Insights into Aerosol Composition, Source Signatures, Chemical Aging, and Transport Dynamics in Delhi–NCR from High-Resolution In-Situ Observations

Keywords: Source Apportionment, SOA, Biomass Burning, PMF, Aerosol Composition, Haze

This study investigates seasonal evolution of aerosol physicochemical properties and source influences in the upwind region of Delhi-NCR using a suite of state-of-the-art instruments deployed at Sonipat, Haryana, India (28.9° N, 77.1° E). Continuous measurements of non-refractory PM₂.₅ using a Time-of-Flight Aerosol Chemical Speciation Monitor (ToF-ACSM) and black carbon (BC) using an Aethalometer (AE-31) were conducted during the peak stubble-burning and Diwali period (25 Oct 2023 to 15 Nov 2023), a time characterized by strong episodic pollution events and regional transport influence. The observational period captured three contrasting regimes: (i) an initial non-haze phase (mean PM₂.₅: 218±90 ug/m³), (ii) an intense haze episode linked to crop-residue burning and meteorological stagnation (haze1: 507±217 ug/m³), followed by a rain-driven dilution event (non-haze2: 166 ±70 ug/m³), and (iii) a a subsequent Diwali-driven haze event (haze2: 311±140 ug/m³). Across all conditions, non-refractory PM₂.₅ was dominated by organic aerosols (OA: 66.9%), with secondary inorganic species such as nitrate (NO³⁻: 8.4%), sulfate (SO₄²⁻:4.7%), ammonium (NH₄⁺:6.6%), and chloride (Cl⁻:2.3%)_, contributing modest fractions. Positive Matrix Factorization (PMF) and Multilinear Engine (ME-2) analysis resolved five distinct OA sources: traffic-related hydrocarbon-like OA (HOA), biomass-burning OA (BBOA), solid fuel combustion OA (SFC-OA), two oxygenated OA components, less oxidized OA (LOOA) and more oxidized OA (MOOA) comprising 37.8% of total OA, indicative of extensive aging during transport. Among primary sources, SFC-OA (23%) and BBOA (11.2%) were most enhanced during pollution episodes, consistent with emissions from wood burning and post-harvest crop-residue fires. Aethalometer-derived BC source apportionment showed a relative decline in fossil–fuel BC during both haze phases, highlighting the strong episodic influence of biomass-burning plumes. Meteorological analysis indicates that the extreme haze1 event was amplified by a pronounced reduction in boundary-layer height and aerosol–radiation feedback, which suppressed vertical mixing and reinforced pollutant accumulation. Aerosols during haze1 exhibited high oxidation states and enhanced aging, pointing to prolonged atmospheric processing and regional transport from source regions upwind of Delhi–NCR. These findings provide a process-level understanding of aerosol evolution during high-pollution periods, illustrating the combined roles of emission variability, atmospheric aging, and meteorological feedback in shaping air quality over the Indo-Gangetic Plain.