Characterization of the atmosphere

Atmospheric gases, aerosols and vapors contribute to absorb, scatter and refract the incident solar radiation and reflected from the surface. This set of phenomena, which depend on wavelength, is called "atmospheric effect" and can be an obstruction to remote sensing because:
  • blurs the image by reducing the contrast between objects;
  • hinders the temporal comparison of image data as contaminated by the variability of the optical properties of the atmosphere at the time of recovery;
  • complicates the derivation of the spectral signature of the observed surfaces.
In many applications that require a physical interpretation of the image, it is therefore often necessary an appropriate correction of atmospheric effects. The methods to be used for the correction of atmospheric effects can be grouped into two broad categories:
  • radiative transfer models;
  • image-based simplified methods.
The first category includes models that solve analytically the equations of radiative transfer of electromagnetic energy through the atmosphere. To accurately describe the propagation of radiation, these models require data on the optical properties of the atmosphere. This information can be obtained from in situ measurements simultaneous to the time of image acquisition or, in case of operational missions (e.g. MODIS) the information can be retrieved from image data itself. The second group includes a series of methods that simplify the equations of propagation of radiation in the atmosphere and derive information about the characteristics of the atmosphere directly from the image.