Environmental Engineering Reference
In-Depth Information
8.8 Evaluation of Toxic Air Contaminants in the
San Francisco Bay Area: Regional Modeling
Philip T. Martien
1
, Saffet Tanrikulu
1
, Yiqin Jia
1
, David Fairley
1
, Cuong
Tran
1
, Jeff Matsuoka
1
, Henry Hilken
1
, Chris Emery
2
, Ed Tai
2
,
and Greg Yarwood
2
1
Bay Area Air Quality Management District, San Francisco, CA, USA
2
ENVIRON International Corporation, Novato, CA, USA
1. Introduction
In 2004, Bay Area Air Quality Management District (BAAQMD) initiated the
Community Air Risk Evaluation (CARE) program to estimate and reduce health
risks associated with exposure to outdoor toxic air contaminants (TAC) within the
San Francisco Bay Area (Bay Area, see
Fig. 1).
Information gathered through a
variety of technical studies has guided the development of measures to reduce
TAC in areas with high health risk. This extended abstract summarizes regional
toxics modeling and model evaluation conducted to support the CARE program.
To produce inputs for regional toxic modeling, we generated meteorological
fields for winter and summer periods in 2000 using the MM5 mesoscale meteoro-
logical model (Grell et al., 1994) and an emissions inventory for year 2005 using
methods described in a companion report (Martien et al., 2009). Initial and boundary
conditions were set using available observations as a guide. The horizontal grid
resolution of both meteorological and air quality models was 4 by 4 km.
Modeling inputs were supplied to the CAMx air quality model (ENVIRON,
2006) to predict concentrations of toxic compounds. Some toxic contaminants,
such as diesel PM, were treated as inert compounds. Other contaminants, such as
formaldehyde and acrolein, underwent chemical reactions in the atmosphere. They
were both directly emitted and formed secondarily from other pollutants. Reactive
chemistry simulations used the SAPRC99 chemical mechanism to predict ozone
and radical concentrations, and applied the reactive tracer chemical mechanism
compiler (RTCMC; Emery et al., 2008) to predict trace toxic compounds. Predicted
concentrations were compared to measurements from a network of TAC monitoring
sites (Martien et al., 2009).
To estimate health risks, predicted pollutant concentrations were weighted by
unit risk factors for cancer-causing compounds
-
such as benzene
-
or normalized
by reference exposure levels for compounds that produce non-cancerous, chronic
and acute heath effects
-
such as acrolein. Total excess cancer risk was estimated
by from cancer-risk-weighted concentrations of five species that collectively represent