Atmospheric corrections of AVIRI8 images with a procedure based on the inversion of the 58 model
Titel:
Atmospheric corrections of AVIRI8 images with a procedure based on the inversion of the 58 model
Auteur:
Zagolski, F. Gastellu-Etchegorry, J.P.
Verschenen in:
International journal of remote sensing
Paginering:
Jaargang 16 (1995) nr. 16 pagina's 3115-3146
Jaar:
1995-11-10
Inhoud:
An algorithm for atmospheric correction was developed for correcting AVIRIS (Airborne Visible and Infrared Imaging Spectrometer) images that were acquired during the 1991 Mac Europe campaign ofNASA/JPL over the 'Landes' (south-west France). The methodology is based on the inversion of the 5S atmospheric model, through an iterative procedure that uses the Gauss Seidel principle. The environmental effect is fully taken into account, on a pixel per pixel basis, by the usc of circular neighbourhoods the radii of which are variable with wavelength. Here, input parameters, i.e. optical characteristics of the atmosphere. are estimated with ill situ atmospheric profile and visibility measurements combined with the 5S model. The visual analysis of AVIRIS spectral bands in the blue region clearly showed a heterogeneous spatial distribution of aerosol etTects. Consequently, a procedure was developed which computes the aerosol optical depth directly from the image. The only assumption is the presence of dense dark vegetation with spectral reflectances lying in narrow intervals the bounds of which, unknown at first, are iteratively determined. Spectral bands centred at 459·8 nm (band 7), 489·4 nm (band 10) and 607·9 nm (band 22) were the most efficient sensor's bands for that approach. The spatial variation of the aerosol optical depth was [0·16-0·26], [0·15-0·23] and [0·10-0·18] in the 459·8 nm, 489·4 nm and 607·9 nm bands, respectively, with a mean 1·4 ångstrom exponent. Spectral aerosol optical depth maps were computed and used as input parameters in the atmospheric correction procedure. This converged after five iterations, for all AVIRIS spectral bands. This correction procedure was conducted with radii of circular neighbourhood ranging from 0 to 100 pixels, which allowed us to compare this approach with those procedures that do not take into account adjacency etTect or assume that the latter can be derived from the pixel value. Moreover, these computations allowed us to determine the minimal sizes of the circular neighbourhoods, for each spectral band, thus ensuring a good approximation of the adjacency effect; e.g., for the 459·8 nm band, a radius of 600m (i.e., 30 pixels) was necessary for obtaining corrected reflectances with an accuracy of 0·5 per cent.
Tijdschrift beschrijving