Some Thoughts on the
Analysis and Interpretation of Lightcurves - Part Two - Periodicities
Lee Anne Willson
Erwin W. Fick Observatory,
Iowa State University, Ames, IA 50010
Part I of this series
showed the consequences of the fact that magnitudes are logarithmic quantities.
In this contribution the question of variable star periods is discussed, in
particular the cases where multiple periods and /or period changes are present.
The O-C diagram can be a
powerful tool for determining periods, period changes, and even the presence of
multiple periods. Some superficially similar light curves can easily be
distinguised by comparison of the O-C diagrams: examples include "AM"
vs. "FM" modulation.
Fourier analysis is the
classic method for establishing the presence of periodicities. This method is
powerful, but has the drawback taht if the light curv eis not strictly
sinusoidal, the Fourier results will include extra periods to account for the
shape of magnitude variation or the flux variation is analyzed.
Recently a number of
relatively straightforward techniques for determining the periods present in a
varying source have become popular among astronomers. These techniques include
"phase dispersion minimization" and "auto-correlation." The
basic assumptions of these techniques are described, examples, shown, and
further references given.