Abstract. Growing ship traffic in Atlantic Canada strengthens the local economy but
also plays an important role in greenhouse gas and air pollutant emissions in
this coastal environment. A mobile open-path Fourier transform infrared
(OP-FTIR; acronyms defined in Appendix A) spectrometer was set up in Halifax
Harbour (Nova Scotia, Canada), an intermediate harbour integrated into the
downtown core, to measure trace gas concentrations in the vicinity of marine
vessels, in some cases with direct or near-direct marine combustion plume
intercepts. This is the first application of the OP-FTIR measurement
technique to real-time, spectroscopic measurements of CO2, CO,
O3, NO2, NH3, CH3OH, HCHO,
CH4 and N2O in the vicinity of harbour emissions
originating from a variety of marine vessels, and the first measurement of
shipping emissions in the ambient environment along the eastern seaboard of
North America outside of the Gulf Coast. The spectrometer, its active mid-IR
source and its detector were located on shore while the passive retroreflector
was on a nearby island, yielding a 455 m open path over the ocean
(910 m two-way). Atmospheric absorption spectra were recorded during day, night,
sunny, cloudy and substantially foggy or precipitating conditions, with a
temporal resolution of 1 min or better. A weather station was co-located
with the retroreflector to aid in the processing of absorption spectra and
the interpretation of results, while a webcam recorded images of the harbour once
per minute. Trace gas concentrations were retrieved from spectra by the MALT
non-linear least squares iterative fitting routine. During field measurements
(7 days in July–August 2016; 12 days in January 2017) AIS information on
nearby ship activity was manually collected from a commercial
website and used to calculate emission rates of shipping combustion products
(CO2, CO, NOx, HC, SO2), which were then
linked to measured concentration variations using ship position and wind
information. During periods of low wind speed we observed extended (∼9 h)
emission accumulations combined with near-complete O3
titration, both in winter and in summer. Our results compare well with a NAPS
monitoring station ∼1 km away, pointing to the extended spatial scale of
this effect, commonly found in much larger European shipping channels. We
calculated total marine sector emissions in Halifax Harbour based on a
complete AIS dataset of ship activity during the cruise ship season (May–October 2015)
and the remainder of the year (November 2015–April 2016) and found
trace gas emissions (tonnes) to be 2.8 % higher on average during the
cruise ship season, when passenger ship emissions were found to contribute
18 % of emitted CO2, CO, NOx, SO2
and HC (0.5 % in the off season due to occasional cruise ships arriving, even in
April). Similarly, calculated particulate emissions are 4.1 % higher
during the cruise ship season, when passenger ship emissions contribute
18 % of the emitted particulate matter (PM) (0.5 % in the off season). Tugs were found to make the
biggest contribution to harbour emissions of trace gases in both cruise ship
season (23 % NOx, 24 % SO2) and the off season
(26 % of both SO2 and NOx), followed by container ships
(25 % NOx and SO2 in the off season, 21 %
NOx and SO2 in cruise ship season). In the cruise ship season
cruise ships were observed to be in third place regarding trace gas emissions, whilst tankers were in third place in the off
season, with both being responsible for 18 % of the calculated emissions. While the
concentrations of all regulated trace gases measured by OP-FTIR as well as
the nearby in situ NAPS sensors were well below maximum hourly permissible
levels at all times during the 19-day measurement period, we find that AIS-based
shipping emissions of NOx over the course of 1 year are
4.2 times greater than those of a nearby 500 MW stationary source emitter and
greater than or comparable to all vehicle NOx emissions in
the city. Our findings highlight the need to accurately represent emissions
from the shipping and marine sectors at intermediate ports integrated into
urban environments. Emissions can be represented as pseudo-stationary and/or
pseudo-line sources.