Mitigation and control of urban air pollution in Beijing

Ansari, Tabish and Wild, Oliver (2020) Mitigation and control of urban air pollution in Beijing. PhD thesis, Lancaster University.

Text (2020tabishphd)
2020tabishphd.pdf - Published Version
Available under License Creative Commons Attribution-NonCommercial-NoDerivs.
Download (30MB)

Abstract

Outdoor air pollution kills more than 4 million people around the world per year and heavy pollution episodes continue to occur especially around large urban centres. In addition to long-term mitigation strategies, short-term emission controls are needed to prevent heavy pollution episodes in megacities such as Beijing. Such controls have been implemented with reasonable success in the past, notably during mega-events, but need to be carefully evaluated to develop a robust mitigation strategy for future. In this work, the 10-day long controls implemented before the APEC summit in Beijing during November 2014 were evaluated for their eectiveness using an online atmospheric chemical transport model WRF-Chem. The controls were found to be only partly responsible for the improvement in air quality during the summit period, while the rest of the improvement was due to favourable meteorology which reduced pollutant levels signicantly as compared to the levels before the control period. The controls were found to be insucient in meeting national air quality standards if applied during periods with more stagnant conditions. Sensitivity studies were performed to identify temporally-resolved source contributions from various sectors and regions. It was found that controls on local emissions benet air quality on the same day, controls on regional emissions show peak benets a day or two after the start of controls and controls on distant emissions show peak benets three to four days later. Local and regional residential and industry sectors were found to dominate contributions to PM2.5 levels in Beijing. A Gaussian statistical technique was used to replace the model behaviour over Beijing with a fast emulator to generate concentration response surfaces for emission reductions across various sectors and regions. These results were utilized to develop an optimal policy for short-term emission controls in Beijing and were implemented in an automatic air quality forecasting and emission prescription system which runs the model successively with reduced emissions to meet daily air quality targets and outputs the magnitude and timing of controls needed across various sectors and regions to prevent heavy pollution episodes in Beijing. This is a novel application of the popular method called Model Predictive Control often used in the petrochemical industry, to air quality modelling. The framework developed here is for Beijing but can be readily adopted for any other polluted region of the world.