The Nepal Ambient Monitoring and
Source Testing Experiment (NAMaSTE) campaign took place in and around the
Kathmandu Valley and in the Indo-Gangetic Plain (IGP) of southern
Nepal during April 2015. The source characterization phase targeted numerous
important but undersampled (and often inefficient) combustion sources that
are widespread in the developing world such as cooking with a variety of
stoves and solid fuels, brick kilns, open burning of municipal solid waste
(a.k.a. trash or garbage burning), crop residue burning, generators,
irrigation pumps, and motorcycles. NAMaSTE produced the first, or rare, measurements of aerosol optical properties, aerosol
mass, and detailed trace gas chemistry for the emissions from many of the
sources. This paper reports the trace gas and aerosol measurements obtained
by Fourier transform infrared (FTIR) spectroscopy, whole-air sampling (WAS),
and photoacoustic extinctiometers (PAX; 405 and 870 nm) based on field work
with a moveable lab sampling authentic sources. The primary aerosol optical
properties reported include emission factors (EFs) for scattering and
absorption coefficients (EF Bscat, EF Babs,
in m2 kg−1 fuel burned), single scattering albedos (SSAs), and absorption
Ångström exponents (AAEs). From these data we estimate black and
brown carbon (BC, BrC) emission factors (g kg−1 fuel burned). The trace
gas measurements provide EFs (g kg−1) for CO2, CO, CH4,
selected non-methane hydrocarbons up to C10, a large suite of oxygenated
organic compounds, NH3, HCN, NOx, SO2, HCl, HF, etc. (up to ∼ 80 gases in all).
The emissions varied significantly by source, and light absorption by both
BrC and BC was important for many sources. The AAE for dung-fuel
cooking fires (4.63 ± 0.68) was significantly higher than for
wood-fuel cooking fires (3.01 ± 0.10). Dung-fuel cooking fires also
emitted high levels of NH3 (3.00 ± 1.33 g kg−1), organic acids
(7.66 ± 6.90 g kg−1), and HCN (2.01 ± 1.25 g kg−1), where the latter
could contribute to satellite observations of high levels of HCN in the
lower stratosphere above the Asian monsoon. HCN was also emitted in
significant quantities by several non-biomass burning sources. BTEX
compounds (benzene, toluene, ethylbenzene, xylenes) were major emissions
from both dung- (∼ 4.5 g kg−1) and wood-fuel (∼ 1.5 g kg−1) cooking fires, and a simple method to estimate indoor exposure to
the many measured important air toxics is described. Biogas emerged as the
cleanest cooking technology of approximately a dozen stove–fuel combinations
measured. Crop residue burning produced relatively high emissions of
oxygenated organic compounds (∼ 12 g kg−1) and SO2 (2.54 ± 1.09 g kg−1). Two brick kilns co-firing different amounts of biomass
with coal as the primary fuel produced contrasting results. A zigzag kiln
burning mostly coal at high efficiency produced larger amounts of BC, HF,
HCl, and NOx, with the halogenated emissions likely coming from the
clay. The clamp kiln (with relatively more biomass fuel) produced much
greater quantities of most individual organic gases, about twice as much
BrC, and significantly more known and likely organic aerosol precursors.
Both kilns were significant SO2 sources with their emission factors
averaging 12.8 ± 0.2 g kg−1. Mixed-garbage burning produced
significantly more BC (3.3 ± 3.88 g kg−1) and BTEX (∼ 4.5 g kg−1) emissions than in previous measurements. For all fossil fuel sources,
diesel burned more efficiently than gasoline but produced larger NOx
and aerosol emission factors. Among the least efficient sources sampled were
gasoline-fueled motorcycles during start-up and idling for which the CO EF
was on the order of ∼ 700 g kg−1 – or about 10 times that of a
typical biomass fire. Minor motorcycle servicing led to minimal if any
reduction in gaseous pollutants but reduced particulate emissions, as
detailed in a companion paper (Jayarathne et al., 2016). A small
gasoline-powered generator and an “insect repellent fire” were also among
the sources with the highest emission factors for pollutants. These
measurements begin to address the critical data gap for these important,
undersampled sources, but due to their diversity and abundance, more work is
needed.