100
C. Arter and S. Arunachalam
11% over the past ten years [1]. This growth is expected to continue with the Federal
Aviation Administration forecasting a 2.1% increase in U.S. carrier passenger growth
each year for the next 20 years, and a 2.1% and 3.5% growth in system traffic in
revenue passenger miles over the next 20 years for domestic travel and international
travel, respectively [2].
This projected growth can place a burden on atmospheric air quality as aircraft
attributable emissions become a larger component of all anthropogenic emissions.
Aircraft emissions are composed primarily of nitrogen oxides (NO X ), sulfur oxides
(SO X ), volatile organic compounds (VOC), primary elemental carbon (PEC), primary organic carbon (POC), and primary sulfate (PSO 4 ). Both LTO and cruise-mode
emissions can lead to the formation of air pollutants such as O 3 and PM 2.5 affecting
populations near airports. One study on 99 U.S. airports estimates an increase in
premature deaths due to aviation emissions from 75 deaths in 2005 to 460 deaths in
2025 [3].
17.1.1 Methods
In this study we utilize an Eulerian atmospheric chemical transport model (CTM),
CMAQv5.0.2 [4], to quantify the concentration and transport of PM, O 3 , and other
pollutants in a 36 × 36 km grid cell resolution domain.
Various sensitivity analyses in the atmospheric CTM framework are used for
guiding policy and environmental scenarios. We make use of a sensitivity analysis
approach, (DDM-3D) [5–7], that excels in describing sensitivities to multiple input
parameters as well as sensitivities to small variations. Both of these conditions are
important for modeling aviation emissions since we aim to describe impacts due to
varying aircraft emission species and the aviation emission sector is much smaller
than other anthropogenic emission sectors. Within DDM-3D, derivatives are taken
at each time step, which calculate the change in concentration of a chemical species
with respect to some change in an input parameter. Output from DDM-3D is in the
form of sensitivity coefficients, which express the derivatives taken at each time step.
First order sensitivity coefficients describe linear changes in concentrations with
respect to changing emissions. In our case of air pollutants attributable to aviation
emissions, we need to be concerned with air pollutant species that may not be linearly
dependent on aviation emissions. The chemistry surrounding tropospheric PM 2.5
and O 3 formation is far more complicated than what can be expressed with only
first order changes. Hence, we extend our DDM-3D analysis framework to higher
order decoupled direct method in three dimensions (HDDM-3D) in order to calculate
second order sensitivity coefficients. In doing so, we hope to capture more of the
non-linearities in the chemistry related to PM 2.5 and O 3 formation from aviation
emissions.
For our study, six precursor emissions that are responsible for the formation of
PM 2.5 and two precursor emissions that are responsible for the formation of O 3
are chosen as sensitivity parameters for our HDDM-3D analyses. Three gas phase
C. Arter and S. Arunachalam
11% over the past ten years [1]. This growth is expected to continue with the Federal
Aviation Administration forecasting a 2.1% increase in U.S. carrier passenger growth
each year for the next 20 years, and a 2.1% and 3.5% growth in system traffic in
revenue passenger miles over the next 20 years for domestic travel and international
travel, respectively [2].
This projected growth can place a burden on atmospheric air quality as aircraft
attributable emissions become a larger component of all anthropogenic emissions.
Aircraft emissions are composed primarily of nitrogen oxides (NO X ), sulfur oxides
(SO X ), volatile organic compounds (VOC), primary elemental carbon (PEC), primary organic carbon (POC), and primary sulfate (PSO 4 ). Both LTO and cruise-mode
emissions can lead to the formation of air pollutants such as O 3 and PM 2.5 affecting
populations near airports. One study on 99 U.S. airports estimates an increase in
premature deaths due to aviation emissions from 75 deaths in 2005 to 460 deaths in
2025 [3].
17.1.1 Methods
In this study we utilize an Eulerian atmospheric chemical transport model (CTM),
CMAQv5.0.2 [4], to quantify the concentration and transport of PM, O 3 , and other
pollutants in a 36 × 36 km grid cell resolution domain.
Various sensitivity analyses in the atmospheric CTM framework are used for
guiding policy and environmental scenarios. We make use of a sensitivity analysis
approach, (DDM-3D) [5–7], that excels in describing sensitivities to multiple input
parameters as well as sensitivities to small variations. Both of these conditions are
important for modeling aviation emissions since we aim to describe impacts due to
varying aircraft emission species and the aviation emission sector is much smaller
than other anthropogenic emission sectors. Within DDM-3D, derivatives are taken
at each time step, which calculate the change in concentration of a chemical species
with respect to some change in an input parameter. Output from DDM-3D is in the
form of sensitivity coefficients, which express the derivatives taken at each time step.
First order sensitivity coefficients describe linear changes in concentrations with
respect to changing emissions. In our case of air pollutants attributable to aviation
emissions, we need to be concerned with air pollutant species that may not be linearly
dependent on aviation emissions. The chemistry surrounding tropospheric PM 2.5
and O 3 formation is far more complicated than what can be expressed with only
first order changes. Hence, we extend our DDM-3D analysis framework to higher
order decoupled direct method in three dimensions (HDDM-3D) in order to calculate
second order sensitivity coefficients. In doing so, we hope to capture more of the
non-linearities in the chemistry related to PM 2.5 and O 3 formation from aviation
emissions.
For our study, six precursor emissions that are responsible for the formation of
PM 2.5 and two precursor emissions that are responsible for the formation of O 3
are chosen as sensitivity parameters for our HDDM-3D analyses. Three gas phase