GE0-CAPE WorkshopUniversity of North Carolina-Chapel Hill

18-20 August 2008

Aerosols: What is measurable and by what remote sensing technique?

Omar TorresHampton University

(with contributions from Lorraine Remer, Ralph Kahn, Shohba Kondragunta,Sundar Christopher, Ana Prados)

Aerosol Types and Origin• Aerosol particles larger than about 1

m in size are produced by windblown dust and sea salt from sea spray and bursting bubbles

• Aerosols smaller than 1 µm are

mostly formed by condensation processes such as conversion of sulfur dioxide (SO2) gas to sulfate particles and by formation of soot and smoke during burning processes.

• After formation, the aerosols are

mixed and transported by atmospheric motions and are primarily removed by cloud and precipitation processes.

MicrophysicalSize distribution Refractive indexShape Vertical distribution

Macrophysical propertiesScattering function (P)Single scattering albedo (SSA)Ext. Optical thickness (AOT)

Particle Scattering Theory

AOText = AOTsca + AOTabs

SSA = AOTsca /(AOTsca + AOTabs)

Scattering Phase Function (P):

AERONET Measurements

Aerosol Physical Properties

Aerosol properties measurable from passive satellite remote sensing

-Visible and Near-IR scattering AOT (usually reported as extinction AOT with an assumption on single scattering albedo) by MODIS, MISR.

-Observations at 412 nm (Deep Blue) used to derive aerosol AOT and SSA over deserts (MODIS)

-Near-UV observations can be used to derive information on aerosol absorptionQualitative (Aerosol Index TOMS / OMI) Quantitative if aerosol vertical distribution is known.

Polarization measurements (POLDER, PARASOL, Glory-APS) n/a to GEO-CAPE

Retrieval Issues

Correction for gas absorption (minor issue)sub-pixel cloud contamination (depends on spatial resolution) (7km resolution better than OMI’s (13X24) but still non-optimum)

‘Surface’ correction:

Ocean:-Ocean color effects (chlorophyll conc., dissolved organic matter)- Glint effects (viewing geometry)- White caps (foam reflectance, wind speed) Land : High surface reflectance -Difficulty to separate aerosol signal from bright background in the vis and near IR -Less of a problem in the near UV BRDF effects Angular dependence of surface reflection

380 nm

440 nm

630 nmSurface Albedo from GOME observations

Aerosol reflectance at 2.1 microns is negligibly small so that TOA measurementsare a direct measurement of surface albedo (R2.1).

Relationships between (R2.1) and R0.47 and R0.66 were developed based onobservations.

R0.47 = 0.25R2.1

R0.66 = 0.5R2.1

Handling of land reflectance issue in MODIS algorithm

(collection 4 model)

In collection 5, proportionality constantsare a function of geographical location.

This parameterization of surface reflectanceallows AOT retrievals over most land surfaces.It does not work over deserts.

Current VIS and near-IR satellite aerosol products

MODIS aerosol product over oceans: - AOT at 0.47, 0.55, 0.66, 0.87, 1.24, 1.63, 2.13 microns - AOTfine / AOTtotal

- Effective radius (Based on spectral dependence of AOT)

MODIS aerosol product over land:

- AOT at 0.47 and 0.66 microns:- AOTfine / AOTtotal

-Qualitative Information on particle size distribution via the Angstrom Exponent

- AOT- SSA over deserts from Deep Blue Algorithm

• Ocean product (10kmx10km):

– Total Spectral Optical thickness

– Effective radius

– Optical thickness of small & large modes/ratio between the 2 modes

• ~ ±0.03±0.05 (dust

excepted)

Aerosol Effective Radius2.00.0 1.00.5 1.5

Aerosol Optical Depth0.80.0 0.40.2 0.6

September, 2000

Small Mode Fraction1.00.0 0.500.25 0.75

One MODIS Aerosol Type Classification: Low AOT (blue), High AOT+Coarse (green), High AOT+Fine (red)

Kaufman et al., JGR, 2005

Absorbing aerosols as seen in the near-UV

Long-range aerosol transport takes place in the free troposphere, frequently above clouds.

The Absorbing Aerosol Index

]))(

)((log)([log100

331

360

331

360calc

sfc

sfcmeas RI

RI

I

IAAI

calcmeas II )()( 331331

Thus, the AAI definition reduces to:

])((

)([log100

360

360

calcsfc

meas

RI

IAAI

Rsfc is a Lambert Equivalent effective surface reflectvity value such that

Rsfc is assumed wavelength independent

Advantages of near-UV aerosol observations

Sensitivity to Aerosol Absorption-Aerosol Detection Capability over all surface types:

All vegetated surfaces deserts Oceans

Ice-snow covered surfaces Above clouds and inter-mingled with clouds

-Sensitive to Aerosol Layer Height

Quantitative Near UV Retrieval Products (OMI)

By means of an inversion algorithm AOD and SSA are derived

March 9, 2007

The observed near-UV spectral contrast is conveniently ‘packed’ as the UV Aerosol Index

June 27-08 (OMI) July 6-06 (OMI)

April 25-98 (TOMS) May 16 1998 (TOMS) Oct 1-07 (OMI)

Long range transport of aerosols into GE0-CAPE’s coverage area

Aerosol type identification with UV-VIS observations

-Near UV-VIS spectral information can be used for aerosol type identification -Requires knowledge of spectral surface albedo

• Satellite data have gaps due to clouds. Problem is particularly bad for polar-orbiting satellites as they see a particular location on the Earth only once a day (in UV-VIS)

• Geostationary satellites due to their rapid refresh rate, can obtain a more complete temporal coverage

Advantage of aerosol observations from Geostationary Satellites

Single snap shot of MODISComposite image from multiple snap shots of GOES-12

No MODIS retrieval due to bright surface

Aerosols close to the surface?

MODIS treats thick smoke as cloud

Combined use of MODIS-OMI observations for aerosol detection

From Shohba Kondragunta