TY - JOUR
T1 - Estimates of Health Impacts and Radiative Forcing in Winter Haze in Eastern China through Constraints of Surface PM2.5 Predictions
AU - Gao, Meng
AU - Saide, Pablo E.
AU - Xin, Jinyuan
AU - Wang, Yuesi
AU - Liu, Zirui
AU - Wang, Yuxuan
AU - Wang, Zifa
AU - Pagowski, Mariusz
AU - Guttikunda, Sarath K.
AU - Carmichael, Gregory R.
PY - 2017/2/21
Y1 - 2017/2/21
N2 - The Gridpoint Statistical Interpolation (GSI) Three-Dimensional Variational (3DVAR) data assimilation system is extended to treat the MOSAIC aerosol model in WRF-Chem, and to be capable of assimilating surface PM2.5 concentrations. The coupled GSI-WRF-Chem system is applied to reproduce aerosol levels over China during an extremely polluted winter month, January 2013. After assimilating surface PM2.5 concentrations, the correlation coefficients between observations and model results averaged over the assimilated sites are improved from 0.67 to 0.94. At nonassimilated sites, improvements (higher correlation coefficients and lower mean bias errors (MBE) and root-mean-square errors (RMSE)) are also found in PM2.5, PM10, and AOD predictions. Using the constrained aerosol fields, we estimate that the PM2.5 concentrations in January 2013 might have caused 7550 premature deaths in Jing-Jin-Ji areas, which are 2% higher than the estimates using unconstrained aerosol fields. We also estimate that the daytime monthly mean anthropogenic aerosol radiative forcing (ARF) to be −29.9W/m2 at the surface, 27.0W/m2 inside the atmosphere, and −2.9W/m2 at the top of the atmosphere. Our estimates update the previously reported overestimations along Yangtze River region and underestimations in North China. This GSI-WRF-Chem system would also be potentially useful for air quality forecasting in China.
AB - The Gridpoint Statistical Interpolation (GSI) Three-Dimensional Variational (3DVAR) data assimilation system is extended to treat the MOSAIC aerosol model in WRF-Chem, and to be capable of assimilating surface PM2.5 concentrations. The coupled GSI-WRF-Chem system is applied to reproduce aerosol levels over China during an extremely polluted winter month, January 2013. After assimilating surface PM2.5 concentrations, the correlation coefficients between observations and model results averaged over the assimilated sites are improved from 0.67 to 0.94. At nonassimilated sites, improvements (higher correlation coefficients and lower mean bias errors (MBE) and root-mean-square errors (RMSE)) are also found in PM2.5, PM10, and AOD predictions. Using the constrained aerosol fields, we estimate that the PM2.5 concentrations in January 2013 might have caused 7550 premature deaths in Jing-Jin-Ji areas, which are 2% higher than the estimates using unconstrained aerosol fields. We also estimate that the daytime monthly mean anthropogenic aerosol radiative forcing (ARF) to be −29.9W/m2 at the surface, 27.0W/m2 inside the atmosphere, and −2.9W/m2 at the top of the atmosphere. Our estimates update the previously reported overestimations along Yangtze River region and underestimations in North China. This GSI-WRF-Chem system would also be potentially useful for air quality forecasting in China.
UR - https://doi.org/10.1021/acs.est.6b03745
U2 - 10.1021/acs.est.6b03745
DO - 10.1021/acs.est.6b03745
M3 - Article
SN - 0013-936X
VL - 51
SP - 2178
EP - 2185
JO - Environmental Science & Technology
JF - Environmental Science & Technology
IS - 4
ER -