High-density volatile organic compound monitoring network for identifying pollution sources

Zehui Li,Zizhen Ma,Zhan Zhang,Lingling Zhang,Enze Tian,Haiteng Zhang,Ruiyao Yang,Diwei Zhu,Hui Li,Ziyi Wang,Yinglei Zhang,Pingchuan Xu,Yuexin Xu,Dongbin Wang,Gang Wang,Minjung Kim,Yi Yuan,Xiaohui Qiao,Mingjie Li,Yangyang Xie,Shaojun Guo,Kaihui Liu,Jingkun Jiang

DOI: https://doi.org/10.1016/j.scitotenv.2022.158872

2023-01-01

Abstract:The elusive sources of air pollution have hampered effective control across all sectors, with long-term consequences for the greenhouse effect and human health. Multiple monitoring systems have been highly desired for locating the sources. However, when faced with extensive sources, diverse air environments and meteorological conditions, the low spatiotemporal resolution, poor reliability and high cost of existing monitors were significant obstacles to their applications. Extending our previous demonstration of sensitive and reliable electrochemical sensors, we here present a machine-learning-assisted sensor arrays for monitoring typical volatile organic compounds (VOCs), which shows the consistent response with gas chromatography-mass spectrometry in the actual air environment. As a proof-of-concept, a low-cost and high-resolution VOC network of 152 sets of monitors across ~55 km² of mixed-used land is established in southwest Beijing. Benefiting from the strong reliability, the pollution sources are revealed by the VOC network and supported by the joint mobile sampling of a vehicle-mounted gas chromatography-mass spectrometry system. With the sustained help of the network, the sources polluted by the local industrial facilities, traffic, and restaurants are effectively site-specific abatement by the local authorities and enterprises during the next half-year. Our findings open up a promising path toward more effective tracing of regional pollution sources, as well as accelerate the long-term transformation of industry and cities.

environmental sciences