excess of 200 of that for 5S. The comparison of the output from the 5S pseudo-code with that 
produced using the 5S code indicate that the operational efficiency of the pseudo-code is 
achieved without an appreciable loss in the accuracy of the result. 
The observed differences between the predictions of the 5S model and those of the 5S pseudo-code 
are associated with the separation of the computation of gaseous transmittance across the band 
from the pre-computation phase for atmospheric parameters. In effect, a systematic variation 
in gaseous transmittance across a band can produce a ramp effect, or spectral weighting, similar 
to a spectra] response function. However, a spectral weighting related to absorption variation 
across the band cannot be incorporated into the pre-computation phase because the levels of 
absorption are related to the atmospheric path length which is itself a function of the illumination 
and viewing geometry. Therefore, discrepanies between the model predictions increase as the 
levels of spectral variation in the gaseous transmittance increase across a spectral band. This 
is most pronounced in channel lb, then V2; channel VI and V3 have uniform levels of gaseous 
transmittance across the spectral bands. This explains the variation in the percentage differences 
indicated on table ( 1 ). 
The proposed operational code could be applied to data from any satellite sensor if pre-computed 
parameters are produced for the particular spectral band. Pre-computation of parameter sets 
has already been done for SPOT HRV-2 bands 1, 2 & 3 and Landsat-5 TM bands 1 to 5 and 
7 (for the same atmospheric and aerosol models as those used for ATSR-2), with gaseous 
absorbing parameters at 5nm increments. The 5S code has to be modified to output these 
pre-computation phase parameters. 
Finally, an important feature of the atmospheric modelling provided by 5S or the 5S pseudo-code 
which relates to the overall atmospheric correction procedure is that by using standard 
atmospheric profiles and aerosol models, the variability in atmospheric optical properties in all 
four spectral bands can be related solely to changes in the aerosol optical depth at the reference 
wavelength x A (550run). Therefore within the overall atmospheric correction procedure, the 
success will depend on the ability of the procedure to solve for this single parameter. 
5 REFERENCES 
McClatchey, R.A., Fenn, R.W., Selby, J.E.A., Volz, F.E., Garing J.S., 1971, Optical properties 
of the atmosphere. AFCRL-TR-71-0279, Environmental research paper, No. 354, L.G. Hancom 
Field, Bedford, Massachusetts. 
Richter, R., (1990), A fast atmospheric correction algorithm applied to Landsat TM images, 
Int. J. Remote Sensing, Vol. 11, No. 1, 159-166. 
Tanre, D., Deroo, C., Duhaut, P., Herman, M., Morcette, J.J., Perbos, J. and Deschamps, P.Y., 
(1986), Simulation of the satellite signal in the solar spectrum (5S), Laboratoire d’Optique 
Atmospherique, Universite des Sciences et Techniques de Lille, 59655 Villeneuve d’Ascq Cedex, 
France. 
Tanre, D., Deroo, C., Duhaut, P., Herman, M., Morcette, J.J., Perbos, J. and Deschamps, P.Y., 
(1990), Description of a computer code to simulate the satellite signal in the solar spectrum: 
the 5S code, Int. J. Remote Sensing, Vol. 11, No. 4, 659-668. 
Teillet, P. M., (1992), An algorithm for the radiometric and atmospheric correction of AVHRR 
data in the solar reflective channels, Remote Sens. Environ., Vol. 41, 185-195. 
WMO (World Meteorological Organization), 1986, A preliminary cloudless standard atmosphere 
for radiation computation. World Climate Program, WCP-112, WMO/TD-No.24.