19 Hamilton Airshed Modelling System
115
industrial, commercial, residential, mobile, and non-road (commercial marine shipping, locomotives and aircraft). Biogenic sources were prepared using the Model
of Emissions of Gases and Aerosols from Nature (MEGAN; [4]) biogenic emissions model. For the on-road mobile sources, the emissions rates were based on
input mobile source activity data, emission factors estimated using the Motor Vehicle Emission Simulator (MOVES2014 [5]), and outputs from transportation traveldemand models. Emissions data sets were derived from multiple agencies (e.g.,
United States Environmental Protection Agency (USEPA), Environment and Climate Change Canada and Ministry of Environment and Climate Change) as well as
other municipal and regional agencies and industries in Hamilton.
The modelling system was based on the Community Multi-scale Air Quality
v5.0.2 (CMAQ v5.0.2; [1] model. CMAQ simulates the various chemical and physical
processes influencing the local air quality, including the emission, dispersion and
chemical transformation of pollutants. The CMAQ modelling system consists of
four pre-processors to the CTM, which provide information on the meteorology,
meteorology-dependant chemical reactions and the concentration of species at the
boundary and in the background (i.e. non-local sources). The modelling system was
developed for four one-way nested modelling tiers with resolutions of 36, 12, 4 and
1.33 km. The highest resolution tier (Tier IV) was selected to allow for the focus
on the Hamilton region. HAMS was run for 2012 with meteorology and emissions
developed to be consistent with the air quality in 2012.
19.3 Modelling Performance Results
Model Performance results were measured against publicly available monitoring
data in Tier IV in both paired and unpaired statistical analyses. The modelled values
are taken from the grid cell that contains the monitoring site. Spatial displays across
the Tier IV domain and time series at monitoring sites were also considered.
The modelling system provides generally reasonable results given the complexity of the model and inputs. CMAQ tends to over-predict concentrations with the
exception PM 10 where the model under-predicts based on the mean bias. The underprediction of PM 10 is likely due to unaccounted local sources generating fugitive
dust. Model results are within a factor of two for all compounds (i.e., a fractional
bias ≤67%).
Time series analysis shows the modelled concentrations following the observed
concentrations, with a tendency to over-predict during the winter season. The overprediction of the NO 2 is evident throughout the year and is consistent at approximately 10 ppb, which suggest that transboundary emissions are influencing the
results.
The spatial displays show a large difference between the maximum daily value
(over the entire year) and the annual average of the daily values. This indicates that the
region experiences high magnitude events but the average conditions are much lower,
by at least an order of magnitude. Generally, maximum concentrations occurred along
115
industrial, commercial, residential, mobile, and non-road (commercial marine shipping, locomotives and aircraft). Biogenic sources were prepared using the Model
of Emissions of Gases and Aerosols from Nature (MEGAN; [4]) biogenic emissions model. For the on-road mobile sources, the emissions rates were based on
input mobile source activity data, emission factors estimated using the Motor Vehicle Emission Simulator (MOVES2014 [5]), and outputs from transportation traveldemand models. Emissions data sets were derived from multiple agencies (e.g.,
United States Environmental Protection Agency (USEPA), Environment and Climate Change Canada and Ministry of Environment and Climate Change) as well as
other municipal and regional agencies and industries in Hamilton.
The modelling system was based on the Community Multi-scale Air Quality
v5.0.2 (CMAQ v5.0.2; [1] model. CMAQ simulates the various chemical and physical
processes influencing the local air quality, including the emission, dispersion and
chemical transformation of pollutants. The CMAQ modelling system consists of
four pre-processors to the CTM, which provide information on the meteorology,
meteorology-dependant chemical reactions and the concentration of species at the
boundary and in the background (i.e. non-local sources). The modelling system was
developed for four one-way nested modelling tiers with resolutions of 36, 12, 4 and
1.33 km. The highest resolution tier (Tier IV) was selected to allow for the focus
on the Hamilton region. HAMS was run for 2012 with meteorology and emissions
developed to be consistent with the air quality in 2012.
19.3 Modelling Performance Results
Model Performance results were measured against publicly available monitoring
data in Tier IV in both paired and unpaired statistical analyses. The modelled values
are taken from the grid cell that contains the monitoring site. Spatial displays across
the Tier IV domain and time series at monitoring sites were also considered.
The modelling system provides generally reasonable results given the complexity of the model and inputs. CMAQ tends to over-predict concentrations with the
exception PM 10 where the model under-predicts based on the mean bias. The underprediction of PM 10 is likely due to unaccounted local sources generating fugitive
dust. Model results are within a factor of two for all compounds (i.e., a fractional
bias ≤67%).
Time series analysis shows the modelled concentrations following the observed
concentrations, with a tendency to over-predict during the winter season. The overprediction of the NO 2 is evident throughout the year and is consistent at approximately 10 ppb, which suggest that transboundary emissions are influencing the
results.
The spatial displays show a large difference between the maximum daily value
(over the entire year) and the annual average of the daily values. This indicates that the
region experiences high magnitude events but the average conditions are much lower,
by at least an order of magnitude. Generally, maximum concentrations occurred along