Why are Radiative Transfer Atmospheric Correction models so hard to use with archived satellite data?