226
P. Makar et al.
forecasting system has been designed in preparation for a follow-up field campaign,
taking place during April and June of 2018.
35.1 GEM-MACH
Version 2 of the Global Environmental Multiscale—Modelling Air-quality and
CHemistry (GEM-MACH) model [5, 6, 8, 10] was used for the simulations described
here. The model uses a 2 or 12-bin sectional representation of aerosol microphysics,
with eight particle species (sulphate, nitrate, ammonium, secondary organic aerosol,
primary organic aerosol, elemental carbon, sea-salt and crustal material). The 2-bin
configuration of the model has been used for ongoing daily experimental forecasts
of pollutants and their deposition since 2012, while the 12-bin configuration has
been used for research simulations. The model configuration employs three levels
of nesting:—analysis-driven meteorological output from the Canadian operational
Regional Deterministic Prediction System is used to simultaneously drive both a
GEM-MACH 10 km resolution grid cell size North American domain simulation
and a Canadian High Resolution Deterministic Prediction System 2.5 km grid cell
size simulation over Canada and the USA. The former and latter provide chemical
and meteorological boundary and initial conditions respectively to a 2.5 km grid
cell size simulation domain encompassing the Canadian provinces of Alberta and
Saskatchewan (Fig. 35.1).
The emissions used in the simulations are the result of a multi-year effort to
improve emissions inventories in the region and include efforts to incorporate emissions data estimated from 2013 research flights around individual oil sands facilities
[13].
The use of these emissions data as well as the 2 and 12-bin configurations of the
model resulted in several important findings:
(1) Simulations employing model-measurement fusion predicted future acidifying
aquatic ecosystem damage at 2013 emissions levels, for an area greater than
3.8 × 10
5 km
2 [9].
(2) A parameterization of bi-directional fluxes of ammonia was required in order
to simulate observed ammonia concentrations [12].
(3) Improved estimates of emitted mass and speciation of volatile organic compounds resulted in significant performance improvements for predicted VOC
chemistry and secondary organic aerosol formation [11].
(4) The standard algorithms for estimating the rise of industrial plumes underestimated plume height, when driven using meteorological observations [4].
However, a revised algorithm showed much greater skill in estimating this key
parameter [1].
(5) The 12-bin size distribution was shown to improve the model’s aerosol performance for all statistical metrics [1].
P. Makar et al.
forecasting system has been designed in preparation for a follow-up field campaign,
taking place during April and June of 2018.
35.1 GEM-MACH
Version 2 of the Global Environmental Multiscale—Modelling Air-quality and
CHemistry (GEM-MACH) model [5, 6, 8, 10] was used for the simulations described
here. The model uses a 2 or 12-bin sectional representation of aerosol microphysics,
with eight particle species (sulphate, nitrate, ammonium, secondary organic aerosol,
primary organic aerosol, elemental carbon, sea-salt and crustal material). The 2-bin
configuration of the model has been used for ongoing daily experimental forecasts
of pollutants and their deposition since 2012, while the 12-bin configuration has
been used for research simulations. The model configuration employs three levels
of nesting:—analysis-driven meteorological output from the Canadian operational
Regional Deterministic Prediction System is used to simultaneously drive both a
GEM-MACH 10 km resolution grid cell size North American domain simulation
and a Canadian High Resolution Deterministic Prediction System 2.5 km grid cell
size simulation over Canada and the USA. The former and latter provide chemical
and meteorological boundary and initial conditions respectively to a 2.5 km grid
cell size simulation domain encompassing the Canadian provinces of Alberta and
Saskatchewan (Fig. 35.1).
The emissions used in the simulations are the result of a multi-year effort to
improve emissions inventories in the region and include efforts to incorporate emissions data estimated from 2013 research flights around individual oil sands facilities
[13].
The use of these emissions data as well as the 2 and 12-bin configurations of the
model resulted in several important findings:
(1) Simulations employing model-measurement fusion predicted future acidifying
aquatic ecosystem damage at 2013 emissions levels, for an area greater than
3.8 × 10
5 km
2 [9].
(2) A parameterization of bi-directional fluxes of ammonia was required in order
to simulate observed ammonia concentrations [12].
(3) Improved estimates of emitted mass and speciation of volatile organic compounds resulted in significant performance improvements for predicted VOC
chemistry and secondary organic aerosol formation [11].
(4) The standard algorithms for estimating the rise of industrial plumes underestimated plume height, when driven using meteorological observations [4].
However, a revised algorithm showed much greater skill in estimating this key
parameter [1].
(5) The 12-bin size distribution was shown to improve the model’s aerosol performance for all statistical metrics [1].