Air quality and health benefits from ultra-low emission control policy indicated by continuous emission monitoring: A case study in the Yangtze River Delta region, China

To evaluate the improved emission estimates from online monitoring, we applied the Models-3/CMAQ (Community Multiscale Air Quality) system to simulate the air quality of the Yangtze River Delta (YRD) region using two emission inventories with and without incorporated data from continuous emission monitoring systems (CEMSs) at coal-fired power plants (cases 1 and 2, respectively). The normalized mean biases (NMBs) between the observed and simulated hourly concentrations of SOspan classCombining double low line"inline-formula"2/span, NOspan classCombining double low line"inline-formula"2/span, Ospan classCombining double low line"inline-formula"3/span, and PMspan classCombining double low line"inline-formula"2.5/span in case 2 were span classCombining double low line"inline-formula"-3.1/span %, 56.3 %, span classCombining double low line"inline-formula"-19.5/span %, and span classCombining double low line"inline-formula"-1.4/span %, all smaller in absolute value than those in case 1 at 8.2 %, 68.9 %, span classCombining double low line"inline-formula"-24.6/span %, and 7.6 %, respectively. The results indicate that incorporation of CEMS data in the emission inventory reduced the biases between simulation and observation and could better reflect the actual sources of regional air pollution. Based on the CEMS data, the air quality changes and corresponding health impacts were quantified for different implementation levels of China's recent "ultra-low" emission policy. If the coal-fired power sector met the requirement alone (case 3), the differences in the simulated monthly SOspan classCombining double low line"inline-formula"2/span, NOspan classCombining double low line"inline-formula"2/span, Ospan classCombining double low line"inline-formula"3/span, and PMspan classCombining double low line"inline-formula"2.5/span concentrations compared to those of case 2, our base case for policy comparisons, would be less than 7 % for all pollutants. The result implies a minor benefit of ultra-low emission control if implemented in the power sector alone, which is attributed to its limited contribution to the total emissions in the YRD after years of pollution control (11 %, 7 %, and 2 % of SOspan classCombining double low line"inline-formula"2/span, NOspan classCombining double low line"inline-formula"iX/i/span, and primary particle matter (PM) in case 2, respectively). If the ultra-low emission policy was enacted at both power plants and selected industrial sources including boilers, cement, and iron and steel factories (case 4), the simulated SOspan classCombining double low line"inline-formula"2/span, NOspan classCombining double low line"inline-formula"2/span, and PMspan classCombining double low line"inline-formula"2.5/span concentrations compared to the base case would be 33 %-64 %, 16 %-23 %, and 6 %-22 % lower, respectively, depending on the month (January, April, July, and October 2015). Combining CMAQ and the Integrated Exposure Response (IER) model, we further estimated that 305 deaths and 8744 years of life loss (YLL) attributable to PMspan classCombining double low line"inline-formula"2.5/span exposure could be avoided with the implementation of the ultra-low emission policy in the power sector in the YRD region. The analogous values would be much higher, at 10 651 deaths and 316 562 YLL avoided, if both power and industrial sectors met the ultra-low emission limits. In order to improve regional airspan idCombining double low line"page6412"/ quality and to reduce human health risk effectively, coordinated control of multiple sources should be implemented, and the ultra-low emission policy should be substantially expanded to major emission sources in industries other than the power industry./p.