by Ann » Wed Jul 10, 2013 5:44 am
Chris Peterson wrote: There is little evidence of systematic errors in Hipparcos measurements
http://en.wikipedia.org/wiki/Pleiades#Distance wrote:
Measurements of the distance have elicited much controversy. Results prior to the launch of the Hipparcos satellite generally found that the Pleiades were about 135 parsecs away from Earth. Data from Hipparcos yielded a surprising result, namely a distance of only 118 parsecs by measuring the parallax of stars in the cluster—a technique that should yield the most direct and accurate results. Later work consistently argued that the Hipparcos distance measurement for the Pleiades was erroneous.[3][4][5][15][16] In particular, distances derived to the cluster via the Hubble Space Telescope and infrared color-magnitude diagram fitting favor a distance between 135–140 pc.[3][15]
I have been a little suspicious of the Hipparcos data ever since the Pleiades controversy occurred. Moreover, it seems to me that the Hipparcos measurements should be less reliable the more distant the star in question is, and the smaller the measured parallax is. A very small adjustment in a very small parallax leads to a huge adjustment in the star's distance and luminosity. And since Gamma Cygni is undoubtedly more distant than the Pleiades, the Hipparcos data should be less reliable for Gamma Cygni than for the Pleiades. And they have been "iffy" for the Pleiades.
Please understand that I don't propose to know better than you do here, Chris, but I wanted to voice the reason for my lack of absolute confidence in the result of tiny Hipparcos parallaxes.
On the other hand, I note that some of the more controversial Hipparcos parallaxes have been adjusted so that they are a bit more in accordance with the distances that have been established, or guessed at, by other means. I believe, although I'm not absolutely certain, that the overall "Hipparcos distance" to the Pleiades has been somewhat expanded. I know for a fact that the "Hipparcos distance" to Deneb, Alpha Cygni, has been slashed. The original Deneb parallax, which was used by my Guide software, was just over one milliarcsecond. The revised parallax is instead 2.31 ± 0.32 milliarcseconds. The luminosity of Deneb, based on the original, tiny parallax, was about 200,000 solar luminosities. Based on the revised parallax, its luminosity has been slashed to about 50,000 Suns. This is in much better accordance with other estimates of Deneb's brilliance than the original Hipparcos parallax was.
And since the parallax of Gamma Cygni is certainly based on the revised Hipparcos catalog, I probably shouldn't be so suspicious of it.
Finally, though, I note that Alpha Cygni is classified as A2Ia and Gamma Cygni as F8Ib. The revised Hipparcos catalog makes Gamma Cygni more distant than Alpha Cygni, and the resulting luminosities are somewhat similar: 49,000 ± 14,000 Suns for Alpha Cygni and 34,000 ± 10,000 Suns for Gamma Cygni. The difference in actual luminosity between the supergiant type Ia and the supergiant type Ib is a bit smaller than I had expected.
Ann
[quote="Chris Peterson"] There is little evidence of systematic errors in Hipparcos measurements[/quote]
[quote]http://en.wikipedia.org/wiki/Pleiades#Distance wrote:
Measurements of the distance have elicited much controversy. Results prior to the launch of the Hipparcos satellite generally found that the Pleiades were about 135 parsecs away from Earth. Data from Hipparcos yielded a surprising result, namely a distance of only 118 parsecs by measuring the parallax of stars in the cluster—a technique that should yield the most direct and accurate results. Later work consistently argued that the Hipparcos distance measurement for the Pleiades was erroneous.[3][4][5][15][16] In particular, distances derived to the cluster via the Hubble Space Telescope and infrared color-magnitude diagram fitting favor a distance between 135–140 pc.[3][15] [/quote]
I have been a little suspicious of the Hipparcos data ever since the Pleiades controversy occurred. Moreover, it seems to me that the Hipparcos measurements should be less reliable the more distant the star in question is, and the smaller the measured parallax is. A very small adjustment in a very small parallax leads to a huge adjustment in the star's distance and luminosity. And since Gamma Cygni is undoubtedly more distant than the Pleiades, the Hipparcos data should be less reliable for Gamma Cygni than for the Pleiades. And they have been "iffy" for the Pleiades.
Please understand that I don't propose to know better than you do here, Chris, but I wanted to voice the reason for my lack of absolute confidence in the result of tiny Hipparcos parallaxes.
On the other hand, I note that some of the more controversial Hipparcos parallaxes have been adjusted so that they are a bit more in accordance with the distances that have been established, or guessed at, by other means. I believe, although I'm not absolutely certain, that the overall "Hipparcos distance" to the Pleiades has been somewhat expanded. I know for a fact that the "Hipparcos distance" to Deneb, Alpha Cygni, has been slashed. The original Deneb parallax, which was used by my Guide software, was just over one milliarcsecond. The revised parallax is instead 2.31 ± 0.32 milliarcseconds. The luminosity of Deneb, based on the original, tiny parallax, was about 200,000 solar luminosities. Based on the revised parallax, its luminosity has been slashed to about 50,000 Suns. This is in much better accordance with other estimates of Deneb's brilliance than the original Hipparcos parallax was.
And since the parallax of Gamma Cygni is certainly based on the revised Hipparcos catalog, I probably shouldn't be so suspicious of it.
Finally, though, I note that Alpha Cygni is classified as A2Ia and Gamma Cygni as F8Ib. The revised Hipparcos catalog makes Gamma Cygni more distant than Alpha Cygni, and the resulting luminosities are somewhat similar: 49,000 ± 14,000 Suns for Alpha Cygni and 34,000 ± 10,000 Suns for Gamma Cygni. The difference in actual luminosity between the supergiant type Ia and the supergiant type Ib is a bit smaller than I had expected.
Ann