Exploring Beijing's urban environment: air quality, haze effects and climate

Towards the end of the last century, significant increases in coal burning for electricity at power plants and for domestic heating purposes led to rapid increases in the levels of harmful gaseous and particulate pollutants in the atmosphere over urban areas in China. The concurrent rise in urban populations has resulted in an ever-increasing human health risk associated with the inhalation of ambient air pollution. Most notably, long- and short-term exposure to PM2.5, describing fine particles suspended in the atmosphere with a diameter ≤ 2.5 μm, is the fourth leading cause of premature death in China. Additionally, very high ambient concentrations of PM2.5 characterising haze pollution events can block a large fraction of incoming sunlight, impacting atmospheric chemistry and reducing heating of the urban surface which drives the mixing of near-surface air. Both of these effects can substantially impact urban air quality. Development of urban areas is also associated with the replacement of vegetative, rural surfaces with urban materials (e.g. concrete). This increases the amount of sunlight absorbed at the ground, reduces cooling associated with evaporation, and thus urban air temperatures are higher than adjacent rural areas. Heat released by human activities (anthropogenic heat) directly into the atmosphere further strengthens the urban -rural temperature contrast (known as the urban heat island). Extreme urban air temperatures, particularly in summer, reduce human thermal comfort and increase the heat-related mortality risk. The main objectives of this study are to (a) quantify the air pollution levels in Beijing in winter at street-scale resolution and understand the key influencing processes, (b) quantify the effects of haze-sunlight interactions on Beijing’s air quality in winter, and (c) understand the relative importance of urban materials and anthropogenic heat emissions in forming Beijing’s urban heat island in winter and summer. For the air quality-related objectives of this work, we use a computer-based model (ADMS-Urban) that replicates the release of air pollutants from a range of activities in urban areas, and accounts for the effects of local weather and chemical reactions on air pollution levels. The ADMS-Urban model can estimate air pollution level variations across very fine spatial extents (< 10 m). Comparisons between modelled and measured air pollutant concentrations inform on the model performance, revealing the success of our model configuration and whether further adjustments are required. To study Beijing’s climate, we use a recently developed variation of the ADMS-Urban model, the ADMS-Urban Temperature and Humidity model (ADMS-UHI). ADMS-UHI is used to calculate air temperature variations across neighbourhood extents (~ 100 m) due to surface thermal characteristics and anthropogenic heat release. The air quality model simulations require estimates of emission rates from roads. For this, we share emissions estimated at grid-level (3 km) resolutions across individual road sections; major roads (e.g. motorway) are apportioned more than minor roads (e.g. residential). Measured air pollutant concentrations are much higher than modelled levels at night, suggesting the emissions estimates are missing a significant evening source; this is most likely related to heavy duty diesel trucks that enter the city at night when travel restrictions are lifted. The model cannot replicate the measured PM2.5 spatial variability. This is likely due to the exclusion of emissions from explicitly represented point sources (e.g. large industry) in the model configuration due to a lack of information on their locations. Also, we assume background air pollutant levels are uniform across Beijing, which is unlikely across such a large urban area. Overall, the measured-modelled air pollutant concentration agreement is strong, indicating that the methodologies and materials implemented for this study can be successfully applied to urban areas elsewhere to help guide air pollution control policies. According to a number of previous modelling studies across China, haze-sunlight interactions have substantial effects on urban air quality. Therefore, to improve the accuracy of urban air quality assessments with ADMS-Urban across polluted megacities such as Beijing, the model, which currently does not account for the impacts of haze, should be adjusted accordingly. Therefore, we alter the modelled air pollution dispersion and chemistry processes to account for reduced sunlight reaching the surface on haze days and analyse the impacts on simulated near-surface air quality. Reduced surface heating on haze days suppresses the vertical mixing of air and therefore increases the modelled concentrations of primary air pollutants such as nitrogen oxides (NOx = NO + NO2). However, near-surface levels of the secondary air pollutant species O₃ reduces substantially due to (a) increased removal from reactions with NO, and (b) suppressed formation via sunlight interactions with NO₂. Measured-modelled O₃ agreement improves significantly when haze effects are considered. The final objective of this thesis is explored with ADMS-UHI. During the winter, the impact of anthropogenic heat emissions on the UHI far outweighs the effects of urban surface thermal properties. UHI intensities (urban-rural temperature difference) are much greater at night than during the day in both winter and summer. In summer, this is mostly related to the evening release of heat stored in the urban fabric throughout the day. Negative daytime UHI intensities in summer highlight the cooling effect of urban green space. Overall, this study provides useful information for urban planners aiming to reduce heat-related health risks across complex urban areas such as Beijing.