An investigation of potential regional and local source regions affecting fine particulate matter concentrations in Delhi, India

In this study, potential regional and local sources influencing PM2.5 (particulate matter with an aerodynamic diameter >2.5 mu m) concentrations in Delhi, India, are identified and their possible impact evaluated through diverse approaches based on study of variability of synoptic and local airflow patterns that transport aerosol concentrations from these emission sources to an urban receptor site in Delhi, India. Trajectory clustering of 72-hr and 48-hr back trajectories simulated at arrival heights of 500 m and 100 m, respectively, every hour for representative years 2008-2010 are used to assess the relative influence of long-distance, regional, and subregional sources on this site. Nonparametric statistical procedures are employed on trajectory clusters to better delineate various distinct regional pollutant source regions. Trajectory clustering and concentration-weighted trajectory (CWT) analyses indicate that regional and subregional PM2.5 emission sources in neighboring country of Pakistan and adjacent states of Punjab, Haryana, and Uttar Pradesh contribute significantly to the total surplus of aerosol concentrations in the Delhi region. Conditional probability function and Bayesian approach used to identify local source regions have established substantial influence from highly urbanized satellite towns located southwest (above 25%) and southeast (above 45%) of receptor location. There is significant seasonal variability in synoptic and local air circulation patterns, which is discerned in variability in seasonal concentrations. Mean of daily averaged PM2.5 concentrations at the Income Tax Office (ITO) receptor site over Delhi at 95% confidence level is highest in winter, ranging between 209 and 185 mu g m(-3) for the entire study period. The annual variability in air transport pathways is more in winter than in other seasons. Year-to-year variability is present in aerosol concentrations, especially during winter, with standard deviations varying from a minimum of 60 mu g m(-3) in winter 2009 to a maximum of 109 mu g m(-3) in winter 2010.