Analysis of aerosol distribution variations over China for the period 2045–2050 under different Representative Concentration Pathway scenarios

Yi Go ,Meigen Zhng ,b,c,?,Chengli Wu

a State Key Laboratory of Atmospheric Boundary Layer Physics and Atmospheric Chemistry (LAPC), Institute of Atmospheric Physics, Chinese Academy of Sciences,Beijing, China

b University of Chinese Academy of Sciences, Beijing, China

c Center for Excellence in Urban Atmospheric Environment, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen, China

d International Center for Climate and Environment Sciences, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China

Keywords:PM2.5 RCP RAMS-CMAQ meteorological influence

ABSTRACT The regional air quality modeling system RAMS-CMAQ was applied to simulate the aerosol concentration for the period 2045–2050 over China based on the downscaled meteorological field of three RCP scenarios from CESM(NCAR’s Community Earth System Model) in CMIP5.The downscaling simulation of the meteorological field of the three RCP scenarios showed that, compared with that under RCP2.6, the difference in near-surface temperature between North and South China is weakened and the wind speed increases over North and South China and decreases over central China under RCP4.5 and RCP8.5.Under RCP2.6, from 2045 to 2050, the modeled average PM 2.5 concentration is highest, with a value of 40–50 μg m ? 3,over the North China Plain, part of the Yangtze River Delta, and the Sichuan Basin.Meanwhile, it is 30–40 μg m ? 3 over central China and part of the Pearl River Delta.Compared with RCP2.6, PM 2.5 increases by 4–12 μg m ? 3 under both RCP4.5 and RCP8.5, of which the SO 4 2 ? and NH 4 + concentration increases under both RCP4.5 and RCP8.5; the NO 3 ? concentration decreases under RCP4.5 and increases under RCP8.5; and the black carbon concentration changes very slightly, and organic carbon concentration decreases, under RCP4.5 and RCP8.5, with some increase over part of Southwest and Southeast China under RCP8.5.The difference between RCP4.5 and RCP2.6 and the difference between RCP8.5 and RCP2.6 have similar annual variation for different aerosol species, indicating that the impact of climate change on different species tends to be consistent.

1.Introduction

In recent years, the frequency of haze and smog weather in China’s major cities and developed regions, such as the North China Plain (NCP),Yangtze River Delta (YRD), and Pearl River Delta (PRD), has become very high.As the main pollutant in China, PMis the main cause of the country’s air pollution.Its high concentration is formed by the combination of emission sources and the meteorological field.It follows,therefore, that the effects of future climate change on the aerosol concentration in China draws considerable attention in the scientific community.

Many studies have investigated the changes in air quality in different regions under future climate change scenarios.For instance,Tai et al.(2012) analyzed the meteorological field from the CMIP5(phase 5 of the Coupled Model Intercomparison Project) multi-model simulations and indicated that, under the SRESA1B scenario, the aerosol concentration in the eastern United States increases due to the decrease in the frequency of the frontal system in the midlatitude region; whereas, the aerosol concentration shows a decreasing trend in the northwestern United States, which is affected by wind from the ocean.Hedegaard et al.(2013) used the Danish Eulerian Hemispheric Model to discover that, under the RCP4.5 scenario, for the period 2000–2100, climate change could result in a decrease in aerosol concentrations in the high latitudes of the Northern Hemisphere, especially in the Arctic, but a slight increase in the aerosol concentration(0–15%) over the Atlantic Ocean and parts of the tropics and subtropics.Allen et al.(2015) analyzed the multi-model results of the International Atmospheric Chemistry–Climate Model Comparison Plan and indicated that future climate change could lead to an increase in total precipitation in the atmosphere and a reduction of large-scale precipitation frequency, which would result in an increase of aerosol in the midlatitude Northern Hemisphere and tropical regions.Jiang et al.(2013) used the atmospheric chemical transport model GEOS-Chem to study the effects of future climate change on aerosol concentrations in China, and the results showed that future climatic change led to a 10%–20% increase in surface concentrations of PM.Fu et al.(2016) applied an atmospheric chemical transport model to analyze the impact of climate change on PMin East Asia from 1980 to 2010 and indicated that,when considering only the effects of climate change, the PM 2.5 concentration in northern China decreased by 4.0–12.0 μg m3 in winter, but increased by 6.0–8.0 μg min summer.

Most of the abovementioned studies were carried out based on global models with very low resolution and relatively simple atmospheric chemistry processes.In addition, there are very few studies in the literature that focused on the future distribution of aerosol concentrations in China.In order to explore the possible future changes of aerosol concentrations in China, especially in some key regions, a regional atmospheric chemistry model was used in the present study to simulate the aerosol concentration for the period 2045–2050 over China based on the downscaled meteorological field of three RCP scenarios from CMIP5.Furthermore, the annual variation of the aerosol concentration is analyzed in the chosen key regions.

2.Data and model description

The CMIP5 simulation results represent the highest level of current climate simulation ( Taylor et al., 2012 ; Guo et al., 2013 ; Joetzjer et al.,2013 ).In this study, the NCAR’s CESM (Community Earth System Model) RCP2.6, RCP4.5 and RCP8.5 future climate scenario data in CMIP5 were used as the model input meteorological data.We extracted the isothermal surface wind speed, temperature, humidity, and potential height in the data as the model input meteorological field data.The corresponding emission source at 2050 under RCP2.6, RCP4.5, and RCP8.5 was taken from the RCP Database (version 2.0).The regional atmospheric chemistry model system RAMS-CMAQ used in this study consists of two parts: CMAQ (Community Multi-Scale Air Quality Modeling System; Byun et al., 1999 ) and RAMS (Regional Atmospheric Modeling System; Pielke et al., 1992 ).The details of NCAR’s CESM future climate scenario data, the emission source data, and the model RAMS-CMAQ are in the supplementary file.The emissions distribution in China under RCP2.6 and the difference between RCP2.6 and both RCP4.5 and RCP8.5 are shown in Figs.S2–S4.

The horizontal grid spacing of the simulated grid was 64 km, covering the entire East Asian region.The simulated area was centered at(110°E, 35°N), with an area of 6654 km × 5440 km.The vertical direction was divided into 15 layers, with the top height at about 20.7 km and the first layer at about 100 m.The model simulation area is shown in Fig.S1.The red boxes from north to south in Fig.S1 represent the NCP (35°–43°N, 111°–121°E), the YRD (27°–34°N, 115°–122°E), and the PRD (21.5°–24°N, 112°–115.5°E).In this study, an 18-year simulation for the period 2045–2050 was conducted under RCP2.6, RCP4.5,and RCP8.5.The meteorological field was the downscaling simulation of CESM RCP2.6, RCP4.5, and RCP8.5 data during 2045–2050 by RAMS.

3.Results and discussion

3.1.Distribution of meteorological conditions in 2045–2050

The distribution of meteorological conditions averaged for 2045–2050 in the three RCPs is discussed in this section.Fig.1 shows the downscaling modeled horizontal distribution of the near-surface (first layer of the model) temperature, relative humidity (RH), and wind speed averaged for 2045–2050 under RCP2.6 and the difference between RCP2.6 and both RCP4.5 and RCP8.5.From the top panel it can be seen that, under RCP2.6, the temperature decreases from south to north,with values of about 295–300 K over the southeast coast and PRD, about 290–295 K over the YRD and central and southwestern China, and about 275–290 K over the NCP and Northeast and Northwest China.Compared with that under RCP2.6, the near-surface temperature increases by 0.9–1.2 K over Northeast and Northwest China, by 0.6–0.9 K over the NCP,by 0.3–0.6 K over part of the YRD and central China, and by 0–0.3 K over South China, including part of the YRD and PRD, under RCP4.5.Under RCP8.5, the temperature increases by more than 1.2 K over Northwest China, by 0.9–1.2 K over Northeast China, part of Northwest China, and part of Southeast China, and by 0.3–0.6 K over other areas of China.Therefore, compared with RCP2.6, the difference in near-surface temperature between North and South China is weakened under RCP4.5 and RCP8.5, especially under RCP4.5.The difference between RCP4.5 and RCP8.5 is that the temperature over South China increases more under RCP8.5 than RCP4.5.

As can be seen from the second panel, the RH under RCP2.6 decreases from east to west.It is about 80%–85% over the south coast of China, including the YRD, part of the PRD, and central China, about 70%–80% over Northeast China, the NCP, and part of the YRD, about 50%–70% over Northwest and Southwest China, and under 50% (its lowest values) over the Tibetan Plateau area.Compared with the nearsurface RH under RCP2.6, it changes slightly, with values between ? 2%and 2%, under RCP4.5.Specifically, it only increases by 0.5%–2% over the NCP and decreases slightly by 1%–3% over Northwest China.Under RCP8.5, compared with RCP2.6, RH decreases more than that under RCP4.5.Specifically, it decreases by 2%–3% over Northwest China and 0.5%–2% over part of central and northern China.Generally, the change in RH between different RCPs is not very significant.

From the third panel of Fig.1,it is apparent that the wind speed under RCP2.6 over South China is relatively lower (about 1–2 m s1 )than that over North China (about 3–6 m s).Wind speed is highest over the Tibetan Plateau, with values of about 6–8 m s1.Compared with that under RCP2.6, the wind speed increases by 0.1–0.2 m sover North China, South China, and Northwest China under RCP4.5.Meanwhile, it decreases by 0.2–0.4 m s1 over central and northeastern China, and by more than 0.6 m sover the Tibetan Plateau.The change in wind speed between RCP8.5 and RCP2.6 is like that between RCP4.5 and RCP2.6 but with larger areas of increase and decrease over North China and central China, respectively.The increase in wind speed over South China under RCP8.5 is smaller than that under RCP4.5.

Fig.1.Horizontal distribution of the near-surface (first layer of the model) temperature, relative humidity, and wind speed averaged for 2045–2050 under RCP2.6,and the differences in these meteorological parameters between RCP2.6 and both RCP4.5 and RCP8.5.

3.2. Distribution of aerosol concentration in 2045–2050

The horizontal distribution of the near-surface aerosol concentration averaged for the period 2045–2050 under RCP2.6, and the differences in these aerosol concentrations between RCP2.6 and both RCP4.5 and RCP8.5, are shown in Fig.2.From the first panel, the average PMconcentration in 2045–2050 is highest, with values of 40–50 μg m,over the NCP, part of the YRD, and the Sichuan Basin.The average PM 2.5 concentration is about 30–40 μg mover central China and part of the PRD, and about 10–20 μg m3 over South China.Under RCP4.5, the change in PMis very small, with values of 0–4 μg mover most areas of China.It only increases by 4–8 μg mover part of the YRD and PRD.The slight change is contributed by the decrease in organic carbon(OC), which is discussed later.Under RCP8.5, the PMconcentration increases more than that under RCP4.5 when compared with RCP2.6.It increases over most areas of China, except part of the PRD and YRD.Specifically, the increase is highest, with values of 12–16 μg m,over the Sichuan Basin, and then there are increases of about 8–12 μg mover central China, and 4–8 μg mover the NCP, Southeast China,and Northeast China.The PM 2.5 concentration decreases by about 2–6 μg mover part of the PRD and YRD.

To illustrate the bias of the modeled aerosol concentration, an additional simulation was conducted with the input of the CESM meteorological data in 2015 and emissions in 2010 under RCP4.5, and the modeled PM 2.5 concentration was compared with observational PMdata gathered by the Ministry of Environmental Protection of China in 2015.Fig.S5 is a scatterplot between the observed and simulated annual average PMsurface concentration for 2015 under RCP4.5 at 1395 sites in seven regions: NC (North China), EC (East China), SC (South China), NWC (northwest China), SWC (southwest China), NEC (northeast China), and CC (central China).The corresponding statistical analysis is listed in Table S1.It can be seen most of the scatter points gather around the 1:1 line.For all the sites, the normalized mean bias (NMB) of the average PMsurface concentration between the observation and simulation is small, with a value of ? 4.68% and correlation coefficient(

R

) of 0.45, indicating that the model can generally capture the regional variation of PM 2.5 over China.The model can also successfully reproduce the PMvariation for NC, EC, SC, SWC, NEC, and CC, with

R

ranging from 0.40–0.79.The NMB for NC, EC, SC, NEC and CC ranges from? 26.47%–19.77%.The PMconcentration is largely underestimated in NWC (NMB = ? 39.13%) and overestimated in SWC (NMB = 37.94%)by the model.Based on this comparison in 2015, the PMconcentration in 2045–2050 tends to be underestimated in NC, NWC, and NEC,but overestimated in SC and SWC.

For SO,as seen in the second panel, its distribution is like that of PM 2.5.The average SO 4? concentration is highest, with values of 5–10 μg m,over the NCP, YRD, PRD, Sichuan Basin, and central China,and ranges from 2–5 μg mover Northeast China and part of Southeast China.The pattern of the SO 4? surface concentration is consistent with the distribution of SOemissions in Fig.S2.Compared with that of RCP2.6, the SO 4? concentration increases over most areas of China under both RCP4.5 and RCP8.5.Under RCP4.5, it increases by 6–8 μg mover the NCP, PRD, YRD, Sichuan Basin, and central China,by 2–4 μg m3 over South China, and by under 2 μg m3 over other areas.This is also consistent with the difference in SOemissions between RCP4.5 and RCP2.6 shown in Fig.S3.Under RCP8.5, the SO 4? concentration increases by 4–6 μg mover the Sichuan Basin and part of central China, and by 2–4 μg m3 over the NCP, Northeast China, and South China.It decreases slightly over part of the YRD and PRD.The increase in SOover the Sichuan Basin, part of central China, Northeast China, and South China, and the decrease over part of the YRD and PRD, is consistent with the change in SOemissions over these areas.However, the increase in SO 4? over the NCP is not consistent with the decrease in SOemissions over the NCP shown in Fig.S4, which reflects the impact of climate change over this area.

The third panel shows that the concentration of NO 3is less than that of SOunder RCP2.6.The average NOconcentration is highest,with values of 5–10 μg m3,over the Sichuan Basin, part of the NCP, and central China, and then it ranges from 2–5 μg m3 over part of the YRD,Northeast China, and North China, while it is under 2 μg mover other areas.Compared with that of RCP2.6, the NOconcentration decreases slightly under RCP4.5 and increases under RCP8.5 over most areas of China.Under RCP4.5, it decreases by 1–2 μg mover part of the NCP and YRD, and by 0.5–1 μg m3 over the NCP and central China, which is consistent with the decrease in NO emissions under RCP4.5 shown in Fig.S3.Under RCP8.5, the NO 3concentration increases by 1–2 μg mover the Sichuan Basin, part of the NCP, in Northeast China, and in part of central China, and decreases slightly over the YRD.These changes are also consistent with the change in emissions under RCP8.5 shown in Fig.S4.

Fig.2.Horizontal distribution of the near-surface (first layer of the model) PM 2.5,SO 4 2 ?,NO 3 ?,NH 4 +,BC, and OC concentrations averaged for 2045–2050 under RCP2.6, and the differences in these aerosol concentrations between RCP2.6 and both RCP4.5 and RCP8.5.

For NH,as shown in the fourth panel, its pattern under RCP2.6 is very similar to that of NO 3.The average NH 4concentration is highest, with values of 5–10 μg mover the Sichuan Basin, part of the NCP,and in central China, and then it ranges from 2–5 μg mover part of the YRD, in Northeast China, and in North China, while it is under 2 μg mover other areas.The NHconcentration increases under both RCP4.5 and RCP8.5 compared with that of RCP2.6, with a higher increase under RCP8.5.Under RCP4.5, the NHconcentration increases by 1–2 μg m3 over the YRD, PRD, central and southwestern China, and by 0–1 μg mover other areas.Under RCP8.5, the NHconcentration increases by 2–3 μg m3 over Southwest China, and by 1–2 μg m3 over the NCP, northeastern, central, and southern China.As shown in Figs.S3 and S4, the NHemissions decrease slightly under RCP4.5 and RCP8.5 compared with RCP2.6, indicating the change in NH 4is more consistent with the change in SOand NOrather than the change in emissions.

The average black carbon (BC) concentration is lowest among all the aerosol species, with values of 2–4 μg mover part of the Sichuan Basin, YRD, and PRD, and under 2 μg m3 over other areas.Compared with RCP2.6, under RCP4.5 it increases by 1–2 μg mover the YRD and PRD, and by 0.5–1 μg m3 over the NCP, central and southwestern China.Under RCP8.5, it increases by 1–2 μg mover southwestern China and by 0.5–1 μg mover southern and central China, while it decreases slightly over the NCP, YRD, and PRD.These changes are consistent with the change in BC emissions under RCP4.5 and RCP8.5 shown in Figs.S3 and S4.

The average OC concentration is highest, with values of 10–15 μg m3 over the Sichuan Basin, and then it ranges from 5–10 μg m3 over the NCP, YRD, PRD, and central China, and from 2–5 μg m3 over Northeast and part of Southeast China.Under RCP4.5, compared with RCP2.6, the OC concentration decreases over most areas of China.Specifically, it decreases by more than 3 μg mover the PRD, YRD,and Sichuan Basin, and by 2–3 μg m3 over the NCP and central China.Under RCP8.5, the OC concentration decreases less than that under RCP4.5, and increases over some areas, when compared with RCP2.6.It decreases by 1–3 μg m3 over the Sichuan Basin, NCP, PRD, and YRD,and increases slightly over part of Southwest and Southeast China.These changes are consistent with the change in OC emissions under RCP4.5 and RCP8.5 shown in Figs.S3 and S4.

Fig.3.Domain-averaged (NCP, YRD, and PRD) annual near-surface PM 2.5,SO 4 2 ?,NO 3 ?,NH 4 +,BC, and OC Concentration for 2045–2050 under RCP2.6, and the differences in these aerosol Concentrations between RCP2.6 and both RCP4.5 and RCP8.5.

3.3. Annual and regional aerosol concentration for the three RCPs

Fig.3 shows the domain-averaged annual near-surface aerosol concentration for 2045–2050 under RCP2.6, and the difference in these aerosol concentrations between RCP2.6 and both RCP4.5 and RCP8.5.The aerosol concentration is averaged over the NCP, YRD, and PRD,which are the most populated and developed regions of China.From the first panel, it can be seen that, under RCP2.6, among all the aerosol species the OC concentration is highest in all three regions; and for the NCP and YRD, the BC concentration is the lowest, while for the PRD it is the NO 3concentration that is the lowest.Among the three regions, the PM,SO,and OC are highest in the YRD during the study period(2045–2050), and lowest in the PRD.Meanwhile, the NH 4and NO 3are highest in the NCP and lowest in the PRD, and the BC is highest in the YRD and lowest in the NCP.There is annual variation of all the aerosol species.For the NCP, the aerosol concentration tends to be higher in 2046 and 2049, especially for OC, while for the YRD it is higher in 2047 and 2048.For the PRD, it is higher in 2047 and 2050.As the quantity of emissions is the same for each year, the annual difference is caused by the meteorological change in different years.

In the second panel, the difference between RCP4.5 and RCP2.6 for the three regions is shown.Compared with RCP2.6, under RCP4.5 the model tends to project higher SO 4?,BC, and NH 4concentrations but lower NOand OC concentrations.This is consistent with the results in Fig.2.Accordingly, under RCP4.5, the model generally projects a higher PMconcentration from 2046–2050 over the three regions.Among the three regions, for NO 3,NH 4and BC, the change is very close, while for OC and SOtheir increase/decrease is largest for the YRD.It is noted that the difference between RCP4.5 and RCP2.6 tends to have similar temporal variation for different aerosol species, indicating that the impact of climate change on different species tends to be consistent.For the NCP, in 2046 and 2049, the increase in SO 4?,BC, and NH 4becomes smaller and the decrease in NOand OC becomes larger.Therefore, the change in PM 2.5 is smallest in 2046 and 2049.For the YRD, the smallest change is in 2048, and for the PRD it is 2045.

In the third panel, for the difference between RCP8.5 and RCP2.6,the changes of different aerosol species are generally similar to those under RCP4.5 but with a lower increase and decrease in SOand OC,respectively.The model projects higher NO 3in the NCP and YRD,which is also different from that under RCP4.5.For the NCP, under RCP8.5, the model generally projects a higher PM 2.5 concentration from 2046–2050 than RCP4.5 and RCP2.6, as a result of the lower decrease in OC and increase in NO 3.For the YRD, the change in PM 2.5 under RCP8.5 is very close to that under RCP4.5, with both the increase in SOand the decrease in OC becoming smaller.For the PRD, the increase in SO 4? is smaller than in the other two regions, and therefore the change in PMis very slight.Like that under RCP4.5, the difference between RCP8.5 and RCP2.6 also has similar temporal variation for different aerosol species.For the NCP and YRD, the change in PMis smallest in 2049, and for the PRD it is 2047.SO 4and OC are the main contributors to the PM 2.5 concentration for the period 2045–2050 in the three RCP scenarios.

4.Conclusion

A regional air quality model, RAMS–CMAQ, was used in this study to simulate the aerosol concentration for the period 2045–2050 over China based on the downscaled meteorological field of three RCP scenarios,and the regional averaged aerosol concentrations over the NCP, YRD,and PRD were investigated.

The downscaling simulation of the meteorological field of the three RCP scenarios showed that, compared with the near-surface temperature under RCP2.6, the difference in near-surface temperature between North and South China is weakened under RCP4.5 and RCP8.5, especially under RCP4.5.The change in RH is very slight between the different RCPs.Under RCP2.6, the wind speed over South China is relatively lower than that over North China.The wind speed increases over North and South China but decreases over central China under RCP4.5 and RCP8.5.

Under RCP2.6, from 2045–2050, the modeled average PM 2.5 concentration is highest, with values of 40–50 μg m,over the NCP,part of the YRD, and the Sichuan Basin, and then it is 30–40 μg mover central China and part of the PRD, and about 10–20 μg m3 over South China.Based on the evaluation of PMsurface concentration in 2015, the PMconcentration in 2045–2050 tends to be underestimated in NC, NWC, and NEC and overestimated in SC and SWC.Compared with RCP2.6, PMincreases by 4–12 μg munder both RCP4.5 and RCP8.5, of which the SO 4? and NH 4concentration increases under both RCP4.5 and RCP8.5, the NOconcentration decreases under RCP4.5 and increases under RCP8.5, the BC concentration changes very slightly, and the OC concentration decreases under RCP4.5 and RCP8.5,with some increase over part of Southwest and Southeast China under RCP8.5.There is annual variation of all the aerosol species.The difference between RCP4.5 and RCP2.6 and the difference between RCP8.5 and RCP2.6 have similar annual variation for different aerosol species,indicating that the impact of climate change on different species tends to be consistent.

Funding

This study was supported by the Strategic Priority Research Program(A) of the Chinese Academy of Sciences [grant number XDA19040204],the National Natural Science Foundation of China [grant number 41830966 ], and the Major Scientific and Technological Innovation Projects of Shandong Province [grant number 2018YFJH0901].

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi: 10.1016/j.aosl.2020.100027 .