2. Basis of the correction procedure
We make no phase shift between the curves but take the effect of the phase shift into account in a correction curve as a function of phase to be determined. It is the determination of the phase variation of these template correction curves that calibrates the method and is the subject of this note.
We postulate, later to be tested using actual data, that we can transfer the information on shape, amplitude, and variable phase shift over the cycle from the complete V light curve to the unknown light curve in B, R, or I by an average correction value C, at every phase , such that
The reason for the postulate is that we expect the individual differences (at any phase) between the observed magnitude and the mean magnitude (intensity mean converted to magnitude units) in wavelength j, relative to the same difference in the V band, will scale as the V amplitude, . If so, the C functions can be determined empirically by applying the procedure to a set of multicolor light curves such as given by Wisniewski & Johnson (1968). We have done that, and list the C functions in the next section.
Once the correction functions C are known, the required mean magnitude, say , can be estimated from any single magnitude measured at phase , by
all data read at the V phase.
Note again that each separate value gives an estimate of . If there are n measurements of at different phases, each gives an value, and the average is taken. An estimate of the error of the resulting can then be made from the rms of the average. We have listed such errors of the values measured in this way in each of our cited HST papers.
© European Southern Observatory (ESO) 1997
Online publication: June 5, 1998