[vsnet-chat 3853] re wavelet analysis software

[crawl@zoom.co.uk]

Peter F Williams

>The prospect of mode switching did come to mind and is mentioned in the
>paper soon to be submitted to the VSS RASNZ, though I do not pretend to
>understand such things.  Tim Bedding of the University of Sydney has an
>interest in stars which show similar behaviour, such as R Doradus (Bedding,
>T.R. et al.: 1998, MNRAS, 301, 1073).  A copy of my data was sent to him
>about two weeks ago for running through his wavelet software but I have not
>heard back yet.  Tim did mention the need for more frequent observations so
>that may answer the question.

A DOS executable enabling wavelet analysis to be performed on data
consisting of 4096 obs or less used to be available at the aavso website,
and probably still is, but I have difficulty navigating it nowadays [it was
a _long_ time ago that I downloaded it].  It's authored by Grant Foster.
The real pig with wavelet analysis mindst is mapping the results,
especially the contour maps, and you need these to look for mode switching.

The above R Dor paper is just about the only one that shows reasonable
evidence for mode switching in the entire literature [I've looked].  I've
never been able to get hold of R Dor material to check it mindst.  With
wavelet analysis, whether or not two signals are contemporaneous or
alternating is difficult to ascertain because it all depends on how "hard"
you look.  This is a technical consequence of what "resolution" of
frequency you're using, how you consider semi-amplitudes, or weighted Z
transforms in Foster's preferred route, to be significant [both these are
markedly non-linear] re "cut off" points, and what have you.

Again, in the above R Dor paper V Boo is also noted as possibly being a
mode switcher as a kind of second opinion, it being cited as a similar-ish
star to R Dor.  V Boo ain't a mode switcher, but if you don't look deep
enough, you miss the shorter period sometimes, cos it is of low amplitude.
Being an amateur I don't always put my point across properly, but I know
full well that professionals are not even remotely interested in refutation
of mode switching occurences.  They don't get to play the game about stars
pulsating simultaneously in 1st and 3rd, or 3rd and 5th, or whatever
overtones so easily without evidence of mode switching.  Since some MACHO
work on LMC LPVs and possible P-L diagrams therefore, this [mode switching]
is very popular.  There is no coincidence that Dr Beddings has shown some
interest in using visual data to find particular cases of Galactic mode
switching LPVs, though the original work on R Dor stems from the fact that
he was part of a team at ESO that interferometrically resolved this star,
which is the _apparently_ largest star in the heavens, I believe [okay, the
sun is, but you know what I mean ;^) ].  Arguments re pulsation mode have
been following two camps, best championed in the past by Feast on one hand
and Wood on the other if I remember rightly [more southern hemisphere
chaps] since the '60s... ...with claims of the definitive proof as to which
coming out every few years [and still continuing].  The thing is, the
theories and maths are all neat and tidy and more or less sown up, it's
just that they don't really describe what LPVs really do, but they might
seem to do that bit better if mode switching really did exist.

Similarly to V Boo, the first paper on mode switching, almost the classic
one, is by Cadmus et al, 1991, where U Boo is suggested to mode switch.
80+ years of BAA data [Howarth & Greaves, JBAA, eventually] shows that this
is not the case, again the shorter period is there all along, though damn
weak I'll admit, such that it is only easy to find when the main period
loses a bit of power on very rare occasions!  The 2:1 period ratio doesn't
help here.

I suppose the best evidence of mode switching is a paper on W Boo by John
Percy, where pep data is used.  However, this is a very low amplitude
shortish period semiregular, and the data run was for only a few years with
a heckmondwike lot of gaps, so whether it is directly analogous to the
normal run of LPVs is summat else.

endtirade

Cheers

John

VSNET Home Page

Return to Daisaku Nogami

---

vsnet-adm@kusastro.kyoto-u.ac.jp