Spectroscopic investigation of the minor and trace constituents of the atmosphere from the vantage point of near-Earth orbit can be made either by diffusely reflected sunlight, by transmitted solar radiation (i.e. by absorption during sunrise or sunset occultations), or by thermal emission. Each of these methods has its advantages and disadvantages. In the Earth orbit case the choice depends on a number of factors, including the required temporal and spatial coverage, and the extent of a priori knowledge of the physical and compositional structure of the atmosphere. Examples of the effect of these considerations on the performance of contemporary absorption and emission instruments are given. These include the Shuttle-borne, high resolution absorption measurements made over the past decade by the ATMOS instrument (0.01 cm^-1, 2 to 16 µm), the emission measurements of meteorology and atmospheric radiation instruments such as IRIS-D and HIS (dv Å 1 cm^-1, 4 to 15 µm) and the planned EOS 0.1 cm^-1 emission instrument TES.

Two particular aspects of these methods are discussed: The first is the effect of spectral interference from the major infrared active gases (H₂O, CO₂, CH₄, N₂O) on the ability to detect and retrieve the abundances of anthropogenically significant trace constituents; the second concerns the effect of the thermal structure of the atmosphere on the appearance (or nonappearance) of features of candidate species (for example, CH₄) in emission spectra. Broadly speaking, the result of these considerations is that, for investigation of the detailed composition of the atmosphere, absorption measurements win hands down whereas, for spatial mapping of the more abundant species, measurements in the emission mode are more efficient.

However, in the case of observations of an unknown atmosphere, made from distances very large compared to the separation between the planet and its parent star, these factors are weighted quite differently. The competition may depend on the outcome of the technological challenge of making ultra high-precision amplitude measurements in the absorption case vs achieving spectrally resolved discrimination of a very weak source in the other. Finally, an opinion will be offered on the required spectral resolution and S/N ratio, in the emission and absorption modes, for the identification of CO₂, H₂O, O₂, O₃, and CH₄ in an unknown planetary atmosphere at 10 ly distance.