WRF/Chem Regional Modeling of the Mid-Atlantic: Comparison with Aura and Ground Based Measurements

1
WRF/Chem Regional Modeling of the Mid-Atlantic
Comparison with Aura and Ground Based Measurements

• Elena Yegorova, Dale Allen, Christopher Loughner,
  and Russell Dickerson of the University of
  Maryland
• Kenneth Pickering of NASA-GSFC

2
Goals of todays talk

• Simulate a Mid-Atlantic (U.S.) pollution event
  using a regional air quality model.
• Compare model-calc sfc layer 8-hr max ozone to
  AIRNOW observations.
• Compare model-calc and satellite-retrieved
  tropospheric ozone columns.
• Obtain insight into the capabilities of OMI and
  TES to detect air pollution on the urban scale.
• What event?
• July 9-11, 2007 pollution event
• What model?
• On-line Weather Research and Forecasting model
  with integrated chemistry (WRF-Chem).

3
Air quality on July 9, 2007 was very bad in the
Mid-Atlantic. 8-hr max O3 of 131 ppbv within
Baltimore NAA
Baltimore, MD visibility during a clear day
Baltimore, MD visibility At 645 PM July 9, 2007
Images obtained from the U.S. Air Quality Smog
Blog (http//alg.umbc.edu/usaq).
4
WRF/Chem Configuration
Atmospheric Processes WRF/Chem
Radiation LW RRTM SW Goddard
Surface Layer Monin-Obukhov
Land Surface Model Noah
Boundary Layer YSU
Cumulus Grell
Microphysics Lin et al.
Gas-phase chemistry RADM2
Aerosol chemistry MADE/SORGAM
Photolysis Fast J

• Domain Eastern U.S.
• 12 km hor res, 23 vertical layers
• Meteorology driven by the North American
  Mesoscale (NAM) model
• Emissions
• Points sources from July 2007 CEMS data.
• Area sources from 2002 U.S. inventory used in SIP
  modeling
• Emissions processed using SMOKE modeling system
• Default (clean) chemical BC and IC are assumed
  (00 UT 6 Jul 2007)
• No lightning NOx or stratospheric chemistry

5
TES and OMI Products

• TES
• Level 2 Nadir Step Stare Special Observations
  on July 9, 2007
• OMI-MLS global, daily tropospheric column
  product
• OMI total column minus stratospheric column from
  the Microwave Limb Sounder (MLS) (Lapse rate
  tropopause product of M.Schoeberl).

6
Build-up to July 9, 2007 Air Pollution Episode
Spatial pattern reasonable outside of Ohio River
Valley. High-bias present
7
WRF/Chem, OMI, and TES Tropospheric Column Ozone
Spatial pattern reasonable TES gt WRF-Chem gt
OMI/MLS Note anomalous peak in TES column over
northeastern New York state
8
SE part of TES track (outflow region) NW part
of TES track (NE/Canada)
9
A priori trpp lower?
SE part of TES track (outflow region)
NW part of TES track (New England/Canada)
10
Preliminary Conclusions

• Spatial pattern of July 9-11 air quality event is
  captured (outside of Ohio River Valley) by WRF.
  Model has high-bias at surface.
• For this particular TES flight track,
• TES tropospheric O3 column gt WRF column gt OMI/MLS
  column
• Low-bias of OMI/MLS column over New
  England/Canada may be due to overestimation of
  stratospheric MLS column (difference between a
  priori and a posteriori upper trop TES col
  suggests higher than normal tropp on this date)
• WRF-calc and satellite-observed tropospheric
  columns are greater in pollutant outflow region
  (over ocean) than over NE/Canada. Contrast of
  retrieved TES columns gt Contrast of a priori TES
  columns.
• Mean TES profiles show little vertical structure.
  However, anomalous (mostly boundary layer) peak
  is seen in TES column near Lake Champlain. Model
  shows small peak in this region.

Acknowledgements Gregory Osterman, Mark
Schoeberl, Dylan Jones, George Grell, and Steven
Peckham
11
This study is a work in progress. Possible next
steps include

• Compare with July 9 ozonesonde from Beltsville,
  MD
• Nested 36/12 km simulation of event with
  climatological BC/IC from GMI model
• Investigate sensitivity of 8-hr max ozone to
  convective parameterization and chemical
  mechanism
• Compare WRF-calc and satellite-observed HCHO and
  NO2 columns
• Apply averaging kernel to model before comparing
  to OMI measurements.
• Compare WRF-calc PM2.5 to measurements
• Other?