How to Measure the Moon and Calibrate using it
In 2002, CERES instruments began lunar raster scans near full moon, as in the animation. In 2007 it was discovered that the raw detector signal in the top plot could be integrated in angular space, to give a measure of the lunar disk averaged sunlight reflected, when divided by the Moon's angular extent. (i.e. found result is the known lunar radiance, multiplied with unknown instrument detector gain)
The short animation above shows a raster scan of the Moon and raw detector signal used for lunar calibration
Top: MERBE measurements of lunar albedo at different phases compared to the ROLO lunar model. Bottom: MERBE lunar albedo normalized to +7 degrees phase using ROLO model.
Lunar percent lunar reflectivity’s were hence recorded initially up to 2015 as shown by the top graph.
Specifically, such bottom graph measurements were defined at +7 degree static phase, normalization to which used the RObotic Lunar Observatory or ROLO model shown in comparison in the top left plot, Essentially the bottom graph is the top blue diamond values, divided by those of the black ROLO dots. This works because the ROLO model, although not absolutely accurate to better than 1%, is stable to 0.1%/decade. The Moon reflects solar UV not emitted by CERES onboard lamps, meaning false calibration UV drifts can be removed from Earth MERBE data.