1
The base case simulation of the AI, including scattering organic carbon as currently used in GEOS-Chem, is compared to corresponding OMI observations. The simulation captures major absorption features compared to OMI over desert regions, however it fails to capture absorption by biomass burning aerosol. The correlation between the simulated and OMI AI is low in all major biomass burning regions with large mean biases. The AI is a method of detecting absorbing aerosols from satellite measurements, and is a product of the Ozone Monitoring Instrument (OMI) [Torres et al., 2007; 2013]. The AI is calculated by separating the spectral contrast of radiances due to aerosol effects and those due to Rayleigh scattering at two wavelengths (354 and 388 nm) in the near-UV region. Simulation We follow the method of Buchard et al., (2014) for simulating the AI using the vector radiative transfer model VLIDORT (Spurr 2006): Wavelength range (nm) Mean simulated AAE values for the four biomass burning regions considered when including BrC Mean AAE literature values for biomass burning aerosol References for the literature values 350-400 2.7 ± 0.06 2.5-3.0 Jethva and Torres, 2011 350-700 2.1 ± 0.04 1.9 Kirchstetter and Thatcher, 2012 450-550 2.3 ± 0.1 1.7 ± 0.07 Corr et al. 2012 400-700 1.9 ± 0.03 1.6 ±0.06 Schnaiter et al. 2005, Russell et al. 2010, Corr et al. 2012, Yang et al. 2009, Clarke et al. 2007 300-1000 1.7 ± 0.02 1.5 ± 0.2 Bergstrom et al. 2007, Yang et al. 2009, Kirchstetter et al. 2004 400-1000 1.3 ± 0.005 1.4 ± 0.02 Russell et al. 2010, Rizzo et al. 2011 1. Buchard, V., A. M. da Silva, P. R. Colarco, A. Darmenov, C. A. Randles, R. Govindaraju, O. Torres, J. Campbell, and R. Spurr (2014), Using the OMI Aerosol Index and Absorption Aerosol Optical Depth to evaluate the NASA MERRA Aerosol Reanalysis, Atmos. Chem. Phys. Discuss., 14(23), 32177–32231, doi:10.5194/acpd-14-32177-2014. 2. Jethva, H., and O. Torres (2011), Satellite-based evidence of wavelength-dependent aerosol absorption in biomass burning smoke inferred from Ozone Monitoring Instrument, Atmos. Chem. Phys., 11(20), 10541–10551, doi:10.5194/acp-11-10541-2011. 3. Mao, J., S. Fan, D. J. Jacob, and K. R. Travis (2013), Radical loss in the atmosphere from Cu-Fe redox coupling in aerosols, Atmos. Chem. Phys., 13(2), 509–519, doi:10.5194/acp-13- 509-2013. 4. Spurr, R. J. D. (2006), VLIDORT: A linearized pseudo-spherical vector discrete ordinate radiative transfer code for forward model and retrieval studies in multilayer multiple scattering media, J. Quant. Spectrosc. Radiat. Transf., 102(2), 316–342, doi:10.1016/j.jqsrt.2006.05.005. 5. Torres, O., A. Tanskanen, B. Veihelmann, C. Ahn, R. Braak, P. K. Bhartia, P. Veefkind, and P. Levelt (2007), Aerosols and surface UV products from Ozone Monitoring Instrument observations: An overview, J. Geophys. Res., 112(D24), D24S47, doi:10.1029/2007JD008809. Comparing Base Case Simulation to OMI AI Interpreting the OMI ultraviolet Aerosol Index to understand absorption by organic carbon aerosols and implications for atmospheric oxidation Melanie S. Hammer 1* , Randall V. Martin 1,2 , Aaron van Donkelaar 1 , Virginie Buchard 3 , Omar Torres 3 , Robert J.D. Spurr 4 1. Dalhousie University 2. Harvard Smithsonian Center for Astrophysics 3. NASA/Goddard Space Flight Center 4. RT Solutions Inc We develop a simulation of the ultraviolet Aerosol Index (AI) using the Vector Linearized Discrete Ordinate Radiative Transfer model (VLIDORT) coupled with GEOS-Chem aerosol fields. The simulation is used to interpret AI observations from the Ozone Monitoring Instrument (OMI). There is good agreement between simulated and observed values in regions where mineral dust dominates the AI, but a large negative bias exists in biomass burning regions. The addition of absorbing brown carbon (BrC) to the model decreases the overall mean bias between simulated and OMI AI values over major biomass burning regions from -0.62 to +0.004. The inclusion of BrC in the model decreases tropospheric OH concentrations by up to 20% over major biomass burning regions, reducing the overestimate of OH concentrations compared to observations. Summary Introduction Aerosols with spectrally independent absorption display an AAE ~ 1. Aerosols with spectrally dependent absorption have an AAE > 1. Black carbon exhibits an AAE ~ 1. In the base case simulation, all biomass burning regions have an AAE ~1. Biomass burning aerosol typically display an AAE~2 over the near-UV to visible and near-IR spectral regions. The AAE values calculated from the simulation including BrC are in good agreement with the literature for major biomass burning regions. Absorption Angstrom Exponent (AAE) AAOD = kλ AAE References Spectral Dependence of Aerosol Absorption AI > 0 : indicates absorbing aerosol AI < 0 : indicates non-absorbing aerosol AI ~ 0 : indicates clouds or aerosol-free atmosphere Depends on: aerosol concentration, aerosol layer height, and aerosol optical properties GEOS-Chem Simulation of Aerosol Composition Coincident with OMI TOMS UV Surface Reflectance (from Omar Torres) OMI Viewing Geometry VLIDORT AI = 100 log 10 I 354 RAY+AER I 354 RAY R 354 Pure Rayleigh Scattering Including Absorbing Aerosols Top of Atmosphere Radiance Wavelength The two main aerosol types detected by the AI are biomass burning and mineral dust aerosol. Jethva and Torres (2011) first documented the need to include absorption by BrC to accurately interpret UV radiances and the AI. Region r Mean Bias North Africa (January) 0.54 -0.60 South Asia (April) 0.39 -0.40 South Africa (July) 0.04 -1.0 South America (September) 0.20 -0.50 *Mean Bias = simulated AI – observed AI Aerosol Index (AI) Biomass burning aerosols consist mainly of black and organic carbon. Traditionally, black carbon is assumed to be the sole absorbing carbonaceous aerosol species in models. Several recent studies have found evidence of absorption by a subset of organic carbon known as “brown carbon” (BrC), which is thought to contribute significantly to the overall absorption by biomass burning aerosol. We introduce BrC to GEOS-Chem and examine its effect on aerosol absorption in biomass burning regions. BrC is introduced to the simulation by constraining the imaginary part of the refractive index for primary organic carbon (POC) between 300 and 500 nm using the spectral dependence of absorption recommended in the literature and the OMI AI. All other properties of BrC are assumed to be the same as the scattering POC currently used in GEOS-Chem. The simulation including BrC is more consistent than the base case simulation at reproducing the OMI AI over major biomass burning regions. Comparing Simulation Including BrC to OMI AI -0.5 0 0.5 1 1.5 2 2.5 3 Simulation Including BrC AI January April July September -0.5 0 0.5 1 1.5 2 2.5 3 AI OMI Observations Base Case Simulation July January April September January April July September Effect of BrC Absorption on OH Concentrations The figure below shows the percent changes in OH concentrations in the lower troposphere due to the addition of absorbing brown carbon to the GEOS-Chem simulation. The tendency of models to overestimate OH concentrations [Mao et al., 2013] is reduced. -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 ΔOH (%) January April July September OH concentrations decrease by up to 20% over the major biomass burning regions. Region r Mean Bias North Africa (January) 0.68 -0.08 South Asia (April) 0.60 -0.003 South Africa (July) 0.55 -0.24 South America (September) 0.54 +0.34 *Mean Bias = simulated AI – observed AI

Interpreting the OMI ultraviolet Aerosol Index to ...acmg.seas.harvard.edu/presentations/IGC7/posters/...July September-0.5 0 0.5 1 1.5 2.5 3 AI OMI Observations Base Case Simulation

  • Upload
    others

  • View
    4

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Interpreting the OMI ultraviolet Aerosol Index to ...acmg.seas.harvard.edu/presentations/IGC7/posters/...July September-0.5 0 0.5 1 1.5 2.5 3 AI OMI Observations Base Case Simulation

• The base case simulation of

the AI, including scattering

organic carbon as currently

used in GEOS-Chem, is

compared to corresponding

OMI observations.

• The simulation captures major

absorption features compared

to OMI over desert regions,

however it fails to capture

absorption by biomass

burning aerosol.

• The correlation between the

simulated and OMI AI is low in

all major biomass burning

regions with large mean

biases.

• The AI is a method of detecting absorbing aerosols from satellite measurements, and is a

product of the Ozone Monitoring Instrument (OMI) [Torres et al., 2007; 2013].

• The AI is calculated by separating the spectral contrast of radiances due to aerosol effects and

those due to Rayleigh scattering at two wavelengths (354 and 388 nm) in the near-UV region.

SimulationWe follow the method of Buchard et al., (2014) for simulating the AI using the vector radiative

transfer model VLIDORT (Spurr 2006):

Wavelength range (nm)

Mean simulated AAE values for

the four biomass burning regions

considered when including BrC

Mean AAE literature

values for biomass

burning aerosol

References for the literature values

350-400 2.7 ± 0.06 2.5-3.0 Jethva and Torres, 2011

350-700 2.1 ± 0.04 1.9 Kirchstetter and Thatcher, 2012

450-550 2.3 ± 0.1 1.7 ± 0.07 Corr et al. 2012

400-700 1.9 ± 0.03 1.6 ±0.06Schnaiter et al. 2005, Russell et al. 2010, Corr et

al. 2012, Yang et al. 2009, Clarke et al. 2007

300-1000 1.7 ± 0.02 1.5 ± 0.2Bergstrom et al. 2007, Yang et al. 2009,

Kirchstetter et al. 2004

400-1000 1.3 ± 0.005 1.4 ± 0.02Russell et al. 2010,

Rizzo et al. 2011

1. Buchard, V., A. M. da Silva, P. R. Colarco, A. Darmenov, C. A. Randles, R. Govindaraju, O. Torres, J. Campbell, and R. Spurr (2014), Using the OMI Aerosol Index and Absorption Aerosol Optical Depth to evaluate the NASA MERRA Aerosol Reanalysis, Atmos. Chem. Phys. Discuss., 14(23), 32177–32231, doi:10.5194/acpd-14-32177-2014.

2. Jethva, H., and O. Torres (2011), Satellite-based evidence of wavelength-dependent aerosol absorption in biomass burning smoke inferred from Ozone Monitoring Instrument, Atmos. Chem. Phys., 11(20), 10541–10551, doi:10.5194/acp-11-10541-2011.

3. Mao, J., S. Fan, D. J. Jacob, and K. R. Travis (2013), Radical loss in the atmosphere from Cu-Fe redox coupling in aerosols, Atmos. Chem. Phys., 13(2), 509–519, doi:10.5194/acp-13-509-2013.

4. Spurr, R. J. D. (2006), VLIDORT: A linearized pseudo-spherical vector discrete ordinate radiative transfer code for forward model and retrieval studies in multilayer multiple scattering media, J. Quant. Spectrosc. Radiat. Transf., 102(2), 316–342, doi:10.1016/j.jqsrt.2006.05.005.

5. Torres, O., A. Tanskanen, B. Veihelmann, C. Ahn, R. Braak, P. K. Bhartia, P. Veefkind, and P. Levelt (2007), Aerosols and surface UV products from Ozone Monitoring Instrument observations: An overview, J. Geophys. Res., 112(D24), D24S47, doi:10.1029/2007JD008809.

Comparing Base Case Simulation to OMI AI

Interpreting the OMI ultraviolet Aerosol Index to understand absorption by organic

carbon aerosols and implications for atmospheric oxidationMelanie S. Hammer1*, Randall V. Martin1,2, Aaron van Donkelaar1, Virginie Buchard3, Omar Torres3, Robert J.D. Spurr4

1. Dalhousie University 2. Harvard Smithsonian Center for Astrophysics 3. NASA/Goddard Space Flight Center 4. RT Solutions Inc

We develop a simulation of the ultraviolet Aerosol Index (AI) using the Vector Linearized Discrete

Ordinate Radiative Transfer model (VLIDORT) coupled with GEOS-Chem aerosol fields. The

simulation is used to interpret AI observations from the Ozone Monitoring Instrument (OMI). There

is good agreement between simulated and observed values in regions where mineral dust

dominates the AI, but a large negative bias exists in biomass burning regions. The addition of

absorbing brown carbon (BrC) to the model decreases the overall mean bias between simulated

and OMI AI values over major biomass burning regions from -0.62 to +0.004. The inclusion of BrC

in the model decreases tropospheric OH concentrations by up to 20% over major biomass burning

regions, reducing the overestimate of OH concentrations compared to observations.

Summary

Introduction

• Aerosols with spectrally independent absorption display an AAE ~ 1.

• Aerosols with spectrally dependent absorption have an AAE > 1.

• Black carbon exhibits an AAE ~ 1.

• In the base case simulation, all biomass burning regions have an AAE ~1.

• Biomass burning aerosol typically display an AAE~2 over the near-UV to visible and near-IR spectral

regions.

The AAE values calculated from the simulation including BrC are in good agreement with the literature for major

biomass burning regions.

Absorption Angstrom

Exponent (AAE)AAOD = kλ−AAE

References

Spectral Dependence of Aerosol Absorption

AI > 0 : indicates absorbing aerosolAI < 0 : indicates non-absorbing aerosolAI ~ 0 : indicates clouds or aerosol-free atmosphereDepends on: aerosol concentration, aerosol layer height, and aerosol optical properties

GEOS-Chem Simulation of Aerosol Composition Coincident with OMI

TOMS UV Surface Reflectance (from Omar Torres)

OMI Viewing Geometry

VLIDORT

AI = − 100 ∙ log10I354RAY+AER

I354RAY

R354∗

Pure Rayleigh Scattering

Including Absorbing Aerosols

Top of

Atmosphere

Radiance

Wavelength

• The two main aerosol types detected by the AI are biomass burning and mineral dust aerosol.

• Jethva and Torres (2011) first documented the need to include absorption by BrC to accurately

interpret UV radiances and the AI.

Region r Mean Bias

North Africa (January) 0.54 -0.60

South Asia (April) 0.39 -0.40

South Africa (July) 0.04 -1.0

South America(September)

0.20 -0.50

*Mean Bias = simulated AI – observed AI

Aerosol Index (AI)

Biomass burning aerosols consist mainly of black and organic carbon. Traditionally, black carbon is

assumed to be the sole absorbing carbonaceous aerosol species in models. Several recent studies

have found evidence of absorption by a subset of organic carbon known as “brown carbon” (BrC),

which is thought to contribute significantly to the overall absorption by biomass burning aerosol. We

introduce BrC to GEOS-Chem and examine its effect on aerosol absorption in biomass burning

regions.

• BrC is introduced to the simulation by constraining the imaginary part of the refractive index for

primary organic carbon (POC) between 300 and 500 nm using the spectral dependence of

absorption recommended in the literature and the OMI AI.

• All other properties of BrC are assumed to be the same as the scattering POC currently used in

GEOS-Chem.

The simulation including BrC is more consistent than the

base case simulation at reproducing the OMI AI over

major biomass burning regions.

Comparing Simulation Including BrC to OMI AI

-0.5 0 0.5 1 1.5 2 2.5 3

Simulation Including BrC

AI

January April

July September

-0.5 0 0.5 1 1.5 2 2.5 3AI

OMI Observations

Base Case Simulation

July

January April

September

January April

July September

Effect of BrC Absorption on OH Concentrations

• The figure below shows the percent changes in OH concentrations in the lower troposphere due

to the addition of absorbing brown carbon to the GEOS-Chem simulation.

• The tendency of models to overestimate OH concentrations [Mao et al., 2013] is reduced.

-20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0

ΔOH (%)

January April

July September

OH concentrations decrease by up to 20% over the major biomass burning regions.

Region r Mean Bias

North Africa (January) 0.68 -0.08

South Asia (April) 0.60 -0.003

South Africa (July) 0.55 -0.24

South America(September)

0.54 +0.34

*Mean Bias = simulated AI – observed AI