Urban meteorology and air quality in a rapidly growing city: Inter-parameter associations and intra-urban heterogeneity

Abstract

Urban air quality is highly dynamic and influenced by macro-, meso- and micro-meteorological conditions as well as urban growth patterns. This study investigates the inter-correlation between meteorological variables and pollutant concentrations in the city of Sydney (Australia) through a below-canopy, mid-priced, and citizen-centred urban network. Year-round observations of 12 different parameters are conducted at 12 different urban sites on account of intra-urban variability, revealing that coastal, but highly urbanized areas tend to accumulate heat and pollutants, reaching the highest median levels. Mutual information analysis is performed to identify linear, non-linear and conditioned correlations, disclosing that: i) seasonal cycles are critical in shaping the inter-correlations between meteorological and air quality parameters, yet anthropogenic mechanisms may take over in the local presence of extensive and compact built features, ii) strong associations exist between temperature and NO2, relative humidity and PM2.5, and wind speed and CO, iii) these interactions are not static as their nature and strength varies in time and space, modulated by the urban metabolism. This study further demonstrates the need for standardized metadata which not only describes the instrument and site main characteristics, but provides details and routine updates on changes and urban developments, critical to decipher otherwise unexplainable correlations.