To calibrate measurements from visible/NIR/SWIR channels, a gain parameter is derived from the pixel counts over the visible calibration target (the VISCAL unit) during the period when it is illuminated by the sun and the illumination level is constant – this occurs once per orbit.
The illumination period is determined according to the calibration window and the average and standard deviation of all valid pixels from the VISCAL unit for all scans during the calibration period are calculated. A similar mean and standard deviation are calculated for all valid pixels from a selected black body over the full orbit, as for the infra-red calibration.
However, as the Level-0 SLSTR input data are granule based, they can cover a duration significantly smaller than a full orbit. As a consequence, most of the SLSTR Level-0 granules provided to the SLSTR Level-1 processor do not contain any data resulting from an illumination of the SLSTR sensor by the sun using the visible calibration (VISCAL) mechanism.
To avoid having the SLSTR Level-1 operational processor in error during VISCAL calibration, a specific processing mode has been defined for creation of the calibration files used in calibration of the visible/NIR/SWIR channels of the next orbit.
This processing is divided into several sub-steps, including product formatting processing. The sub-steps are copied from the "Calibration Coefficients Computation for Visible/NIR/SWIR Channels" processing step, included in the Level-1A observation mode processing (see Algorithms Overview).