Delta Trianguli

[bdm]

cel://Follow/DEL Tri/2008-07-19T17:22:51.33359?x=AAD0+kP0MpGXE/o&y=AACwuM23JzU0FXk&z=AAAQvIZAy5ZO2RH//////w&ow=0.789967&ox=-0.344092&oy=0.237083&oz=0.448715&select=DEL Tri&fov=32.334709&ts=1.000000&ltd=0&p=0&rf=300975&lm=10&ver=2

In the above URL (copy and paste for best results), the star Del Tri B flickers on and off. To see it best, speed up the simulation to 100,000 times.

(Also, the smaller star has no spectral type defined, and the masses of the components appear out of whack given that the B component is probably less massive due to its spectral type (tentatively K4?). That's understandable given that the spectroscopic components of this system have not been measured accurately.)

[t00fri]

cel://Follow/DEL Tri/2008-07-19T17:22:51.33359?x=AAD0+kP0MpGXE/o&y=AACwuM23JzU0FXk&z=AAAQvIZAy5ZO2RH//////w&ow=0.789967&ox=-0.344092&oy=0.237083&oz=0.448715&select=DEL Tri&fov=32.334709&ts=1.000000&ltd=0&p=0&rf=300975&lm=10&ver=2

In the above URL (copy and paste for best results), the star Del Tri B
flickers on and off. To see it best, speed up the simulation to 100,000
times.

(Also, the smaller star has no spectral type defined, and the masses of
the components appear out of whack given that the B component is
probably less massive due to its spectral type (tentatively K4?). That's
understandable given that the spectroscopic components of this
system have not been measured accurately.)

If both the luminosity class of the B component and it's apparent
magnitude were well known, these data would have been entered in the
spectbins database. However, D. Poubaix has preferred NOT to quote these
values (despite possible speculations) in his publication (AppMag=5.0
is a default dummy value).

Yet, the orbit parameters of A and B including their mass ratio should
be fine.

If you know of solid new /published/ measurements, please let me
know.

F.

[bdm]

cel://Follow/DEL Tri/2008-07-19T17:22:51.33359?x=AAD0+kP0MpGXE/o&y=AACwuM23JzU0FXk&z=AAAQvIZAy5ZO2RH//////w&ow=0.789967&ox=-0.344092&oy=0.237083&oz=0.448715&select=DEL Tri&fov=32.334709&ts=1.000000&ltd=0&p=0&rf=300975&lm=10&ver=2

In the above URL (copy and paste for best results), the star Del Tri B
flickers on and off. To see it best, speed up the simulation to 100,000
times.

(Also, the smaller star has no spectral type defined, and the masses of
the components appear out of whack given that the B component is
probably less massive due to its spectral type (tentatively K4?). That's
understandable given that the spectroscopic components of this
system have not been measured accurately.)

If both the luminosity class of the B component and it's apparent
magnitude were well known, these data would have been entered in the
spectbins database. However, D. Poubaix has preferred NOT to quote these
values (despite possible speculations) in his publication (AppMag=5.0
is a default dummy value).

Yet, the orbit parameters of A and B including their mass ratio should
be fine.

If you know of solid new /published/ measurements, please let me
know.

F.

I mentioned the lack of detail for the star because it may have been pertinent to the flickering issue. The other star has complete information, and this star does not flicker on and off in this way.

If I do find more accurate information from published papers, I will surely mention it, but I'm pretty sure that someone will beat me to it.

[bdm]

In the file spectbins.stc, the Del Tri barycenter has an incorrect distance:

Barycenter 10644 "DEL Tri"
{
RA 34.260074
Dec 34.224831
Distance 23.982609
}

In the NSTARS database, this system has the following distance:
10.8213 pc x 3.26158 = 35.2945 ly

Solstation reports 35.4 ly

[t00fri]

In the file spectbins.stc, the Del Tri barycenter has an incorrect distance:

Barycenter 10644 "DEL Tri"
{
RA 34.260074
Dec 34.224831
Distance 23.982609
}

In the NSTARS database, this system has the following distance:
10.8213 pc x 3.26158 = 35.2945 ly

Solstation reports 35.4 ly

Thanks for noting the discrepancy, which was known to
me, since my PERL script checks all coordinates and
distances against the HIP catalog and revisions thereof.

Code: Select all

Distance mismatch of 47.5070217038916 % with (revised) stars.txt for HIP 10644
Distance mismatch of 15.5672815750115 % with (revised) stars.txt for HIP 14328
Distance mismatch of 36.5868384870275 % with (revised) stars.txt for HIP 91636
Distance mismatch of 23.2407717835893 % with (revised) stars.txt for HIP 99473
Distance mismatch of 18.4920819544622 % with (revised) stars.txt for HIP 108917


Unfortunatley, it does not help me, since it does not
prove that the distance in spectbin.stc is incorrect.
The publication of Pourbaix is from 2000, AFTER the HIP
data were taken. Simbad gives the
parallax from the HIP catalog: 92 MAS, while in Pourbaix'
publication 136 MAS are used. This is what I
consistently use in in spectbin.stc, since anonymous
hand-editing does NOT make sense from a scientific
point of view. There are many incorrect values also in
the HIP catalog.

Hence I would need an independent published
measurement that one can believe and quote! and that
agrees with the HIP parallax.

So please, let me know:
++++++++++++++++++++
On what independent, PUBLISHED measurement
did you base your claim that the value in Pourbaix' paper
is wrong? Why did you believe the earlier HIP data more
for the parallax than the later, renowned catalog by
Pourbaix?
You did not quote the scientific publication on which the distance value from SolStation was based? Usually such sites just copy HIP.
++++++++++++++++++++

F.

[bdm]

In the file spectbins.stc, the Del Tri barycenter has an incorrect distance:

Barycenter 10644 "DEL Tri"
{
RA 34.260074
Dec 34.224831
Distance 23.982609
}

In the NSTARS database, this system has the following distance:
10.8213 pc x 3.26158 = 35.2945 ly

Solstation reports 35.4 ly

Thanks for noting the discrepancy, which was known to
me, since my PERL script checks all coordinates and
distances against the HIP catalog and revisions thereof.

Code: Select all

Distance mismatch of 47.5070217038916 % with (revised) stars.txt for HIP 10644
Distance mismatch of 15.5672815750115 % with (revised) stars.txt for HIP 14328
Distance mismatch of 36.5868384870275 % with (revised) stars.txt for HIP 91636
Distance mismatch of 23.2407717835893 % with (revised) stars.txt for HIP 99473
Distance mismatch of 18.4920819544622 % with (revised) stars.txt for HIP 108917


Unfortunatley, it does not help me, since it does not
prove that the distance in spectbin.stc is incorrect.
The publication of Pourbaix is from 2000, AFTER the HIP
data were taken. Simbad gives the
parallax from the HIP catalog: 92 MAS, while in Pourbaix'
publication 136 MAS are used. This is what I
consistently use in in spectbin.stc, since anonymous
hand-editing does NOT make sense from a scientific
point of view. There are many incorrect values also in
the HIP catalog.

Hence I would need an independent published
measurement that one can believe and quote! and that
agrees with the HIP parallax.

So please, let me know:
++++++++++++++++++++
On what independent, PUBLISHED measurement
did you base your claim that the value in Pourbaix' paper
is wrong? Why did you believe the earlier HIP data more
for the parallax than the later, renowned catalog by
Pourbaix?
You did not quote the scientific publication on which the distance value from SolStation was based? Usually such sites just copy HIP.
++++++++++++++++++++

F.

I checked against the NSTARS database, as I mentioned in my earlier posting. I don't know how accurate this database is. However it's worth noting that I am inclined to believe the 35 light-year figure because the primary star is a G0-star roughly as luminous as the sun, yet it has an apparent magnitude that doesn't make sense if it was actually 23 light years away. Can G0 main-sequence stars be underluminous to this degree? Of course, the spectroscopic nature of this system may interfere with this assessment ...

[t00fri]

I checked against the NSTARS database, as I mentioned in my earlier posting. I don't know how accurate this database is. However it's worth noting that I am inclined to believe the 35 light-year figure because the primary star is a G0-star roughly as luminous as the sun, yet it has an apparent magnitude that doesn't make sense if it was actually 23 light years away. Can G0 main-sequence stars be underluminous to this degree? Of course, the spectroscopic nature of this system may interfere with this assessment ...

It's not a matter of checking against another database, we need another INDEPENDENT MEASUREMENT of the distance! So, what's the published reference for that distance value in NSTARS?

Luminosity-distance relations are empirical and can have big errors! Such arguments are hardly decisisve.

F.

[bdm]

I cannot say what the source for the parallax is for the system (the layout of the NSTARS page doesn't make it clear where each specific piece of information is coming from), but here's a link to the NSTARS page for the system:
Delta Trianguli. It gives a parallax of 0.09241 +/- 0.00082 (roughly 35.29 +/- 0.32 light years).

The bibliography page is here:
NSTARS bibliography

VizieR gives a trig parallax of 92.20+/-0.84 mas (which roughly translates to 35.38 +/- 0.32 light years).

These are as close as I can get to accessing primary references (I'm not familiar with the literature), but all the secondary references that I can find give roughly the same figures.

[ajtribick]

Some detective work on Delta Trianguli:

Taking a look at the results given for this system from VizieR, we have from Hipparcos and Tycho (I am using milliarcseconds as the units here)

• Yale Trigonometric Parallaxes, Fourth Edition (van Altena+ 1995): 96.5 ?± 3.7
• The Hipparcos Main Catalogue (ESA 1997): 92.20 ?± 0.84
• The main part of Tycho Catalogue (ESA 1997): 96.90 ?± 10.10

The Spectroscopic survey in solar neighborhood (Allende Prieto+ 2004), Companions to close spectroscopic binaries (Tokovinin+, 2006) and LSPM-North proper-motion catalog nearby stars all use the Hipparcos parallax.

The All-Sky Compiled Catalog of 2.5 million stars (Khachenko, 2001) uses 92.23 ?± 0.83, which turns out to be the weighted mean of the Hipparcos and Tycho values. Same goes for the 2nd Cat. of Radial Velocities with Astrometric Data (Kharchenko+, 2007)

The NSTARS value of 92.41 ?± 0.82 appears to be the weighted mean of the Hipparcos and Yale values.

Purely for interest's sake, the weighted mean of Hipparcos, Tycho and Yale is 92.44 ?± 0.82.

Hmmm... perhaps for clarity it would be a good idea to split the Delta Trianguli discussion out of this thread?

[t00fri]

Some detective work on Delta Trianguli:

Taking a look at the results given for this system from VizieR, we have from Hipparcos and Tycho (I am using milliarcseconds as the units here)

• Yale Trigonometric Parallaxes, Fourth Edition (van Altena+ 1995): 96.5 ?± 3.7
• The Hipparcos Main Catalogue (ESA 1997): 92.20 ?± 0.84
• The main part of Tycho Catalogue (ESA 1997): 96.90 ?± 10.10

The Spectroscopic survey in solar neighborhood (Allende Prieto+ 2004), Companions to close spectroscopic binaries (Tokovinin+, 2006) and LSPM-North proper-motion catalog nearby stars all use the Hipparcos parallax.

The All-Sky Compiled Catalog of 2.5 million stars (Khachenko, 2001) uses 92.23 ?± 0.83, which turns out to be the weighted mean of the Hipparcos and Tycho values. Same goes for the 2nd Cat. of Radial Velocities with Astrometric Data (Kharchenko+, 2007)

The NSTARS value of 92.41 ?± 0.82 appears to be the weighted mean of the Hipparcos and Yale values.

Purely for interest's sake, the weighted mean of Hipparcos, Tycho and Yale is 92.44 ?± 0.82.

Hmmm... perhaps for clarity it would be a good idea to split the Delta Trianguli discussion out of this thread?

Excellent, Andrew,

that's the kind of information we need to make a sensible decision!

I would also vote for moving that discussion out of here,
since it is certainly NOT a Celestia bug issue, but rather
a delicate astrophysics/astronomy decision.

Thanks!
Fridger

[chris]

I would also vote for moving that discussion out of here,
since it is certainly NOT a Celestia bug issue, but rather
a delicate astrophysics/astronomy decision.

Agreed. Here we are in the Physics and Astronomy forum.

--Chris

[ajtribick]

The Pourbaix paper discusses delta Trianguli:

The simultaneous adjustment of the visual and spectroscopic data is a complete nightmare! In addition to the natural correlation between ?‰ and T due to the nearly circular orbit, there is another strong, ??’0.996, correlation between i and ?–. Within the confidence interval on the inclination (7?° -width), sin i ranges between 0.22 and 0.34. Therefore, a small variation on i implies a large variation on ?–. That is the reason why our results are quite imprecise (and rather unreliable): ?– = 136 ?± 30 mas, M_A = 0.25 ?± 0.16 M_??™ and M_B = 0.23 ?± 0.14 M_??™ . The parallax after Hipparcos (ESA 1997) is 92.2 ?± 0.84 mas is consistent neither with our value nor with the estimate of Van Altena et al. (1991).

So it seems that the parallax given in that paper may be quite unreliable... as has been noted by bdm it implies a suspiciously underluminous star (about half the luminosity of the Sun, despite a similar spectral type). In contrast, the Hipparcos and Yale parallaxes result in a luminosity fairly similar to that of our Sun.

Presumably the masses given in the Pourbaix paper include the sin i degeneracy, since a mass of 0.25 solar masses for a G star is ridiculous.

On the other hand, whether the Hipparcos parallax is anomalously low (perhaps due to confusion caused by the binary nature of the star?), I don't know.

[PlutonianEmpire]

What is the age of this system?