The methods of the dark signal determination by direct contemporaneous measurements using a light spectrum and modelling of the dark signal based on the dark signal characterisation data were discussed. These techniques were tested with two charge-couple detectors (CCD) array spectroradiometers used in solar UVR measurements. The sensitivity of both instruments was significantly reduced when shutters were used; the measured signal varied by up to 12% depending on the orientation of the shutter. The shutters should be permanently attached to the SSR, so that the orientation cannot be changed to prevent an increase in uncertainty. The method of using blind pixels from the optically inactive part of the CCD array in a light spectrum could be used to derive the dark signal with some limitations for integration times <10 s for the QE65000. An alternative method of deriving the dark signal from light measurements using out-of-range pixels has been proved impossible due to out-of-range stray light in both instruments. The dark signal was characterised for the range of integration times and ambient temperatures of 15-35°C. Based on these data, the model of the dark signal was developed so that a single value of the dark signal can be subtracted over the whole spectral range if the instrument temperature is known.
© Crown copyright 2014.