Ozone pollution in bay areas of China : investigation on atmospheric dynamics and photochemistry with field observations and model simulations

Abstract

Due to the rapid industrialization and urbanization in recent decades, ozone (O3) pollution has become an intractable environmental problem in China. The occurrence of intensive photochemical reactions is conventionally acknowledged in the continental regions with substantial anthropogenic emissions. However, severe O3 pollution has been reported over the bay areas of China, including Bohai Bay (BHB), Hangzhou Bay (HZB) and Pearl River Estuary (PRE), despite sparse anthropogenic emissions. As connection areas between city agglomerations and marine environment, the three bays suffer from not only inconstant dynamic circulations but also complicated chemical conditions with polluted plumes transported from continent together with the aged air masses from marine environment, making the photochemical pollution in bay areas intricate. This thesis aims to understand (1) the dynamic and chemical processes for the build-up of O3 over the three bay areas of China, (2) the interaction of photochemical pollution between the two banks of the PRE, and (3) the long-term variation of O3 and its underlying causes in Hong Kong, a coastal city within the PRE. To understand the causes leading to the high O3 concentrations over BHB, HZB and PRE, O3 and its precursors from 2016 to 2018 were collected and analyzed. Due to the influence of the Asian monsoon, high concentrations of O3 were found in these bays in different seasons. In BHB, thirteen high O3 episode days (hourly O3 concentration exceeding 100 ppbv at no less than one suburban sites in each of the seven, six, and seven coastal cities of BHB, HZB and PRE, respectively) were dominated by uniform pressure fields. Eleven O3 episode days over HZB were associated with high pressure systems, West Pacific Subtropical High and the cold front, while twelve O3 episode days over PRE were related to continental anticyclones and typhoons. The simulated results revealed that the interaction between synoptic winds and mesoscale breezes resulted in slow wind speeds over HZB and PRE, where the substantial air pollutants transported from coastal cities gained a long residence time and subsequently participated in intensive photochemical reactions. Consequently, the O3 production rates within the bay areas were even comparable to those in surrounding cities which experienced much stronger anthropogenic emissions. This finding was also applicable to BHB, but the O3 production rates were lower and the contributions of man-made emissions were less. Moreover, the regional transport from the farther upwind areas of the BHB partially elevated the O3 concentrations. Compared with HZB and PRE, O3 pollution was intensified by less dry deposition over the BHB. Hence, these three bay areas served as O3 "pools" which caused the accumulation of air pollutants via atmospheric dynamics and subsequent intense photochemical reactions under certain meteorological conditions. To further explore the photochemical pollution over PRE and the interaction between two banks of the estuary, intensive measurements of O3 and its precursors were simultaneously conducted at the east (TC) and west (QAI) banks of PRE from Sept. to Nov. of 2016. On the three days when O3 episodes (days with the maximum hourly O3 higher than 100 ppbv) occurred at TC, QAI also suffered from O3 episodes, which were all associated with typhoons. As a downwind site of PRE when northwesterly winds prevailed, TC displayed a higher maximum hourly O3 than QAI on these three days. In contrast, the other thirteen episodes that solely occurred at QAI were generally under northeasterly winds, which brought more coastal air to TC and led to lower levels of O3 at this site. In addition, the stronger titration of the locally emitted NO to O3 also accounted for the lower O3 at TC on these days. The simulated results indicated that the interactions of synoptic winds and mesoscale breezes led to relatively weak winds and long residence time for air pollutants over the PRE, thereby air pollution accumulation. This provided the likelihood of intensive chemical productions of O3 over PRE, with the production rates even higher than those in the surrounding land areas. Moreover, to investigate the temporal variation of O3 in the coast of bay areas, the interannual and seasonal variations of surface O3 from 2005 to 2017 in Hong Kong, a coastal city within the PRE, were examined. O3 pollution has been a persistent problem in Hong Kong, particularly in autumn when severe O3 pollution events are often observed. In this study, analysis of long-term O3 data revealed that the variation of autumn O3 obviously leveled off during 2005-2017, mainly due to the significant decrease of autumn O3 in 2013-2017 (period II), despite the increase in 2005-2012 (period I). In addition, the rise of O3 in summer and winter also ceased since 2013. In contrary, O3 continuously increased throughout the spring of 2005-2017, especially in period II. Consequently, an incessant increase of overall O3 was observed during 2005-2017. The annually increased O3 from 2005 to 2017 was mainly enhanced by the local O3 production due to the reduction of NOx. Apart from the local production, the meteorological conditions and regional transport facilitated the O3 growth in period Ι. In contrast, the unchanged O3 level in period II was attributable to the low contributions of meteorology and regional transport to O3 formation and accumulation, as well as the insignificant change in local production. Seasonally, the reduction of NOx was the main driving force for the remarkable rise of spring O3, especially in the spring of period II. Furthermore, although meteorological conditions elevated the overall O3 in summer and winter, the contributions of regional transport and local production in summer and winter of period II to O3 concentration were negative, respectively. Moreover, the decrease of O3 in the autumn of period II was caused by the inhibitory meteorological conditions, reduced regional transport, and alleviated local production. It is noteworthy that the increased spring O3 and decreased autumn O3 in period II were mainly because 1) the TVOC concentration significantly decreased in the autumn while remained stable in the spring of period II; and 2) the autumn NO kept steady in period II (p = 0.32), whereas the spring NO remarkably declined in this period. Thus, the reduced NOx have less impact on autumn O3. This was further confirmed by the simulated results, in which the difference in O3 trends between the base and the unchanged NOx scenarios in autumn of period II was insignificant (p = 0.49). To sum up, these studies primarily unraveled the intensive O3 pollution and its formation mechanisms over bay areas in China, which served as O3 "pools" that led to the accumulation of air pollutants through atmospheric dynamics and subsequent intensive photochemical reactions. The findings may be applicable to other similar ecotones around the world. In addition, the driving forces for the interannual and seasonal variations of O3 discovered in Hong Kong may be used as a guide for future O3 mitigation measures in cities of bay areas.