[vsnet-campaign-be 123] Re: Delta Sco: Provisional Spectroscopic Period and Orbit

[Thom Gandet <tgandet@mindspring.com>]

Sebastian et al,

   A couple of interesting points:

      1)  There is a pronounced period of between 68 and 80 days in the
light curve of the visual/PEP observations shown in the second figure
of IBVS 5026.at about JD 2448500 and ending with an apparent minimum on
about JD 2448825.  Prior to the start of that drop, the data strongly
suggest another local minimum around JD 2448200, almost exactly 8 
cycles of the 78.86-day spectroscopic period before the possible
minimum on 2448825.  From the spectroscopic ephemeris I gave, Min. I
should have occurred on JD 2448825 and on JD 2448194.

      3)  If the 68-80 day cyclic behaviour shown in the second
figure in IBVS 5026 is apparent in the recent photometry, and if each
successive max and min is slightly brighter than the preceding one,
then the luminosity is increasing slowly with time.  At the same
time, a periodicity in the light variation is emerging that hasn't
been seen before.  (Your Figure 1 also suggests that a lomg-period 
variation in magnitude has occurred before.)

   The predicted times of maximum and minimum that the spectroscopic
"orbit" yields are not necessarily times of *eclipses*, but they
do reproduce the recently observed times of minimum very well.
(They are offset from the actual observed light minima by about
90in phase, which is a characteristic of Beta CMa stars.)
The Delta Sco light variations are roughly sinusoidal in shape,
which is also a Beta CMa-like characteristic.

   An analogue to Delta Sco might be EN Lac: Several, very short,
pulsation periodicities are tangled up with long-period varitations
of from 331-days to 74-years.  EN Lac is also an eclipsing binary
with line-profile variations, and the radial velocitie curve shows
a lot of scatter close to periastron passage in the eclipsing binary
orbit.  So there's a lot to untangle!  But it can be done - see
Lehmann, Harmanec, et al (2001), A&A 367, pp 236-249.

   The provisional spectroscopic "orbital elements" merely describe
in a convenient way, for now, the periodic nature of the velocities.
They also allow for some testable predictions to be made.  The fact
that the value for a sin i places the secondary *inside* the
primary pretty much quashes the interpretation of the gross radial
velocity variations as being due to binarity.

   While there are many elements of Delta Sco's  behaviour that
are indicative of Beta CMa-type variability, the projected
rotational velocity of Delta Sco is 175 km/sec, but the maximum
rotational velocity observed for Beta CMa stars is about 40 km/sec.
Also, there are only a couple of other Beta CMa stars (that I can
recall) which show any kind of emission lines in their spectra, and
then only weakly (and not of the hydrogen lines).

   It is too early to say what is causing the periodic behaviour in
the Delta Sco velocities, or why they have persisted for over ninety
years.  I believe they are very complex, much like those of EN Lac,
but that they can be untangled.  There are several possibilities
involving one or a combination  of causes, and the radial velocities
do not entirely rule much shorter, Beta CMa-like variability
underlying the data.  This witness deponeth not further - I am
still committing crimes against the radial velocity data! <G>

   One final note: Periastron passage in the long-period 
(speckle) orbit is rapidly approaching.  None of the published
radial velocity data covers the rapid descending branch of the
velocity curve near periastron passage.  Radial velocities,
line profiles, and photometry are badly needed during periastron
passage.


Regards,
Thom Gandet

 
> Hi, Thom:
>                  No trace of eclipses can be found in the PEP and visual
> recent photometry, and most important in the three year Hipparcos data.
> Although the lightcurve may have been affected by the eruption, certainly
> the eclipses wouldn't have appeared or disappeared because of this.
>                  However, maybe there IS an eclipse but not of a star but
> the ejected ring. Maybe that's the reason of the cyclic variations seen. The
> same goes for the BCEP possibility: how could the amplitude be increased
> that much? And, of course, a 78 days cycle has nothing to do with the less
> than a day periods of those stars. So the star is a BCEP variable but the 75
> or 78 days cycle needs to be explained in other way.
> 
> Cheers,
> Sebastian.
> 
> ---
> Outgoing mail is certified Virus Free.
> Checked by AVG anti-virus system (http://vsnet.grisoft.com).
> Version: 6.0.249 / Virus Database: 122 - Release Date: 13/04/01

-- 
************************************************************************
                      Lizard Hollow Observatory
                      Thomas L. Gandet, Director
                            PO Box 77021                   
                      Tucson, AZ 85703-7021  USA
************************************************************************

VSNET Home Page

Return to Daisaku Nogami

---

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