A match of ROLO data to an old formula had been achieved by adding some constants: -0.95 to the old formula I use, 0.05 to the blue filter, -0.11 to the red one (otherwise the blue luminance would be the lowest, the red one the largest one). The vertical axis is green reflectance as defined for the whole Moon by the ROLO team. This way the old curve I use and the ROLO one best agree around full moon, apart from the very opposition, where the differences climb up to 0.11 in log10 (0.28 mag).
At large phase angles the differences are up to 0.03 in log10 (0.075 mag, or 7 %). The old curve should become the ROLO one for phase angles below 123 degrees.
ROLO seems to give the waning Moon brighter when its lower left edge is toward us (max. negative subterrestrial long. and lat., exposing most of non-maria surface there). The span is up to 0.07 mag.
Data are taken from
Kieffer, H.H., Stone, T.M.: The spectral irradiance of the Moon, Astronom. J, 129 2887-2901, June 2005 ( see it at www.journals.uchicago.edu). (Numeric values of wavelength-dependent coefficients have been kindly sent by Tom Stone.)
jh, Jan 2007