Multivariante analyses of multicolour photometry for mode identification
Margit Paparó and Lajos G. Balázs
Konkoly Observatory, Budapest, Hungary
A generally accepted view that for mode identification of stars we need
longer and longer data sets with higher and higher precision for reaching
a similar systematic arrangement of modes that we have for the Sun (small
and large separations). Hopefully the space mission's data will fulfill these
criteria but the fact still exists that the stars are much further than
our Sun. The precision maybe not as high as we wish and the resolution is not
two-dimensional.
Preparing ourself for the space mission's era an independent line of
observational investigation is suggested that is not supported by any
theoretical prediction at this moment (non-radial, non-linear models are
missing). If some regularities are searched and found among the modes
with largest amplitudes, the critical point of mode identification for most
of the astroseismologically interesting targets is getting to be more
treatable.
Tucanae is one of the best studied Scuti star based on
Stromgren multicolour photometry. For the excited modes with largest amplitudes
the richness of observational parameters (frequencies, amplitudes and phases)
is at our disposal. The regularities of excited modes of Tucanae
(frequency spacing, amplitude and phase behaviour) has been checked on
two-dimensional planes (Paparó & Sterken, AA, 362. pp. 245).
A result of multivariante analyses of multicolour photometry of
Tucanae is presented. Multivariate clustering methods are used to find
structures. Hierarchical clustering was performed for finding similarities,
assuming squared Euclidean distances among cases (frequencies) and variables.
Several agglomeration methods of SPSS statistical software package were used
for getting groups or any other definite structures in n-dimensional space.
Print this abstract
Last changed: 2002/Apr/26
©1996-2002 CAUP
|