A pre-processing module is systematically performed before CO and CH₄ retrieval. This pre-processing module exploits spectral information from the SWIR spectral ranges to compute Lambert-Equivalent Reflectivity (LER) filter and three different cloud filters.

• LER filter: measurements (ground pixels) with a continuum LER value below a certain threshold are discarded (for both CO and CH4 retrieval). For example, clear-sky ocean measurements are rejected by this inquiry due to the very low ocean surface reflection in the SWIR spectral range.
• A priori CH₄ cloud filter: a comparison of a non-scattering methane column retrieval with a model a priori methane estimate is used to reject spectral observations contaminated by high optically thick clouds. The extent of light path shortening and enhancement due to atmospheric scattering by clouds and aerosols can be indicated by comparing the retrieved CH₄ column with the a priori knowledge. If the difference ?CH₄ exceeds a certain threshold, observations are strongly contaminated by clouds. This cloud filter is required for the SICOR CO retrieval, which is designed to process clear sky (over land only) SWIR observation and measurements of scenes with low water clouds and optically thin cirrus and aerosol loads.
• Two-band CH₄ cloud filter: This filter is based on the comparison of two CH₄ non-scattering retrievals that use two different spectral bands – a strong and a weak absorption band respectively. The difference between the two retrieved columns indicates changes of the light path due to atmospheric scattering and so can be used as a cloud filter. This filter is particularly valuable for cirrus screening needed by the CH₄ data processing.
• Two-band H₂O cloud filter: This filter uses the same concept as the two-band CH₄ cloud filter but with two H₂O absorption bands. This filter can be used as a complement to the other cloud filters.