Analysis of Ozone Modeling for May July 2006 in PNW using AIRPACT3 CMAQ and CAMx'

1
Analysis of Ozone Modeling for May July 2006 in
PNW using AIRPACT3 (CMAQ) and CAMx. Robert
Kotchenruther, Ph.D. EPA Region 10 Nov. 5 2006
Current AIRPACT3 12km domain
CMAQ O3 Prediction 7/22/2006 00Z
2
Background The basis of the Model Runs was
WSUs AIRPACT3 air quality forecast system.
Daily 24 hour prediction of air quality (gases
and aerosols). 95 x 95 grid at 12km
resolution. CMAQ v4.4 model with SPARC99
chemical mechanism. MM5 12km meteorological
forecasts from UW. Uses the SMOKE emissions
model with emissions files developed by WSU. For
this retrospective run, AIRPACT3 was used as
above except Archived MM5 forecasts were
concatenated into a continuous meteorological
dataset using hours 12-23 from each UW run.
The CAMx (v4.31) model was also run by converting
CMAQ IC, BC, and emissions files into CAMx input
format. CMAQ and CAMx were run for the period
May 8 July 31, 2006. First 4 days of run
were removed for model spin-up. Some CMAQ and
CAMx differences CAMx uses SAPRC99, but does
not include aerosol chemistry. Differences in
formulation of chemistry numerical solvers
Differences in treatment of vertical and
horizontal advection and diffusion (minor)
Differences in treatment of plume rise from
elevated point sources (minor) CMAQ and CAMx
user different MM5/met pre-processors.
3
Background(2) The motivation for this
retrospective run of AIRPACT3 was that Washington
Oregon experienced exceedances of the 8 hour
ozone standard at 5 sites this summer. Provided
and opportunity to explore how well AIRPACT3 is
performing for O3
Max 8hr average O3, WA OR sites, 2006 (ppbv)

• Main purpose of presentation
• A snapshot of current AIRPACT3 ozone
  performance.
• A lead in to a discussion of where we should
  place efforts to improve
• - our understanding of models current
  performance
• - where best to place efforts at improving model
  performance

4
This analysis will focus on the five sites with
8hr O3 exceedances
(WA Sites) ISSLKSAM ENUMUDMT PAC_FOR (OR
Sites) CARUS TURNER
WA OR O3 monitoring sites available for analysis
5
The plots on the following slides show daily max
(and min) ozone (and temperature). The maximum
1hr average was taken for both models and
monitor, unpaired in time. This example is from
Enumclaw, Mud Mountain Road.
Hourly Average O3, 2006
Daily Max O3, 2006
6
ENUMUDMT (WA) Enumclaw, Mud Mountain Road
exceedance day Plots show daily max and min
ozone, 1hr average, model and monitor unpaired in
time.


Daily Max O3, 2006


Daily Min O3, 2006
7
ENUMUDMT (WA) Enumclaw, Mud Mountain Road
exceedance day Plots show daily max ozone and
temp, 1hr average, model and monitor unpaired in
time.


Daily Max O3, 2006


Daily Max Temp., 2006
8
CARUS (OR) Carus, Spangler Rd. exceedance
day Plots show daily max ozone and temp, 1hr
average, model and monitor unpaired in time.

Daily Max O3, 2006
Daily Max Temp., 2006
9
TURNER (OR) Turner Cascade Jr. High
exceedance day Plots show daily max ozone and
temp, 1hr average, model and monitor unpaired in
time.

Daily Max O3, 2006
Daily Max Temp., 2006
10
PAC_FOR (WA) Pack Forest exceedance
day Plots show daily max and min ozone, 1hr
average, model and monitor unpaired in time.


Daily Max O3, 2006

ISSLKSAM (WA) Issaquah Lake Sammamish


Daily Max O3, 2006
11
Scatter plot comparing daily 1hr ozone maximum
observations and model predictions, unpaired in
time. Combined data from 5 sites reporting O3
exceedances in OR WA.
12
Scatter plot comparing daily 1hr ozone minimum
observations and model predictions, unpaired in
time. Combined data from 5 sites reporting O3
exceedances in OR WA.
13

• Discussion
• Ideas for where to look to understand model over
  prediction of ozone when ambient ozone is below
  80 ppb and under prediction when over 80 ppb?
• Biases in pbl or other met variables?
• Emissions inventories?
• Nesting to a finer resolution?
• ????

• Discuss above in two contexts
• Given current available measurements.
• If we begin making measurements not currently
  available.
• i.e., mobile lab to be discussed later today