How Far to Andromeda? The First Direct Distance Determination

For the first time, an eclipsing binary has revealed the precise distance to
the Andromeda Galaxy (M31), say astronomers in Spain, America, and Scotland.
The distance is in excellent agreement with other, less direct
determinations.

The full story is at http://KenCroswell.com/AndromedaDistance.html .

Correct email: MagnificentUniverse "at" yahoo "dot" com.

Magnificent Universe wrote:
For the first time, an eclipsing binary has revealed the precise distance to
the Andromeda Galaxy (M31), say astronomers in Spain, America, and Scotland.
The distance is in excellent agreement with other, less direct
determinations.

The full story is at http://KenCroswell.com/AndromedaDistance.html .


"By comparing the absolute and apparent magnitudes, Ribas's team concluded
the Andromeda Galaxy is 2.52 ± 0.14 million light-years from Earth. This
agrees perfectly with the Cepheid-based distance to Andromeda--2.5 million
light-years".

On Sat, 05 Nov 2005 16:58:46 GMT, "Magnificent Universe"
wrote:

For the first time, an eclipsing binary has revealed the precise distance to
the Andromeda Galaxy (M31), say astronomers in Spain, America, and Scotland.
The distance is in excellent agreement with other, less direct
determinations.

The full story is at http://KenCroswell.com/AndromedaDistance.html .

Correct email: MagnificentUniverse "at" yahoo "dot" com.

Interesting story and a great result to have another distance
measurement agree, however I don't understand what is any more
"direct" about the measurement than the Cepheid method.

Cepheid light curves and luminosities have been "calibrated" by the
Hipparcos data from nearby stars (less than 1,000 ly).

The method described in the article above is still based in
assumptions regarding the luminosity/temperature relationships
established by measurements and distributions of nearby stars. A good
assumption I agree, that stars in M31 should show the same H-R
characteristics as those in the Milky Way, but this is not any more
"direct" in my opinion.

In my view, a direct measurement would be parallax.

---
Michael McCulloch

"Michael McCulloch" wrote in message
...
On Sat, 05 Nov 2005 16:58:46 GMT, "Magnificent Universe"
wrote:

For the first time, an eclipsing binary has revealed the precise distance
to
the Andromeda Galaxy (M31), say astronomers in Spain, America, and
Scotland.
The distance is in excellent agreement with other, less direct
determinations.

The full story is at http://KenCroswell.com/AndromedaDistance.html .

Correct email: MagnificentUniverse "at" yahoo "dot" com.

Interesting story and a great result to have another distance
measurement agree, however I don't understand what is any more
"direct" about the measurement than the Cepheid method.

The new measurement depends only on observations of stars in Andromeda.

Cepheid light curves and luminosities have been "calibrated" by the
Hipparcos data from nearby stars (less than 1,000 ly).

Actually, Cepheids are calibrated by assuming a distance (of 163,000
light-years) to the Large Magellanic Cloud and observing that galaxy's
Cepheids. The LMC's Cepheid period-luminosity relation is then applied to
observations of Cepheids in other galaxies, such as Andromeda. (Except for
Polaris, all Cepheids are too far from Earth to have reliable Hipparcos
parallaxes.)

The method described in the article above is still based in
assumptions regarding the luminosity/temperature relationships
established by measurements and distributions of nearby stars.

No, it's based on physics: L (star's luminosity) is proportional to R**2
(star's radius) and T**4 (star's temperature). An eclipsing binary allows
the determination of both R and T, and thus of L. Comparing L to apparent
brightness yields the distance.

A good
assumption I agree, that stars in M31 should show the same H-R
characteristics as those in the Milky Way, but this is not any more
"direct" in my opinion.

No, the H-R diagram is not used.

In my view, a direct measurement would be parallax.

Someday!

---
Michael McCulloch

On Sat, 05 Nov 2005 23:18:57 GMT, "Magnificent Universe"
wrote:

Actually, Cepheids are calibrated by assuming a distance (of 163,000
light-years) to the Large Magellanic Cloud and observing that galaxy's
Cepheids. The LMC's Cepheid period-luminosity relation is then applied to
observations of Cepheids in other galaxies, such as Andromeda. (Except for
Polaris, all Cepheids are too far from Earth to have reliable Hipparcos
parallaxes.)

Nevertheless the Cepheid parallax effort was in good agreement with
other distance estimations including the one in question here. It is
only a matter of time before a next generation instrument will be
launched that can reach further distances via parallax.

I'm all for more distance estimation methods -- I'm only arguing that
this further validates Cepheid methods -- it doesn't supersede.

The method described in the article above is still based in
assumptions regarding the luminosity/temperature relationships
established by measurements and distributions of nearby stars.

No, it's based on physics: L (star's luminosity) is proportional to R**2
(star's radius) and T**4 (star's temperature). An eclipsing binary allows
the determination of both R and T, and thus of L. Comparing L to apparent
brightness yields the distance.

How do you remove/determine reddening or extinction in the temperature
measurement?

---
Michael McCulloch

Michael McCulloch wrote:

I'm all for more distance estimation methods -- I'm only arguing that
this further validates Cepheid methods -- it doesn't supersede.

Well, I'd put it this way: this most recent distance measurement
is more trustworthy than the Cepheid ones.

How do you remove/determine reddening or extinction in the temperature
measurement?

By taking high resolution spectra, calibrating them carefully,
and comparing them to models of stellar atmospheres. Yes, any time
one brings models into play, one adds uncertainty. But in my
experience,
stellar models can do a very good job on the relatively simple
atmospheres of very hot stars (which these are). I'd guess that
the uncertainty in the models is at the 2-3 percent level in places.

Michael Richmond