Real Time Source Apportionment in Four Chinese Cities Integrating Data from the AXA (ACSM-Xact-Aethalometer) Set Up

OLGA ZOGRAFOU, Manousos Ioannis Manousakas, Yuemei Han, Shaofei Kong, Yang Chen, Shuyan Xing, Jay G. Slowik, Kaspar R. Daellenbach, Rui Wang, Ningning Zhang, Qiyuan Wang, Jie Tian, Zhen Yang, Lei Cao, Fei Gao, Chongzhi Zhai, Zhenliang Li, Chao Peng, Feng Ding, Linjun Li, Konstantinos Eleftheriadis, Junji Cao, André S. H. Prévôt, NCSR Demokritos, Athens, Greece

     Abstract Number: 327
     Working Group: Source Apportionment

Abstract
Real-time source apportionment (RT SA) has emerged as a powerful tool for the identification of atmospheric Particulate Matter (PM) pollution sources under real-time conditions offering the advantage of immediate response to air quality threats. In the framework of the Sino-Swiss Cooperation on Air Pollution Source Apportionment for Better Air project, an RT SA software was developed to meet the demands of environmental authorities, researchers, and policymakers for real-time access on ambient PM sources. Online real-time instruments outclass the traditional offline methods, offering higher temporal resolution, enabling insights into the daily patterns of pollutants. Similarly, RT SA analysis outclasses the traditional offline SA methods which usually take up to months until post-measurements comprehensive results are obtained, even in the case of real-time online measurements. While valuable information in the long haul can be obtained from offline SA methods, RT SA offers access to real-time knowledge and the ability to take immediate actions.

The RT SA software integrates continuous data streams from a suite of online monitoring instruments collectively consisting the AXA set-up, including the Aerosol Chemical Speciation Monitor (ACSM), the Aethalometer, and the Xact elemental analyzer. These instruments provide high-resolution complementary chemical information, including organic aerosol (OA), inorganic ions (sulphate, ammonium, nitrate, and chloride), equivalent Black Carbon (eBC), and elemental composition.

For comparison, a second model, SoFi RT, served as the reference approach. They operate continuously at monitoring stations in Xi’An and Wuhan, China, where the required instruments are installed. After an initial seven-day calibration, both models deliver source apportionment results for total PM within minutes of each measurement.

The accuracy of the two models was evaluated by comparing their outputs to results from advanced, well-established SA techniques. The comparison showed relatively low uncertainty, supporting the reliability of these models as effective tools for real-time air quality monitoring and pollution mitigation.