August 10, 1990

 
MEMORANDUM

SUBJECT:  Phoenix CO Design Concentrations

FROM:     James L. Dicke, Chief
          Techniques Evaluation Section, SRAB, TSD (MD-14)

TO:       John Vimont, Regional Meteorologist
          Air Programs Branch, Region IX (A-2-1)


     This memorandum provides a response to your August 1, 1990
memorandum on determining design CO concentrations for the
Phoenix FIP.  It is also intended to reflect the subsequent
discussions with the Model Clearinghouse.

     We agreed that a combination of areawide and hot spot
modeling ultimately will be used for determining the 1989 design
concentration for the Phoenix FIP.  Since the Urban Airshed Model
(UAM) has already been applied successfully in Phoenix for an
October 1985 design day, it should be utilized for the 1989
areawide modeling.  The UAM modeling for the December 1989 design
day should utilize 1989 meteorological, traffic, and emissions
data.  The MOBILE4 model should be used to develop the emissions
inputs for the modeling.  The hot spot modeling could use the
modeling technique used for modeling the 1985 design day.  In
addition to determining the base case 1989 design concentrations,
the UAM and hot spot modeling should be used to model control
strategies to demonstrate attainment.

     As we discussed on the telephone, it is recognized that this
modeling cannot be completed in time for the Notice of Proposed
Rulemaking (NPRM), but would be completed in time for the FIP
promulgation or shortly thereafter.  Thus, the NPRM should
indicate UAM and hot spot modeling will be used to determine the
design concentration and evaluate control strategies for the
Phoenix FIP, but this analysis will not be completed until late
1990 or early 1991.  In the meantime, the use of a design 
concentration based on monitoring data is appropriate.

     The design concentration based on monitoring data should be
normalized to remove the effect of the oxygenated fuels program. 
Furthermore, the design concentration based on normalized 
monitoring data would need to be adjusted upward to account for
the increase in the design concentration that would result if
modeling was performed using a spatial grid of receptors.  This
upward, spatial adjustment can be estimated using the results of
the UAM and hot spot modeling performed for the October 1985
design day.

     Since the NPRM is needed in mid-August, rollback can be used
to identify control measures to demonstrate attainment for the
NPRM.  However, as soon as the UAM and hot spot modeling is
completed, those results should replace the use of rollback in
identifying control measures for attainment.

     If you have any questions or need further information,
please contact Tom Braverman of my staff at FTS 629-5383.

cc:  T. Helms
     W. Laxton
     J. Tikvart
     D. Wilson
     J. Vitas