Aerosol Properties and Behavior Studies for Atmospheric Correction

Document Sample
Aerosol Properties and Behavior Studies for Atmospheric Correction Powered By Docstoc
					         Draft proposal to NASA Project SIMBIOS (Jan 19, 2000)


  Physically-Based Aerosol Models
                 for
 Atmospheric Correction Algorithms




     Washington University, St. Louis, MO
Center for Air Pollution Impact and Trend Analysis (CAPITA).
        Table of Contents

•   1
•   2
•   3
•   4
                                           Abstract
The objective of the project is to improve the quality of atmospheric correction algorithms by applying
    physically based aerosol models, multi-sensory data fusion and signal decomposition procedures.
The aerosol is composed of multiple aerosol layers, each consisting of a specific aerosol type such as
    windblown dust, biomass smoke, sea salt, volcanic, biogenic and urban-industrial haze aerosols.
    Each aerosol type tends to occur over specific global regions and atmospheric strata.
Algorithms will be developed to retrieve the spatio-temporal pattern of each aerosol type. The available
    optical properties (spectral extinction, backscattering and single scatter albedo) as well as the
    dependence on ambient weather parameters (e.g. humidity) will be updated and extended. A key
    challenge will be to decompose the total reflectance signal into aerosol and surface components and
    subsequently separating the contribution of different aerosol types. The apportionment of the total
    aerosol signal into dust, smoke, salt, volcanic, biogenic and industrial haze will be based on the
    combination of aerosol climatological data and instantaneous satellite and in-situ observations.
    Surface-based monitoring (e.g. AERONET, size distribution and chemical analysis) will provide the
    detailed aerosol characteristics. Satellite data from SeaWiFS, POLDER, MODIS, MISR, TOMS,
    GOES/GMS/METEOSAT etc will provide the spatial pattern as well as the spectral, angular
    backscattering and polarization parameters. The multi-sensory data fusion will use a physical model
    for aerosol size distribution dynamics, a chemical composition model for refractive index and Mie
    theory for optical properties.
The benefits of the new aerosol models, aerosol data fusion and signal decomposition methodology will
    include (1) Improved aerosol correction by physically and geographically constraining the aerosol
    selection procedure, () Delivering separate aerosol products for dust, smoke, salt, volcanic, biogenic
    and haze that will be useful for climate change biogeochemical and other studies and (3)
    Extendibility through multi-sensory fusion and multi-institutional collaboration.
                             Background
• Aerosols are a nuisance to satellite remote sensing for surface retrieval
   – Broad-band interference throughout the spectrum
   – Variable optical properties dependent on aerosol types and environment
     therefore it is difficult to accurately account for aerosol interference
   – Space-time variation due to sources, transport, transformation processes
• The problem is separating the surface and atmospheric signals
   – Proper separation yields an undistorted surface color
   – It also provides an atmospheric aerosol signal useful for other studies
• Past atmospheric corrections were hampered by inadequate data
   – No aerosol spectral-angular scattering and absorption properties. No global
     distribution, No temporal covarage
   – Atmospheric correction was restricted to data obtained by that particular remote
     sensor,
   – Single sensor in inherently limited not enough on spectral angular and
     absorption characteristics of aerosol
   – Limited auxiliary data and virtually no data fusion
                Opportunities
• Global satellite remote sensors covering ocean
  and land
• Aerosol properties from many regional fields
  studies
• Technological advances that facilitated multi-
  sensory data exchange, integration and fusion
     Aerosol Detection over Land and The Oceans
• Ocean color is particularly interesting near the continental shores
  where much of the biological activity is takes place.
• Most of the atmospheric aerosols are generated over land. Even over
  the oceans most of the radiative perturbation is due to aerosols from
  the continents.
• The new ocean color sensors allow remote sensing of both oceans
  and land.
        The Objectives of the Project:
• Integrate the communal knowledge on the properties and
  environmental behavior of the major aerosol types that perturb
  satellite remote sensing of surface color: windblown dust,
  biomass smoke, sea salt, volcanic, biogenic and urban-industrial
  haze aerosols
• Formulate and test algorithms on data fusion and decomposition
  procedures for the separate retrieval of each of the above aerosol
  types using data from multiple satellite and surface sensors
• Deliver the new aerosol retrieval algorithms to the SIMBIOS and
  broader community for incorporation into atmospheric correction
  algorithms on operational and research level.
Expected Significance of the Project:
• The project is expected to improve the quality of aerosol correction algorithms
  by better characterization of the aerosol as a multi-component, dynamic
  physical system
• The project will facilitate separate retrieval of dust, smoke, sea salt, volcanic,
  biogenic and industrial haze aerosol products which will be useful for global
  climate change, biogeochemical and other studies.
• The approach is based on physical principles and it can evolve into an
  integrated data-model assimilation procedure for atmospheric aerosols.
• The resulting aerosol models will evolve through open consensus-based
  approach using the Internet and will incorporate the experience of both the
  atmospheric aerosol and the atmospheric correction research communities.
• The disadvantages of the proposed approach include (1) more demand on
  aerosol and environmental input data, (2) more intense computation and (3)
  several years needed before it will be fully operational
  Past Aerosol Models and Atmospheric Correction
Past Aerosol Models
    Junge, 1963 – power law
    Whitby and Husar 1972 – bimodal
    Shettle and Fenn, 1978 – bimodal, RH, components
    D‟Almeida and Koepke 1980 –components, spatial
Past Aerosol Models for Atmospheric Correction
    CZCS, SeaWiFS - Gordon, Wang et al
    AVHRR – Stowe, Ignatov, Durkee
    POLDER – Leroy, Tanre, Vermote
    MODIS – King, Kaufman et al
               Recent Work Related to this Project




Global Oceanic Aerosol based on AVHRR         Global Continental Aerosol Based on Surface Visibility




Haze Retrieval over Ocean and Land from SeaWiFS   Asian Dust Size Distribution by Different Methods
     Approach to the Aerosol Model Development
              and Implementation Procedures
1.    Assume that the vertical aerosol at any give point is the sum of windblown
      dust, biomass smoke, sea salt, volcanic, biogenic and urban-industrial haze
2.    Gather the best available knowledge on the general physico-optical properties
      of each aerosol type including dependence on humidity, atmospheric
      residence time,..
3.    Regionalize and seasonalize each aerosol type
4.    Define properties for each regional/seasonal aerosol type by constraining its
      functional form. Some loose scaling factors are included to “fine tune” the
      aerosol type model.
5.    Gather and integrate instantaneous aerosol data from space and surface
      sensors to determine the model scaling constants
6.    Statistically fit the remaining unknown factors to determine the appropriate
      form for each aerosol type model
7.    Apply the appropriate optical model to each aerosol type and calculate aerosol
      optical parameters relevant to atmospheric correction.
8.    Deliver „best estimate‟ aerosol-optical properties for atmospheric correction.
              Vertical Pattern of Global Aerosol
•   Windblown Dust (crustal elements)
•   Biomass Smoke (organics, H20 )
•   Sea H20 salt (NaCl. H20)
•   Stratospheric (Volcanic) (H2SO4)
•   Biogenic (Non-sea salt sulfate, org)
•   Urban-Industrial Haze (SO4, org.
    H20)

• Dust, smoke, volcanic aerosol and
  industrial haze originate from land
• The global aerosol concentration is
  highest over land and near the
  continents over the oceans
  (coastal regions)
• Sea salt is significant over some
  of the windy oceanic regions and
  biogenic sulfate and organic
  aerosols also occur …
         Regional Aerosol Studies:
ACE Australia, ACE Africa, SAFARI, SCAR-B

•   ACE - Australia
•   ACE - W. Africa
•   SCAR – Brazil
•   SAFARI – S. Africa
                 Windblown Dust
•   Aerosol Properties
     – Size distribution
     – Chemical composition and refractive index
     – Spectral extinction, backscattering and single scatter albedo
•   Variation of properties with environmental conditions
     – Humidity effect including clouds
     – Transformation/removal effects (chemical, settling)
•   Spatio-temporal distribution
     – Climatological/seasonal pattern over the ocean and land
     – Daily distribution
     – Vertical distribution
                   Biomass Smoke
•   Aerosol Properties
     – Size distribution
     – Chemical composition and refractive index
     – Spectral extinction, backscattering and single scatter albedo
•   Variation of properties with environmental conditions
     – Humidity effect including clouds
     – Transformation/removal effects (chemical, settling)
•   Spatio-temporal distribution
     – Climatological/seasonal pattern over the ocean and land
     – Daily distribution
     – Vertical distribution
                              Sea Salt
•   Aerosol Properties
     – Size distribution
     – Chemical composition and refractive index
     – Spectral extinction, backscattering and single scatter albedo
•   Variation of properties with environmental conditions
     – Humidity effect including clouds
     – Transformation/removal effects (chemical, settling)
•   Spatio-temporal distribution
     – Climatological/seasonal pattern over the ocean and land
     – Daily distribution
     – Vertical distribution
                 Volcanic Aerosol
•   Aerosol Properties
     – Size distribution
     – Chemical composition and refractive index
     – Spectral extinction, backscattering and single scatter albedo
•   Variation of properties with environmental conditions
     – Humidity effect including clouds
     – Transformation/removal effects (chemical, settling)
•   Spatio-temporal distribution
     – Climatological/seasonal pattern over the ocean and land
     – Daily distribution
     – Vertical distribution
Marine Biogenic Haze Aerosol
•   Aerosol Properties
     – Size distribution
     – Chemical composition and refractive index
     – Spectral extinction, backscattering and single scatter albedo
•   Variation of properties with environmental conditions
     – Humidity effect including clouds
     – Transformation/removal effects (chemical, settling)
•   Spatio-temporal distribution
     – Climatological/seasonal pattern over the ocean and land
     – Daily distribution
     – Vertical distribution
          Urban-Industrial Haze
•   Aerosol Properties
     – Size distribution
     – Chemical composition and refractive index
     – Spectral extinction, backscattering and single scatter albedo
•   Variation of properties with environmental conditions
     – Humidity effect including clouds
     – Transformation/removal effects (chemical, settling)
•   Spatio-temporal distribution
     – Climatological/seasonal pattern over the ocean and land
     – Daily distribution
     – Vertical distribution
Aerosol and Surface Radiative Transfer
Extraterrestial      Attenuated                     Attenuated
Radiation            Surface Reflection             AerosolReflection

    Io            I o Ro e                            I o P1  e  
                                                                      
                                                                      


                                 Aerosol                PIo
                      Ro I o


    The intensity of reflected radiation, I is the sum
    of the attenuated surface and aerosol signals :

                                          
         I  I o R  I o Ro e   I o P 1  e    
         Apparent reflectance detected by the sensor:
          R  ( R0  (e  1) P )e 
                 Apparent Surface Reflectance, R
R is the perturbed reflectance sensed at the top of the atmosphere
The critical parameter whether aerosols will increase or decrease the apparent reflectance, R,
is the ratio of aerosol to surface reflectance, P/ R0

Aerosols will increase the apparent                                 1
surface reflectance, R, if P/R0 < 1.
                                                                   0.9
                                                                                                        R  ( R0  (e  1) P)e
For this reason, the reflectance of
                                                                                                                 For Haze P  0.38
ocean and dark vegetation increases                                0.8

with τ.                                                                                            Clouds at all wavelengths, P/Ro<o.5
                                                                   0.7




                                          Apparent Reflectance,R
When P/R0 > 1, aerosols will decrease                              0.6
the surface reflectance. Accordingly,
the brightness of clouds is reduced by                             0.5
                                                                             Soil at >0.6 um
overlying aerosols.                                                0.4       Vegetation at >0.6 um, P/Ro=1

The reflectance of soil and vegetation                             0.3

at 0.8 um is unchanged by haze
                                                                   0.2                          Vegetation at 0.5 um, P/Ro=2-5
aerosols since P~ R0.
                                                                   0.1
At large τ (radiation equilibrium),                                                Ocean at >0.6 um
                                                                                   Vegetation at 0.4 um, P/Ro>10
both dark and bright surfaces                                       0
                                                                         0       0.5      1       1.5       2        2.5       3   3.5   4
asymptotically approach the aerosol                                                           Aerosol Optical Thickness, AOT
reflectance, P
                Aerosol Effect on Surface Color
•   Aerosols add to the reflectance and sometimes reduce the reflectance of surface objects
•   Aerosols always diminish the contrast between dark a bright surface objects
•   They change the color of surface objects
•   Haze adds a bluish while dust adds yellowish tint to the surface color of surface objects.
                                           Role of Dust Aerosol on Reflectance


                                     Water                                                     Vegetation                                                     Soil

                       350                                                         350                                                         350




                                                                                                                          Reflectance x 1000
                                                              Reflectance x 1000
                       300                                                         300                                                         300
  Reflectance x 1000




                       250                                                         250                                                         250
                       200                                                         200                                                         200
                       150                                                         150                                                         150
                       100                                                         100                                                         100
                       50                                                          50                                                          50
                        0                                                           0                                                           0
                             0.3     0.5       0.7      0.9                              0.3     0.5        0.7     0.9                              0.3     0.5       0.7     0.9

                        CleanWater         DustyWater   P                          CleanVeget          DustyVeget     P                         Clean Soil         DustySoil    P




• tttt
Examples of Aerosol Reflectance Decomposition



 •   Sahara – Mediterranean
 •   Haze – EUS
 •   Smoke – Africa, Asia
 •   Dust & Haze – Mediterranean
                   Reflectance Decomposition
R = Rdust + Rsmoke + Rsalt + Rbiogenic + Rvolcanic + Rhaze


• Reflectance over the
  Mediterranean
                                Deliverables
• Updated Aerosol Models (size distr., chem. Comp, opt.)
    –   Windblown dust
    –   Biomass smoke
    –   Sea Salt
    –   Biogenic
    –   Volcanic
    –   Urban-industrial haze
• Methodologies and Data to Drive each of the Models
    – Satellite data from SeaWiFS, POLDER, MODIS, MISR
    – Visibility, AERONET, Mass and chemical conc.
    – Global model-derived parameters
• Open Facilities for the Evolution and Use of Aerosol Models
    – Incorporation into atmospheric correction algorithms
    – Incorporation into aerosol retrieval algorithms
    – Interactive web-based community stuff
   Management Approach to the Project

• Strong interaction with atmospheric aerosol and
  atmospheric correction communities
• Integration and updating the best available knowledge
  through open, participatory consensus-building process
  using internet
• Facilitation of strong international participation with
  emphasis on local participation
• There will be a „best available‟ aerosol model and
  methodology available to the community after year1

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:4
posted:7/17/2011
language:English
pages:27