Assimilation of Satellite Derived Aerosol Optical Depth

1
Assimilation of Satellite Derived Aerosol Optical
Depth

• Udaysankar Nair1, Sundar A. Christopher1,2
• 1 Earth System Science Center, University of
  Alabama in Huntsville
• 2 Department of Atmospheric Science, University
  of Alabama in Huntsville

2
Outline

• Prior research
• Numerical modeling of Saharan Dust storm, Use of
  GOES derived AOD
• Long range transport of smoke, Central American
  biomass
• Long range transport of smoke, Georgia fires
• Future plans, Inverse modeling of Ammonia and
  fire emissions

3
Numerical simulation of Saharan Dust Storm

• Used Regional Atmospheric Modeling System (RAMS)
  to simulate passage of Saharan Dust Storm over
  Puerto Rico during the PRIDE field experiment
  (Wang et al. 2004)
• Explore radiative impacts of dust aerosols
• Utilized hourly observation of GOES retrieved AOT
  to initialize and nudge the lateral boundaries

4
Numerical simulation of Saharan Dust Storm

• Utilized vertical distribution from aircraft
  measurements

5
Numerical modeling of smoke transport from
Central America

• In the case of episodic events, such as forest
  fires and biomass burning in Central America,
  simple smoke transport modeling is adequate for
  predicting air quality category
• RAMS incorporating satellite derived smoke
  emissions, used to simulate long range transport
  of smoke from Central America
• Used thirty minute, Fire Locating and Modeling of
  Burning Emissions (FLAMBE)

6
Numerical modeling of smoke transport from
Central America

• Fire emissions for the period 20 April to 21 May
  was utilized.

7
Numerical modeling of smoke transport from
Central America

• Fire emissions for the period 20 April to 21 May
  was utilized
• Specification of injection height required

8
Numerical modeling of smoke transport from
Central America
9
Numerical modeling of smoke transport from
Central America
10
Numerical modeling of smoke transport from
Central America
11
Numerical modeling of Georgia fires

• Smoke emissions for 23-25th of May 2007

12
Numerical modeling of Georgia fires

• Smoke emissions underestimated by 70

13
Numerical modeling of Georgia fires

• Surface concentrations underestimated during day
  time, vertical mixing? Injection height?

14
Future work, Inverse modeling approach using
Ensemble Kalman filtering

• Apply to Georgia fire simulations
• Assimilate 30 minute GOES AOD/ MODIS AOD, use
  multipliers to surface emissions, update in a
  manner similar to state variables
• Ammonia emissions from animal agriculture

15
Future work, Inverse modeling approach using
Ensemble Kalman filtering

• Use AOD as a constraint
• Assimilate 30 minute GOES AOD/ MODIS AOD

East N.C. East N.C. Urban Urban Europe Arctic Arctic Atlantic
Case I Case II UK US Average Marine Continental
SO42- 14.17 1.93 13.80 16.47 3.150 1.95 2.320 2.577
NO3- 1.81 23.95 3.00 9.7 0.920 0.022 0.055 0.050
Cl- 0.58 0.65 3.18 0.73 0.112 0.174 0.013 4.625
NH4 4.52 7.87 4.84 6.93 1.295 0.152 0.226 0.162