『Abstract
Samples of airborne particulate matter (PM2.5)
were collected at a site in Lahore, Pakistan from November 2005
to January 2006. A total of 129 samples were collected using an
Andersen Reference Ambient Air Sampler 2.5-400 sampler and analyzed
for major ions, trace metals, and organic and elemental carbon
concentrations. The data set was then analyzed by positive matrix
factorization (PMF) to identify the possible sources of the atmospheric
PM collected in this urban area. Six factors reproduced the PM2.5 sample compositions with meaningful physical
interpretation of the resolved factors. The sources included secondary
PM, diesel emissions, biomass burning, coal combustion, two-stroke
vehicle exhaust, and industrial sources. Diesel and two-stroke
vehicles contributed about 36％, biomass burning about 15％, and
coal combustion sources around 13％ of the PM2.5
mass. Nearly two thirds of the PM2.5 mass
is carbonaceous material. Secondary particles contributed about
30％ of PM2.5 mass. The conditional probability
function (CPF) was then used to help identify likely locations
of the sources present in this area. CPF analysis point to the
east and northeast, which are directions of urban and industrial
areas located across the border near Amritsar, India as the most
probable source for high PM2.5 concentration
from diesel and two-stroke vehicles exhaust in Lahore. Analysis
of those days within three different ranges of PM2.5
concentration shows that most of the measured high PM2.5
mass concentrations were driven by diesel and two-stroke vehicle
emissions including the associated primary sulfate. The use of
the potential source contribution function (PSCF) to find the
source locations of regionally transported particles is inapplicable
in situations when high PM2.5 concentrations
are dominated by local sources and local meteorology.
Keywords: Receptor modeling; Two-stroke emissions; Lahore; PM2.5; PDCF; CPF』
1. Introduction
2. Experimental methods
2.1. Positive matrix factorization
2.2. Conditional probability function analysis
3. Results and discussions
3.1. Monthly variations in measured PM2.5
concentration
3.2. Source apportionment
3.3. Source contributions to ranges of PM2.5
concentrations
3.4. Analysis of monthly and diurnal variations in PMF resolved
sources
3.5. Analysis of PM2.5 source locations
identified by PSCF
4. Conclusions
Acknowledgements
References