TY - JOUR
T1 - Spatial Resolved Surface Ozone with Urban and Rural Differentiation during 1990-2019
T2 - A Space-Time Bayesian Neural Network Downscaler
AU - Sun, Haitong
AU - Shin, Youngsub Matthew
AU - Xia, Mingtao
AU - Ke, Shengxian
AU - Wan, Michelle
AU - Yuan, Le
AU - Guo, Yuming
AU - Archibald, Alexander T.
N1 - Funding Information:
This study is funded by UK Natural Environment Research Council (NERC) and UK National Centre for Atmospheric Science (NCAS). H.S. and M.W. receive funding from Engineering and Physical Sciences Research Council (EPSRC) via the UK Research and Innovation (UKRI) Centre for Doctoral Training in Application of Artificial Intelligence to the study of Environmental Risks (AI4ER, EP/S022961/1). ATA acknowledges funding from NERC (NE/P016383/1) and through the Met Office UKRI Clean Air Programme. Y.G. is supported by a Career Development Fellowship of the Australian National Health and Medical Research Council (APP1163693). We thank Ushnish Sengupta (University of Cambridge) and Matt Amos (Lancaster University) for sharing their demonstration codes in Python. We are also grateful to the editors and three anonymous reviewers for their insightful revision comments to help us substantially improve the manuscript.
Publisher Copyright:
© 2021 American Chemical Society.
PY - 2022/6/7
Y1 - 2022/6/7
N2 - Long-term exposure to ambient ozone (O3) can lead to a series of chronic diseases and associated premature deaths, and thus population-level environmental health studies hanker after the high-resolution surface O3 concentration database. In response to this demand, we innovatively construct a space-time Bayesian neural network parametric regressor to fuse TOAR historical observations, CMIP6 multimodel simulation ensemble, population distributions, land cover properties, and emission inventories altogether and downscale to 10 km × 10 km spatial resolution with high methodological reliability (R2 = 0.89-0.97, RMSE = 1.97-3.42 ppbV), fair prediction accuracy (R2 = 0.69-0.77, RMSE = 5.63-7.97 ppbV), and commendable spatiotemporal extrapolation capabilities (R2 = 0.62-0.76, RMSE = 5.38-11.7 ppbV). Based on our predictions in 8-h maximum daily average metric, the rural-site surface O3 are 15.1±7.4 ppbV higher than urban globally averaged across 30 historical years during 1990-2019, with developing countries being of the most evident differences. The globe-wide urban surface O3 are climbing by 1.9±2.3 ppbV per decade, except for the decreasing trends in eastern United States. On the other hand, the global rural surface O3 tend to be relatively stable, except for the rising tendencies in China and India. Using CMIP6 model simulations directly without urban-rural differentiation will lead to underestimations of population O3 exposure by 2.0±0.8 ppbV averaged over each historical year. Our original Bayesian neural network framework contributes to the deep-learning-driven environmental studies methodologically by providing a brand-new feasible way to realize data fusion and downscaling, which maintains high interpretability by conforming to the principles of spatial statistics without compromising the prediction accuracy. Moreover, the 30-year highly spatial resolved monthly surface O3 database with multiple metrics fills in the literature gap for long-term surface O3 exposure tracing.
AB - Long-term exposure to ambient ozone (O3) can lead to a series of chronic diseases and associated premature deaths, and thus population-level environmental health studies hanker after the high-resolution surface O3 concentration database. In response to this demand, we innovatively construct a space-time Bayesian neural network parametric regressor to fuse TOAR historical observations, CMIP6 multimodel simulation ensemble, population distributions, land cover properties, and emission inventories altogether and downscale to 10 km × 10 km spatial resolution with high methodological reliability (R2 = 0.89-0.97, RMSE = 1.97-3.42 ppbV), fair prediction accuracy (R2 = 0.69-0.77, RMSE = 5.63-7.97 ppbV), and commendable spatiotemporal extrapolation capabilities (R2 = 0.62-0.76, RMSE = 5.38-11.7 ppbV). Based on our predictions in 8-h maximum daily average metric, the rural-site surface O3 are 15.1±7.4 ppbV higher than urban globally averaged across 30 historical years during 1990-2019, with developing countries being of the most evident differences. The globe-wide urban surface O3 are climbing by 1.9±2.3 ppbV per decade, except for the decreasing trends in eastern United States. On the other hand, the global rural surface O3 tend to be relatively stable, except for the rising tendencies in China and India. Using CMIP6 model simulations directly without urban-rural differentiation will lead to underestimations of population O3 exposure by 2.0±0.8 ppbV averaged over each historical year. Our original Bayesian neural network framework contributes to the deep-learning-driven environmental studies methodologically by providing a brand-new feasible way to realize data fusion and downscaling, which maintains high interpretability by conforming to the principles of spatial statistics without compromising the prediction accuracy. Moreover, the 30-year highly spatial resolved monthly surface O3 database with multiple metrics fills in the literature gap for long-term surface O3 exposure tracing.
KW - CMIP6
KW - downscaling
KW - environmental justice
KW - space-time Bayesian neural network
KW - surface ozone
UR - http://www.scopus.com/inward/record.url?scp=85119455965&partnerID=8YFLogxK
U2 - 10.1021/acs.est.1c04797
DO - 10.1021/acs.est.1c04797
M3 - Article
C2 - 34751030
AN - SCOPUS:85119455965
SN - 0013-936X
VL - 56
SP - 7337
EP - 7349
JO - Environmental Science and Technology
JF - Environmental Science and Technology
IS - 11
ER -