1' FY10 GOESR3 Project Proposal Title Page

1
1. FY10 GOES-R3 Project Proposal Title Page

• Title Air Quality Applications of Satellite Data
• Project Type GOES-R data utilization project
• Status Renewal
• Duration 2 years
• Leads
• Shobha Kondragunta (NESDIS/STAR)
• Qiang Zhao (IMSG)
• Other Participants
• Istvan Laszlo (NESDIS/STAR)
• R. Brad Pierce (NESDIS/STAR)
• Jeff McQueen (NWS/NCEP)
• Pius Lee (NWS/NCEP-SAIC)
• Rohit Mathur (NOAA/ARL)
• James J. Szykman (EPA)

1
2
2. Project Summary

• Use current operational GOES Aerosol Optical
  Depth (AOD) products to improve air quality
  forecasts. Research and development work done
  under this project will investigate the
  usefulness of aerosol assimilation in improving
  air quality forecasts and pave the way for using
  enhanced aerosol products from GOES-R ABI
• Operational GOES AOD data
• GOES-R ABI like retrievals obtained from MODIS
  radiances
• EPA-NWS numerical air quality forecast system
  (research version) Weather Research and
  Forecasting/Community Multiscale Air Quality
  (WRF/CMAQ) system
• Tasks
• Develop AOD assimilation methodologies
• Conduct WRF/CMAQ simulations for multiple case
  studies
• Demonstrate impact
• Expected Outcome
• Demonstration of improved air quality predictions

2
3
3. Motivation/Justification

• Supports NOAA Mission Goal(s)
• Weather and water
• GOES-R ABI aerosol products will support
  Memorandum of Understanding (MOU) and Memorandum
  of Agreement (MOA) between EPA and NOAA
• Current GOES aerosol products have limitations.
  Only AOD retrieval from a single channel is
  possible. Retrieval has uncertainties associated
  with surface reflectance retrieval and other
  assumptions. GOES-R ABI aerosol products are
  expected to be of better quality than current
  GOES.
• Emissions inventory in the forecast model for
  ozone and aerosol precursors is not updated
  during the simulations. This leads to forecast
  errors in the model when episodic events such as
  forest fires and dust outbreaks introduce high
  amount of aerosols into the atmosphere.
  Satellite derived column AOD measurements
  correlate with surface PM2.5 (particles smaller
  than 2.5 microns in diameter) and can be used to
  tune model initial conditions. Satellite data
  available in near real time can circumvent that
  problem by providing improved initial conditions
  during the analysis part of the forecast cycle.

Correlation between GOES-12 AOD and Surface PM2.5
for a mid-western site
GOES AOD
PM2.5 (µg/m3)
3
4
4. Methodology

• Assimilation experiments
• 1st phase
• Choose air pollution episodes corresponding to
  urban/industrial, dust outbreak, and biomass
  burning events
• Conduct base case model simulations
• Use GOES AODs to derive modeled/observed ratios
  and tune PM2.5 concentrations in the model
• Expected to work well for situations when
  aerosols are uniformly distributed in the
  boundary layer compared to situations when
  aerosols are elevated
• Use NESDIS Automated Smoke Detection Algorithm
  (ASDA) to identify smoke aerosols and modify
  assimilation approach accordingly
• 2nd phase
• Determine observational errors and model
  background errors
• Conduct a full 3-D variational assimilation study
• Analyze the assimilation runs by comparing model
  predicted surface PM2.5 to observed PM2.5 across
  the whole CONUS
• Repeat 1st and 2nd phase with GOES-R ABI like
  retrievals obtained using MODIS as a proxy dataset

4
5
6. Expected Outcomes

• This is a feasibility study to show the
  usefulness of operational aerosol products from
  GOES platform in improving air quality forecasts
• Risks involved inability of satellite retrievals
  to specifically identify aerosol speciation, lack
  of vertical resolution in retrieved aerosol
  products
• This study will assess the merits and limitations
  of satellite data in improving air quality
  forecasts
• If successful, research and development work has
  the potential to be transferred to NWS for
  operational implementation
• NWS/NCEP and NESDIS/STAR meet once a month to
  discuss chemical data assimilation issues.
  Timely updates of this project are provided to
  NWS

5
6
7. Major Milestones

• FY09
• GOES-R3
• Complete analysis of model and background errors
  and observational errors
• Model background errors will be determined by
  using forecasts and analysis
• Observational background errors determined by
  comparing satellite AOD data with ground truth
• Investigate aerosol vertical profiles in CMAQ
  model

6
7
7. FY09 Accomplishments
No Assimilation
Assimilation
Improved surface PM2.5 predictions due to GOES
AOD assimilation
8
7. FY09 Accomplishments
Day Time Minimum in Surface PM2.5

• Comparisons of model predicted AOD agrees with in
  situ observations better for assimilation run
  than base run
• PBL dynamics appear to modulate surface PM2.5
  concentrations

9
Comparison of CMAQ and CALIPSO Vertical Profiles
of Aerosols (Sulfate Episode)
7. FY09 Accomplishments
CALIPSO
Impact of GOES AOD assimilation on improving
PM2.5 predictions depend on CMAQ vertical profiles
BASE
DA-GOES
10
7. FY09 Accomplishments

• 3DVar
• Combining background forecasts and observations
  in a statistically sound way
• Computationally practical for operational use
• Relatively easy to implement
• Potential for including different types of
  observations

10
11
7. Major Milestones

• FY10
• GOES-R3
• Obtain and test a newer version of CMAQ from NWS
  that has biomass burning sources FY09 activity
  ongoing
• Complete base case and test case runs for a
  biomass burning event - FY09 activity ongoing
• Complete analysis of the results
• Complete comparison of model simulations and
  observations to demonstrate improvements in model
  predictions - FY09 activity ongoing
• Complete code development for 3-D var
  assimilation - FY09 activity ongoing
• Complete compilation of aerosol profile
  climatology for potential testing in assimilation
  experiments - FY09 activity ongoing
• Conduct assimilation runs using 3-D var
  assimilation scheme
• Analyze model assimilation runs with assimilation
  runs using Cressman analysis
• Evaluate assimilation runs using ground
  observations
• FY11
• GOES-R3
• Continue the development and testing of aerosol
  assimilation runs
• Assimilate ABI retrievals
• Generate CMAQ based proxy datasets for ABI
  aerosol retrieval studies
• Complete assimilation experiments using GOES-R
  ABI retrievals using proxy MODIS data FY10
  activity moved to FY11

11
12
8. Funding Profile (K)

• Summary of leveraged funding
• STAR base funding for Shobha Kondragunta
• Computer infrastructure from NWS

12
13
9. Expected Purchase Items

• FY09 120,000 Total Project Budget
• (120K) STAR contractor at full time from Apr 09
  to Mar 10
• 120K for IMSG contract
• FY10 125,000 Total Project Budget
• (125K) STAR contractor at full time from Apr 10
  to Mar 11
• 125K for IMSG contract
• FY11 135,000 Total Project Budget
• (135K) STAR contractor at full time from Apr 11
  to Mar 12
• 135K for IMSG contract

13