Identifying the sources of PM2.5 at a COALESCE site in the Brahmaputra Valley Region using dispersion normalized PMF

Abstract

This study presents the results from PM2.5 (particulate matter ≤2.5 μm) source apportionment of a full-year sampling campaign at a Carbonaceous Aerosol Emissions, Source Apportionment and Climate Impacts (COALESCE) network site in the Brahmaputra Valley Region (BVR) of India. Conventional positive matrix factorization (C-PMF) and dispersion normalized PMF (DN-PMF) was applied for the PM2.5 source apportionment using the receptor model USEPA PMF5.0. Six sources of PM2.5 pollution were identified at the receptor site: industrial, re-suspended crustal dust, combustion, secondary aerosols, biomass burning and chloride-salt. Annually, the major contributors to PM 2.5 were biomass burning (23% from C-PMF and 30% from DN-PMF), secondary aerosols (15% from C-PMF and 15% from DN-PMF) and combustion processes (20% from C-PMF and 18% from DN-PMF). Seasonal dominance of combustion aerosols (CA) and biomass burning (BB) was observed in winter and post-monsoon. The pre-monsoon season witnessed enhanced contribution of re-suspended crustal dust to PM2.5. Winter and post-monsoon enhancement in BB was observed in terms of percentage contribution to PM2.5 in DN-PMF owing to greater localized BB emissions. Remote transport and transboundary pollution were identified at the receptor site during the pre-monsoon season. The study can be utilized to implement source-specific control strategies for PM2.5 pollution in the BVR.