Environmental Engineering Reference
In-Depth Information
may be especially influential in multi-pollutant planning due to the tradeoffs and
co-benefits between controls for each pollutant (Chestnut et al., 2006). However,
such efforts are complicated by lack of clarity as to which temporal metric is best
suited for assessing the health benefits of controls. Epidemiological studies in the
scientific literature have applied a wide variety of temporal averaging periods (e.g.,
1-h, 8-h, daily, or annual average) and threshold values in developing concentration-
response functions for air pollutant health effects. For ozone, the choice of a
temporal metric may influence health effect magnitude and trend estimates (Stedman
and Kent, 2008) and shift which emissions scenario is perceived to most impact
human health (Bell et al., 2005).
Here, we investigate how alternate temporal metrics and thresholds for ozone
concentration-response functions could influence the optimization of control
strategies for health benefits. The work presented here is a preliminary component
of a broader study that is exploring how uncertainties in control costs, photochemical
sensitivities, and concentration-response functions can be jointly considered to
inform the development of air quality attainment strategies.
2. Methods
The modeling approach was designed to simulate the health implications that could
be considered in the development of a state-level ozone attainment plan. Photo-
chemical sensitivity modeling was conducted with the Community Multi-scale Air
Quality (CMAQ) model version 4.5 (Byun and Schere, 2006) with the CB-IV
chemical mechanism. Meteorology and emissions inputs were developed and
evaluated by the Visibility and Improvements State and Tribal Association of the
Southeast (VISTAS) (Morris et al., 2007). VISTAS applied MM5 to simulate
meteorological conditions for 2002; here we focus on the period May 30-June 6,
2002 (first 2 days discarded for model initialization). Emissions are from VISTAS'
Year 2009 projections (projected from a 2002 base inventory), with updates to
Georgia emissions projections based on Georgia Environmental Protection Division's
(GA EPD) SIP modeling.
Sensitivity analysis was conducted with the decoupled direct method (DDM) in
CMAQ (Dunker, 1984). First-order DDM sensitivity coefficients
S
i,j
(1)
represent
the responsiveness of pollutant
i
to an incremental change in input
j
. A sensitivity
coefficient of
S
i,j
(1)
=
y ppb indicates that an x% reduction in
j
would yield an
(xy/100) ppb reduction in
i
.
Sensitivities were simulated to emissions from three regions: “Atlanta,” defined
as the 20-county ozone non-attainment region; “Macon,” defined as a 7-county
region centered on the city of Macon; and Plant Scherer, the largest NO
x
emitting
power plant in Georgia in the projected emissions inventory.
Sensitivity results were processed based on one of the following temporal
metrics, each of which has been the basis of some ozone health studies:
