Abstract
The temporal granularity of data in receptor models plays a key role in identifying emission patterns and episodic pollution events, which are essential for robust source apportionment. Low-time resolution measurements may overlook short-term variations, leading to an incomplete representation of pollution sources. This study investigated the source characteristics and contributions of PM2.5 across different time resolutions (1-h, 2-h, 4-h, 8-h, 12-h, and 24-h) and seasons, using multi-time-season-resolved chemical composition data collected at Major Dhyan Chand National Stadium in Delhi from May 2019 to February 2020. Positive Matrix Factorization (PMF) identified eight consistent factors across all time resolutions and seasons: two solid fuel combustion sources (SFC1 and SFC2), an S-rich source, traffic (exhaust and non-exhaust), dust, and three anthropogenic industrial/combustion plume events (Cl-Br, Pb-Se, and Cu-Cd). The comparison of source profiles and contributions at different time resolutions revealed that SFC1 exhibited the highest temporal variability across all seasons, followed by Cl-Br, traffic, Pb-Se, and Cu-Cd, while S-rich and dust factors remained relatively stable. The variation in source profiles over time, influenced by species mixing, posed challenges for source identification. A sensitivity analysis using the coefficient of divergence (CoD) showed that heterogeneity in source profiles increased with coarser time resolution, indicating the need for high-resolution data to capture dynamic source variations. Given these findings, a rolling-PMF approach with high-resolution data could improve real-time source apportionment by updating source profiles at regular intervals to reflect ever-changing emissions sources better. This study highlights the importance of high-resolution data in achieving accurate and temporally resolved source apportionment, essential for air quality management and policy development.
[Display omitted]
•Identified scenarios where the characteristics of particulate matter sources vary•PMF analysis underscores the challenge of source identification amidst species mixing•Distinct diurnal patterns reveal the operational dynamics of sources•Sensitivity analysis reveals heterogeneous source profiles•Recommendation: adopt rolling-PMF for real-time air quality management