outflow from eastern China to the West Pacific, the meridional plane that along 75E from 35 to 55N to show the inflow from Europe and Central Asia to China, and the latitudinal plane along 21.7N from 90 to 125E to show the transport to or from South Asia and Southeast Asia.

Fig. 10 .
Fig. 10.The locations of the 3 vertical planes through which fluxes of transboundary aerosols are calculated: the meridional plane along 135 • E from 20 • to 55 • N to show the outflow from eastern China to the West Pacific; the meridional plane along 75 • E from 35 • to 55 • N to show the inflow from Europe and Central Asia to China; and the latitudinal plane along 21.7 • N from 90 • to 125 • E to show the transport to or from South Asia and Southeast Asia.

Fig. 11 .
Fig. 11.(a) Simulated present-day mass fluxes of PM 2.5 and zonal winds.Projected changes in mass fluxes of PM 2.5 and zonal winds from the present day (1996-2005) to future (2046-2055) owing to (b) climate change alone, (c) changes in anthropogenic emissions alone, and (d) changes in both climate and anthropogenic emissions.Mass fluxes of PM 2.5 are shown by shades (Units: Tg) and winds are represented by contours (Units: m s -1 ).Both mass fluxes and winds are those through the meridional plane along 135E from 20 to 55N.The dotted areas are statistically significant at the 95% level, as determined by the student's two sample t-test.

Fig. 11 .
Fig. 11.(a) Simulated present-day mass fluxes of PM 2.5 and zonal winds.Projected changes in mass fluxes of PM 2.5 and zonal winds from the present day (1996-2005) to future (2046-2055) owing to (b) climate change alone, (c) changes in anthropogenic emissions alone, and (d) changes in both climate and anthropogenic emissions.Mass fluxes of PM 2.5 are shown by shades (units: Tg), and winds are represented by contours (units: m s −1 ).Both mass fluxes and winds are those through the meridional plane along 135 • E from 20 • to 55 • N. The dotted areas are statistically significant at the 95 % level, as determined by Student's two-sample t test.

a
Seasonal and annual total fluxes averaged over 1996-2005; b year 2050 fluxes simulated with climate change alone; c year 2050 fluxes simulated with changes in anthropogenic emissions alone; d year 2050 fluxes simulated with changes in both climate and anthropogenic emissions; e NonL= (2050 b −2000 a )+(2050 c −2000 a ) 2050 d −2000 a .

a
Seasonal and annual total fluxes averaged over 1996-2005; b year 2050 fluxes simulated with climate change alone; c year 2050 fluxes simulated with changes in anthropogenic emissions alone; d year 2050 fluxes simulated with changes in both climate and anthropogenic emissions; e NonL = (2050 b −2000 a )+(2050 c −2000 a ) 2050 d −2000 a .

Fig. 12 .
Fig. 12. Same as Figure 11 but mass fluxes and winds are those through the meridional plane along 75E from 35 to 55N.

Fig. 12 .
Fig. 12. Same as Fig. 11 but mass fluxes and winds are those through the meridional plane along 75 • E from 35 • to 55 • N.
and anthropogenic emissions Simulated fluxes of aerosols through the latitudinal plane along 21.7 • N from 90 • to 125 • E with the future changes in both anthropogenic emissions and climate are listed in Table 5.The year 2050 annual fluxes of SO 2− 4 , NH + 4 , BC, OC, and PM 2.5 are simulated as northward into China, and the annual flux of NO − 3 is southward out of China.The 2000-2050 changes in fluxes of all aerosol species are dominated by the contributions from changes in anthropogenic emissions.However, the role of future climate change can exceed that of future changes in emissions for the fluxes of NO − 3 in DJF, BC in JJA, as well as OC in DJF (Table

Fig. 13 .
Fig. 13.(a) Simulated present-day mass fluxes of PM 2.5 and meridional winds.Projected changes in mass fluxes of PM 2.5 and meridional winds from the present day (1996-2005) to future (2046-2055) owing to (b) climate change alone, (c) changes in anthropogenic emissions alone, and (d) changes in both climate and anthropogenic emissions.Mass fluxes of PM 2.5 are shown by shades (Units: Tg) and winds are represented by contours (Units: m s -1 ).Both mass fluxes and winds are those through latitudinal plane along 21.7N from 90 to 125E.Positive (negative) fluxes indicate northward (southward) transport.The dotted areas are statistically significant at the 95% level, as determined by the student's two sample t-test.

Fig. 13 .
Fig. 13.(a) Simulated present-day mass fluxes of PM 2.5 and meridional winds.Projected changes in mass fluxes of PM 2.5 and meridional winds from the present day (1996-2005) to future (2046-2055) owing to (b) climate change alone, (c) changes in anthropogenic emissions alone, and (d) changes in both climate and anthropogenic emissions.Mass fluxes of PM 2.5 are shown by shades (units: Tg), and winds are represented by contours (units: m s −1 ).Both mass fluxes and winds are those through latitudinal plane along 21.7 • N from 90 • to 125 • E. Positive (negative) fluxes indicate northward (southward) transport.The dotted areas are statistically significant at the 95 % level, as determined by Student's two-sample t test.

Table 5 .
Simulated transport of aerosols to or from southern China through the latitudinal plane along 21.7 • N from 90 • to 125 • E. Positive (negative) values indicate northward (southward) transport.Percentage change is listed in the parentheses as the 2050 flux has the same sign as 2000 flux.The units are Tg season −1 for seasonal fluxes and Tg yr −1 for annual fluxes of whole species (e.g., Tg(SO 2− 4 ) season −1 , Tg(SO 2− 4 ) yr −1 ).

a
Seasonal and annual total fluxes averaged over 1996-2005; b year 2050 fluxes simulated with climate change alone; c year 2050 fluxes simulated with changes in anthropogenic emissions alone; d year 2050 fluxes simulated with changes in both climate and anthropogenic emissions; e NonL = (2050 b −2000 a )+(2050 c −2000 a ) 2050 d −2000 a .nitrate concentrations are predicted to increase in eastern China over 2000-2050, with the largest increases of 1-2 µg m −3 in DJF, 3.5-4.5 µg m −3 in JJA, and 2-3.5 µg m −3 in MAM and SON.The projected changes in PM 2.5 owing to the changes in emissions alone show decreases in PM 2.5 concentrations of 1-8 µg m −3 in eastern China over 2000-2050 at a 4 • × 5 • resolution.Under the IPCC A1B scenario, future changes in anthropogenic emissions exert a larger effect on year 2050 PM 2.5 concentrations in eastern China than does projected future climate change.The estimated year 2050 transboundary fluxes of aerosol species to and from China are sensitive to future changes in climate and emissions.The present-day flux of PM 2.5 from eastern China to the western Pacific (through the meridional plane along 135 • E from 20 • to 55 • N) is simulated to be 10.8 Tg yr −1 , and the annual outflow of PM 2.5 is estimated to change by −7.0 %, −0.7 %, and −9.0 %, respectively, over 2000-2050 owing to future climate change alone, future changes in emissions alone, and future changes in both climate emissions.Climate change is predicted to have a larger impact on future fluxes of aerosols than changes in emissions.While annual fluxes of all aerosol species show reductions by climate change alone, future increases in outflow of nitrate and ammonium offset to a large extent the future decreases in outflow of sulfate, BC, and OC, leading to a small www.atmos-chem-phys.net/13/7937/2013/Atmos.Chem.Phys., 13, 7937-7960, 2013

Table 3 .
Simulated outflow of aerosols from eastern China through the meridional plane along 135 • E from 20 • to 55 • N. Numbers in the parentheses are percentage changes relative to the present-day fluxes.The units are Tg season −1 for seasonal fluxes and Tg yr −1 for annual fluxes of whole species (e.g., Tg(SO 2− 4 ) season −1 , Tg(SO 2− 4 ) yr −1 ).
• E from 20 • to 55 • N with the future changes in both anthropogenic emissions and climate are listed in Table 3.The annual fluxes of SO 2− 4 , NO − 3 , NH + 4 ,

Table 4 .
Simulated inflow of aerosols to western China through the meridional plane along 75 • E from 35 • to 55 • N. Numbers in the parentheses are percentage changes relative to the present-day fluxes.The units are Tg season −1 for seasonal fluxes and Tg yr −1 for annual fluxes of whole species (e.g., Tg(SO 2− 4 ) season −1 , Tg(SO 2− 4 ) yr −1 ).