nature geoscience

Article https://doi.org/10.1038/s41561-024-01425-1

Amplified positive effects on air quality,

health, and renewable energy under China’s
carbon neutral target
In the format provided by the
authors and unedited
 Supplementary Information

This PDF file includes:

Section 1: Supplementary Notes
Section 2: Supplementary Figures 1 to 15
Section 3: Supplementary Tables 1 to 7
Supplementary Notes
As we explained in the main text, aerosols in the atmosphere, which include particles like dust,
black carbon, sulfate and nitrate aerosols, absorb and scatter sunlight, leading to changes that
increase atmospheric stability by: (1) Reducing solar radiation reaching the ground: by blocking
sunlight, aerosols reduce the amount of incident solar energy that makes it to the Earth's surface.
This leads to a cooling effect at the surface. (2) Heating the air in the atmosphere: this effect is
primarily due to absorbing aerosols, such as black carbon. When these particles absorb sunlight,
they heat the surrounding air, contributing to a warmer atmosphere.
As a result of these processes, there is an increase in the atmosphere’s vertical stability due to
aerosols (a vertically stable structure in the atmosphere generally features a warmer top layer and
a cooler bottom layer). This increased stability is characterized by less convective motion or
turbulence in the atmosphere. Consequently, this reduces the vertical exchange among different
layers of the atmosphere. Since winds higher up (aloft) are generally faster than at the surface (due
to the friction at the Earth’s surface), the reduction in vertical motion means there is less mixing
between these fast upper-level winds and the slower surface winds. Therefore, the reduced vertical
exchange leads to a decrease in the momentum transfer from the faster winds aloft to the slower
surface winds, resulting in slower surface winds relative to the circumstances when vertical mixing
is not reduced by aerosols.
Therefore, numerical simulations based on WRF-Chem, which seamlessly couples meteorology
and chemistry and which is capable of capturing the feedbacks of changing aerosols on
meteorology1,2, are used to assess future wind and solar performance responses to achieving
carbon neutrality.
Our study is subject to several uncertainties and limitations. First, largely due to an incomplete
understanding of the atmospheric chemistry and physics mechanisms and emission inventory
uncertainties, WRF-Chem model would induce inherent uncertainties on simulated PM2.5 surface
concentrations. Validation against PM2.5 observations shows that our model simulations have good
agreements with the spatial and temporal patterns of China’s PM2.5 surface concentrations
(Supplementary Fig. 1). Second, PM2.5 associated health impacts quantification are subject to
uncertainties in relative risks estimation parameters, changing socio-economics (e.g., aging
population), and human health models applied. We therein generate 10,000 shapes of relative risks
(RR) based on the distribution of key parameters in determining RR and quantify the 95%
confidence interval uncertainty ranges of health impacts (Supplementary Tables 2 and 6,
Supplementary Fig. 10). We find both the diseases-specific GEMM model and the Integrated
Exposure-Response (IER) model identify the same trends of avoided premature deaths under
China’s carbon neutral target as the GEMM NCDs+LRIs model; in addition, aging population will
often completely offset the health benefits from China’s carbon neutrality target, though it could
be fully compensated by reduced baseline mortality rate. Third, we only demonstrate the air quality,
human health, and renewable energy co-benefits from a specific renewable energy-driven carbon
neutrality pathway in this study. In doing so we are able to reveal and characterize the amplifying
co-benefits due to aerosol-meteorology interactions and the self-reinforcing mechanisms between
renewable energy and aerosols mitigation. Future work can further explore the impacts of different
carbon neutrality pathways. Fourth, to isolate the impacts of emission reduction under China’s
carbon neutrality, we fix meteorology in the base year following early studies1,3-6. Nevertheless,
our changing meteorology sensitivity analyses identify comparable air quality, human health, and
renewable resource impacts (Supplementary Fig. 13). Likewise, there are uncertainties in
estimating wind and solar resource due to uncertainties in solar radiation and wind speed
simulation such as under different meteorology application (Supplementary Figs. 14-15), yet the
same trends of increasing solar and wind power resource and stability under China’s carbon neutral
target are often observed.
Supplementary Figure 1 | Validation of model performance on PM2.5 spatial distribution and
temporal variation. (a) Comparison of annual mean surface PM2.5 concentrations of observations
(left) and model simulations (right) across China in 2015. Model output is oversampled to the
locations of surface measurement sites. Annual mean PM2.5 concentrations for observations and
simulations averaged across China are shown in the upper left corner of each panel. Mean error
(ME) and normalized mean bias (NMB), as well as the spatial correlation coefficient (R) between
simulated and observed PM2.5 concentrations are shown in blue. Rectangles from north to south
indicate the locations of three major city clusters in China: the Beijing-Tianjin City Clusters, the
Yangtze River Delta City Clusters, and the Peral River Delta City Clusters. (b) Daily variation in
PM2.5 simulation (red) and observations (black) in three city clusters (shown in a) across China.
Dots are the daily concentrations averaged across all stations in each cluster, and the lines denote
the minimum and maximum daily concentrations of these stations. Annual mean PM2.5
concentrations for observations and simulations averaged across each city cluster are shown in the
upper left corner of each panel. Mean error (ME) and normalized mean bias (NMB), as well as the
temporal correlation coefficient (R) between simulated and observed PM2.5 concentration are
shown in blue. Base map data are from the Resource and Environment Science and Data Center,
Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences
(https://doi.org/10.12078/2023010103, 2023)7.
Supplementary Figure 2 | Resultant changes in temperatures due to radiative forcing of
atmospheric aerosols over China and their impacts on boundary layer height and ventilation
condition. (a) Averaged near-surface (i.e., 2m) air temperature changes caused by atmospheric
aerosols in China in 2015. (b) Same as (a), but for temperature differences between 1000m and
2m (i.e., T minus T ). (c) Annual averaged changes in planetary boundary layer height (PBLH)
1000m 2m

due to the radiative effects of atmospheric aerosols. (d) Annual averaged changes in ventilation
coefficient due to the radiative effects of atmospheric aerosols. Note that the ventilation coefficient
is calculated based on near-surface wind speed and the PBLH (Lu et al., 2012) . (e) and (f) are
1

relative changes in PBLH and ventilation coefficient due to the radiative effects of atmospheric
aerosols. Base map data are from the Resource and Environment Science and Data Center, Institute
of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences
(https://doi.org/10.12078/2023010103, 2023)7.
Supplementary Figure 3 | Anthropogenic CO2 and air pollutants emission pathways and
energy mix. (a) Anthropogenic CO2 and major air pollutants (e.g., SO2, NOx, and primary PM2.5)
emission pathways from 2015 to 2060 under the baseline (dashed lines) and the carbon neutral
(solid lines) scenarios. CO2, SO2, NOx, and primary PM2.5 emissions are marked in red, green,
yellow, and blue colors, respectively; (b) Energy mix under corresponding scenarios.
Supplementary Figure 4 | 2060 Provincial PM2.5 surface concentration changes in mainland
China. Area-weighted (A-W) and population-weighted (P-W) PM2.5 surface concentration
changes between 2060 Baseline scenario and 2060 Carbon neutral scenario, with (ARIon) and
without (ARIoff) considering aerosols’ radiation interactions respectively. Base map data are from
the Resource and Environment Science and Data Center, Institute of Geographic Sciences and
Natural Resources Research, Chinese Academy of Sciences (https://doi.org/10.12078/2023010103,
2023)7.
Supplementary Figure 5 | 2060 Provincial PM2.5 surface concentration changes in mainland
China due to ARI. Provincial area-weighted (A-W) and population-weighted (P-W) PM2.5
concentration changes and relative percentage changes due to aerosols’ radiation interactions
(ARI’s impacts are represented by ARIon – ARIoff). Base map data are from the Resource and
Environment Science and Data Center, Institute of Geographic Sciences and Natural Resources
Research, Chinese Academy of Sciences (https://doi.org/10.12078/2023010103, 2023)7.
Supplementary Figure 6 | Avoided provincial PM2.5 associated premature deaths in mainland
China. PM2.5 associated premature death reduction (a) with and (b) without considering aerosols’
radiation interactions (ARI) due to China’s carbon neutral target, and (c) ARI contributed
premature death reductions between 2060 Carbon neutral scenario and 2060 Baseline scenario.
Base map data are from the Resource and Environment Science and Data Center, Institute of
Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences
(https://doi.org/10.12078/2023010103, 2023)7.
Supplementary Figure 7 | Population exposure to different air pollution levels. (a) population
and (b) population share exposed to different levels of air pollution under different scenarios.
Supplementary Figure 8 | The geographic centroid of energy production and consumption
across mainland China. National centroids of mainland China’s energy production (green dot),
energy consumption (blue dot), and aerosol radiation interactions (ARI)-driven enhanced
renewable energy resource under carbon neutrality (red dot). The centroid of energy consumption
is notably closer to that of enhanced renewable energy performance than to the centroid of energy
production, indicating the potential alleviation of the geospatial mismatch between China’s energy
demand and supply with enhanced renewable energy resources under China’s carbon neutral target.
Base map data are from the Resource and Environment Science and Data Center, Institute of
Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences
(https://doi.org/10.12078/2023010103, 2023)7.
Supplementary Figure 9 | Remaining air pollutant emissions upon achieving carbon
neutrality. Major air pollutants (a) SO2, (b) NOx, and (c) primary PM2.5 emissions under the
2060 carbon neutrality scenario.
Supplementary Figure 10 | PM2.5 associated premature deaths in mainland China. PM2.5
associated premature death a) with (ARIon) and b) without (ARIon) considering aerosols’
radiation interactions under three different health models: GEMM, GEMM-5 diseases, and IER
(Integrated exposure response functions). Red dash line indicates future population aging, and
yellow dash line indicates both aging population and changes in baseline mortality rate. GEMM-
5 diseases model are constructed in Burnett et al. (2018)8 that include the same five diseases as
the IER models: ischemic heart disease (IHD), stroke, chronic obstructive pulmonary disease
(COPD), lung cancer, and lower respiratory infections (LRIs). Data are presented as mean values,
with the lower and upper bound of error bars representing the 2.5th and 97.5th percentiles,
respectively. n=10,000 for GEMM and GEMM-5 diseases functions as we generate 10,000 shapes
of relative risks; and n=1000 for IER estimations from Yang et al., 20239.
Supplementary Figure 11 | Relationships between PM2.5 surface concentrations and 10-meter
wind speed in eastern China. Each dot is daily mean values for observation stations (PM2.5 and
wind speed) in the individual city. Black lines indicate the log-linear regression relationship
between PM2.5 and wind speed. Please note this only indicates the correlation between surface
PM2.5 concentrations and 10-meter wind speed.
Supplementary Figure 12 | Model domain setup and topography distributions. Model domain
coverage in all the WRF-Chem model simulations. Base map data are from the Resource and
Environment Science and Data Center, Institute of Geographic Sciences and Natural Resources
Research, Chinese Academy of Sciences (https://doi.org/10.12078/2023010103, 2023)7.
Supplementary Figure 13 | Air quality, human health and renewable energy under 2060
meteorology. 2060CN and 2060MET represent 2060 carbon neutrality scenario with fixed
present-day meteorology and 2060 meteorology, respectively. ARIon and ARIoff indicate with
and without considering aerosols’ radiation interactions, respectively. A-W and P-W PM2.5
indicate area-weighted and population-weighted PM2.5surface concentrations, respectively. Base
map data are from the Resource and Environment Science and Data Center, Institute of Geographic
Sciences and Natural Resources Research, Chinese Academy of Sciences
(https://doi.org/10.12078/2023010103, 2023)7.
Supplementary Figure 14 | Spatial correlation between WRF-Chem simulation and ERA5
reanalysis data. The spatial correlation between WRF-Chem simulated and ERA5 reanalysis
dataset of solar radiation and wind speed for four seasons, which are generally reasonable. Base
map data are from the Resource and Environment Science and Data Center, Institute of Geographic
Sciences and Natural Resources Research, Chinese Academy of Sciences
(https://doi.org/10.12078/2023010103, 2023)7.
Supplementary Figure 15 | Temporal correlation between WRF-Chem simulation and ERA5
reanalysis data for solar radiation in China’s seven power grids. The temporal correlation
between WRF-Chem simulated (blue lines) and ERA5 reanalysis dataset (red lines) of daily solar
radiation and wind speed for four seasons, which demonstrate generally good correlation.
Supplementary Table 1. National average PM2.5 surface concentration reductions under
carbon neutrality.
Scenarios DAW*-PM2.5 DPW*-PM2.5
2060 Carbon Neutral - 2015 Baseline -16.85 (-48.4%) -47.06 (-76.2%)
2060 Carbon Neutral - 2060 Baseline -14.56 (-44.8%) -38.75 (-79.5%)
*
AW: area-weighted; PW: population-weighted
Supplementary Table 2. Scenarios-specific national PM2.5 associated premature deaths#.
ARIon ARIoff
Model Scenario
mean (2.5per. - 97.5per.) mean (2.5per. - 97.5per.)
2.45×106 2.41×106
GEMM 2015BA
(2.08×106 - 2.81×106) (2.04×106 - 2.77×106)
GEMM- 1.88×106 1.85×106
2015BA
5 diseases (1.32×106 - 2.35×106) (1.30×106 - 2.32×106)
1.54×106 1.52×106
IER 2015BA
(7.97×105 - 2.09×106) (7.85×105 - 2.07×106)
1.98×106 1.94×106
GEMM 2060BA
(1.67×106 - 2.27×106) (1.64×106 - 2.24×106)
2060BA 7.04×106 6.93×106
GEMM
(age) (5.93×106 - 8.11×106) (5.84×106 - 7.99×106)
GEMM- 1.51×106 1.49×106
2060BA
5 diseases (1.06×106 - 1.91×106) (1.04×106 - 1.88×106)
GEMM- 2060BA 5.54×106 5.45×106
5 diseases (age) (3.83×106 - 7.03×106) (3.77×106 - 6.93×106)
GEMM- 2060BA 2.04×106 2.37×106
5 diseases (age and MR) (1.58×106 - 3.12×106) (1.55×106 - 3.07×106)
1.27×106 1.26×106
IER 2060BA
(6.50×105 - 1.75×106) (6.41×105 - 1.73×106)
2060BA 4.50×106 4.45×106
IER
(age) (2.28×106 - 6.27×106) (2.25×106 - 6.21×106)
2060BA 1.69×106 1.67×106
IER
(age and MR) (8.46×105 - 2.39×106) (8.31×105 - 2.36×106)
8.42×105 8.34×105
GEMM 2060CN
(6.99×105 - 9.85×105) (6.92×105 - 9.76×105)
2060CN 3.55×106 3.52×106
GEMM
(age) (2.94×106 - 4.16×106) (2.91×106 - 4.12×106)
GEMM- 5.72×105 5.66×105
2060CN
5 diseases (4.00×105 - 7.40×105) (3.95×105 - 7.31×105)
GEMM- 2060CN 2.47×106 2.44×106
5 diseases (age) (1.70×106 - 6.93×106) (1.68×106 - 3.17×106)
GEMM- 2060CN 8.01×105 7.92×105
5 diseases (age and MR) (5.05×105 - 1.09×106) (4.99×105 - 1.07×106)
3.98×105 3.90×105
IER 2060CN
(1.71×105 - 6.87×105) (1.66×105 - 6.77×105)
2060CN 1.62×106 1.58×106
IER
(age) (6.97×105 - 2.83×106) (6.76×105 - 2.78×106)
2060CN 4.18×105 4.09×105
IER
(age and MR) (1.65×105 - 7.49×105) (1.60×105 - 7.38×105)
#
age indicates future population aging is considered; age and MR indicate both population aging and baseline
mortality rate changes are considered.
Supplementary Table 3. National average meteorology changes under carbon neutrality.
Scenarios DSolar radiation (W/m2) DWind speed (m/s)
2060 Carbon Neutral - 2015 Baseline 6.78 (3.1%) 0.0133 (0.31%)
2060 Carbon Neutral - 2060 Baseline 6.19 (2.8%) 0.0140 (0.30%)

Supplementary Table 4. WRF-Chem modeling configuration options and settings.

Domain setting
Horizontal grid 200 × 160
Grid spacing 27 km × 27 km
Number of vertical layers 37 levels from the surface to 50hPa
Center point 100 oE, 37 oN
Map projection Lambert Conformal Conic
Paramerization configuration
Long-wave radiation RRTMG Scheme
Short-wave radiation Goddard Scheme
Cumulus parameterization Grell 3D Scheme
Land-surface Noah Scheme
PBL Yonsei University Scheme
Microphysics Morrison Scheme
Photolysis Fast-J Scheme
Gas chemistry CBMZ Scheme
Aerosol chemistry MOSAIC-4bins Scheme
Supplementary Table 5. China’s inter-provincial regional power grids10.
Power grids provinces
Northeastern Heilongjiang, Jilin, Liaoning, Eastern Inner Mongolia
Northern Western Inner Mongolia, Beijing, Tianjin, Hebei, Shanxi, Shandong
Central Henan, Hubei, Chongqing, Sichuan, Hunan, Jiangxi
Eastern Shanghai, Jiangsu, Anhui, Zhejiang, Fujian
Southern Guizhou, Yunnan, Guangxi, Guangdong, Hainan
Northwestern Shaanxi, Gansu, Ningxia, Qinghai, Xinjiang, Tibet

Supplementary Table 6. Global Exposure Mortality Model (GEMM) parameters used in this
study8.
Age theta theta std alpha miu pai Cause
25 to 29 0.1585 0.01477 1.6 15.5 36.8 NCD_LRI
30 to 34 0.1577 0.0147 1.6 15.5 36.8 NCD_LRI
35 to 39 0.157 0.01463 1.6 15.5 36.8 NCD_LRI
40 to 44 0.1558 0.0145 1.6 15.5 36.8 NCD_LRI
45 to 49 0.1532 0.01425 1.6 15.5 36.8 NCD_LRI
50 to 54 0.1499 0.01394 1.6 15.5 36.8 NCD_LRI
55 to 59 0.1462 0.01361 1.6 15.5 36.8 NCD_LRI
60 to 64 0.1421 0.01325 1.6 15.5 36.8 NCD_LRI
65 to 69 0.1374 0.01284 1.6 15.5 36.8 NCD_LRI
70 to 74 0.1319 0.01234 1.6 15.5 36.8 NCD_LRI
75 to 79 0.1253 0.01174 1.6 15.5 36.8 NCD_LRI
80 plus 0.1141 0.01071 1.6 15.5 36.8 NCD_LRI
Supplementary Table 7. Variables used for solar and wind power estimation

Variables Description Units

I Downward shortwave radiation W/m2
T2m Air temperature at 2 meters above ground ℃
u10m Wind speed at 10 meters above ground m/s
u100m Wind speed at 100 meters above ground m/s
uin Cut-in speed for the wind turbine (3m/s) m/s
uout Cut-out speed for the wind turbine (25m/s) m/s
urated Rated speed for the wind turbine (12m/s) m/s
ISTC Shortwave radiation on PV panels under standard test conditions W/m2
（1000W/m2）
PR Performance ratio 1
Tcell Cell temperature ℃
TSTC Cell temperature under standard test conditions（25℃） ℃
γ 0.005 -1
℃
c1 4.3 ℃
c2 0.943 1
c3 0.028 ℃/(W/m2)
c4 1.528 ℃/(m/s)
CF Capacity factor 1
CV Coefficient of variation 1
i Latitudinal position of the grid point 1
j Longitudinal position of the grid point 1
k Moment of the grid point 1
References
1 Zhou, M. et al. Environmental benefits and household costs of clean heating options in
northern China. Nature Sustainability 5, 329-+ (2022).
2 Qin, Y. et al. Environmental Consequences of Potential Strategies for China to Prepare
for Natural Gas Import Disruptions. Environ Sci Technol 56, 1183-1193 (2022).
3 Qin, Y. et al. Air quality, health, and climate implications of China's synthetic natural gas
development. P Natl Acad Sci USA 114, 4887-4892 (2017).
4 Qin, Y. et al. Air quality–carbon–water synergies and trade-offs in China's natural gas
industry. Nature Sustainability 1, 505-511 (2018).
5 Tang, R. et al. Air quality and health co-benefits of China's carbon dioxide emissions
peaking before 2030. Nature Communications 13, 1008 (2022).
6 Shi, X. R. et al. Air quality benefits of achieving carbon neutrality in China. Sci Total
Environ 795, 148784 (2021).
7 Xu, X. China's multi-year provincial administrative division boundary data. The
Resource and Environment Science and Data Center, Institute of Geographic Sciences
and Natural Resources Research, Chinese Academy of Sciences.
(http://www.resdc.cn/DOI), DOI:10.12078/2023010103. (2023).
8 Burnett, R. et al. Global estimates of mortality associated with long-term exposure to
outdoor fine particulate matter. P Natl Acad Sci USA 115, 9592-9597 (2018).
9 Yang, H., Huang, X. Y., Westervelt, D. M., Horowitz, L. & Peng, W. Socio-demographic
factors shaping the future global health burden from air pollution. Nature Sustainability
6, 58-+, doi:10.1038/s41893-022-00976-8 (2023).
10 Zhou, X. X. et al. An overview of power transmission systems in China. Energy 35,
4302-4312 (2010).