A method is presented to correct for the effects of dust opacity on the der
ivation of water-vapor abundances from observed spectral absorption lines.
We show that the dust opacity can be inferred by use of limb-to-limb measur
ements of a CO2 band. Ideally, simultaneous CO2 and water-vapor measurement
s at the same location should be obtained; in practice, we make these two m
easurements sequentially. The effects of the dust opacity are computed by a
doubling-adding program that has been applied to many planetary atmospheri
c problems, one of which is obtaining the phase function and single-scatter
ing albedo of the dust in Mars' atmosphere observed by the cameras on the M
ars Pathfinder lander. Sample analyses of data from three nights in March 1
999 are shown. For these results there does appear to be a real diurnal var
iation in the H2O abundance. (C) 2000 Academic Press.