This page's content is no longer actively maintained, but the material has been kept on-line for historical purposes.
The page may contain broken links or outdated information, and parts may not function in current web browsers.

GACP Meetings

First Science Team Meeting

Notes from Discussion Group A: Development and Use of Satellite Retrieval Algorithms

Moderator: Ralph Kahn; Rapporteur: Philip A. Durkee

How might we refine the AVHRR algorithm being used for near-term and retrospective satellite constraints on aerosol column optical depth and microphysical properties? [Direct Effect]

1. INPUT DATA

  • AVHRR
    • 1-channel (Stowe et al.) or
    • 2-channel (Nakajima et al.; Durkee; Mishchenko)
    • [0.63 and 0.91 microns]
  • Particle Microphysical Models (assumed)
    • Mode 1 particle size distribution (reff1, sigma1)
    • Mode 2 particle size distribution (reff2, sigma2) [Nakajima; Durkee]
    • Particle index of refraction ( nr(r,z), ni(r,z) )
    • Particle shape (s)
    • Particle vertical distribution ( N(z) )
  • Environmental Model (partly constrained by data)
    • Ozone column abundance for radiance correction (L1)
    • Water vapor column abundance for radiance correction (L2)
    • Relative humidity profile for particle hydration state
    • Ocean surface reflectivity ( depends on wind speed, silt, phytoplankton)
    • Land surface reflectivity

2. OUTPUTS

  • Aerosol Column Optical Depth
  • Aerosol Column Microphysical Constraint
    • Ratio of mode 1 to mode 2 contributions to optical depth [Nakajima; Durkee]
    • Angstrom Coefficient, related to effective particle size [Mishchenko, Stowe]

3. ISSUES

  • Spectral calibration of AVHRR
  • Particle and Environmental Models must be region-specific, and depend on Season, Time-of-Day, and Meteorology
  • Need to determine which attributes of the aerosol column are most effectively constrained by the observations (sensitivity studies)
  • Need to determine which attributes of the assumed particle and environmental models matter most to the retrieval results, and how (quantitatively) the uncertainties can be diminished by specific, additional measurement programs, and further analysis of existing data.
  • Need to assess the bias introduced since retrievals can be done only over "cloud-free" regions (also need to quantify "cloud-free")
  • Need to assess the tradeoffs between retrieving aerosol properties and directly retrieving radiation parameters (e.g., Li)
  • Assimilation of multiple data sets
  • Need to revise these notes, to include other research efforts

Return to Conference Homepage