APOD: The Gaia Stars of M15 (2019 Mar 28)

[APOD Robot]

The Gaia Stars of M15

Explanation: Messier 15 is a 13 billion year old relic of the early formative years of our galaxy, one of about 170 globular star clusters that still roam the halo of the Milky Way. About 200 light-years in diameter, it lies about 35,000 light years away toward the constellation Pegasus. But this realistic looking view of the ancient globular star cluster is not a photograph. Instead it's an animated gif image constructed from remarkably precise individual measurements of star positions, brightness, and color. The astronomically rich data set used was made by the sky-scanning Gaia satellite which also determined parallax distances for 1.3 billion Milky Way stars. In the animated gif, twinkling stars are M15's identified RR Lyrae stars. Plentiful in M15, RR Lyrae stars are evolved pulsating variable stars whose brightness and pulsation period, typically less than a day, are related.

<< Previous APOD

This Day in APOD

Next APOD >>

[Ann]

This is a fantastic APOD!

The twinkling stars are RR Lyrae stars. They are, as the APOD caption said, evolved stars that have left the main sequence. An important point is that the RR Lyrae stars are very metal-poor.

RR Lyrae stars are found only in such globular clusters that contain a blue horizontal branch. Globulars that only have a short red horizontal branch, like 47 Tuc, lack RR Lyrae variables. See this post:

http://asterisk.apod.com/viewtopic.php?p=290804#p290804

Click to view full size image

Color-magnitude diagram of a globular cluster.
Note the "gap" in the horizontal branch. This is where
you find the RR Lyrae stars. Source:
http://burro.cwru.edu/academics/Astr221 ... sters.html

RR Lyrae stars typically belong to spectral classes F or late A. They are therefore not the bluest stars in globular cluster, but they typically only appear in globulars with a blue horizontal branch. In the picture at left, note the "gap" in the horizontal branch where the white stars are located. This is where you find the RR Lyrae variables.

In today's APOD, all the orange stars are shining with a steady, unmoving light, and so are the bluest stars. The white or blue-white RR Lyrae stars are the ones that vary.

Ann

[simonmago]

This is a brilliant post, thank you for sharing. I found a small typo, so here is my contribution: the prominent property of RR Lyrae is the relation between period and luminosity - not brightness. Brightness and luminosity are related by distance, so the importance of RR Lyrae is that we measure a physical property (period) and a distance-dependent property (brightness). Given that period also gives us the luminosity (through the relation), we can use luminosity and brightness to calculate the distance.

[Boomer12k]

Terrific way to show a different type of star...


:---[===] *

[Alex_g444]

I'm wondering how it looks like when you are at, say around 1AU, from such a pulsating star.

[orin stepanek]

Could planets be plentiful in globular clusters?

[bls0326]

Clicked on the image to get the larger version. Wow! Quite a few twinkling stars are easily seen. I was wondering how many actual stars are represented in the gif. M15 has over 100,000 stars according to one of the links. The gif has about 7 million pixels so a large number of actual stars could be (and probably are) represented.

[pferkul]

Why do the pulsations appear to be somewhat in phase? Here is a plot of average image brightness for each of the 10 frames.

[BDanielMayfield]

Could planets be plentiful in globular clusters?

No, not plentiful, it's thought, for at least two reasons. (1) Not much planet forming material (metals) were around back in the early days of the universe when globular cluster stars were forming. Gas giants may have formed, but (2) The stars in a globular will tend to loose its planets over time as they are ejected due to close encounters between star systems.

Bruce

[neufer]

Click to play embedded YouTube video.

Why do the pulsations appear to be somewhat in phase?

<<Phase synchronization is the process by which two or more cyclic signals tend to oscillate with a repeating sequence of relative phase angles. One example of phase synchronization of multiple oscillators can be seen in the behavior of Southeast Asian fireflies. At dusk, the flies begin to flash periodically with random phases and a gaussian distribution of native frequencies. As night falls, the flies, sensitive to one another's behavior, begin to synchronize their flashing. After some time all the fireflies within a given tree (or even larger area) will begin to flash simultaneously in a burst.>>

[Ann]

Could planets be plentiful in globular clusters?

Yes and no. I know I have read that Earth-sized planets can form quite easily around metal-poor stars, but gas giants are thought to be rare in them.

The problem is that the planets in globular clusters are subjected to strong gravitational forces all the time, which tend to either cast them out of their solar systems altogether, or at least fling them into very elongated orbits - or send them plunging into their suns.

Globular clusters are probably not very good incubators for advanced life forms.

Ann

[Cousin Ricky]

Why do the pulsations appear to be somewhat in phase? Here is a plot of average image brightness for each of the 10 frames.

Because it’s an animation reconstructed from data, rather than an actual video?

[neufer]

Globular clusters are probably not very good incubators for advanced life forms.

• Which is actually good since truly advanced life forms would have probably already figured out ways to block robocalls:

<<The Arecibo message is a 1974 interstellar radio message carrying basic information about humanity and Earth sent to globular star cluster M13 in the hope that extraterrestrial intelligence might receive and decipher it. The message was aimed at the current location of M13 some 25,000 light years away because M13 was a large and close collection of stars that was available in the sky during a ceremony to mark the remodeling of the Arecibo radio telescope in Puerto Rico on 16 November 1974. The message consisted of 1,679 binary digits, approximately 210 bytes, transmitted at a frequency of 2,380 MHz and modulated by shifting the frequency by 10 Hz, with a power of 450 kW. The "ones" and "zeros" were transmitted by frequency shifting at the rate of 10 bits per second. The total broadcast was less than three minutes.>>

[BDanielMayfield]

Why do the pulsations appear to be somewhat in phase? Here is a plot of average image brightness for each of the 10 frames.

Nice data analysis pferkul. There is no way such a smooth brightness curve could be natural. These variable stars wouldn't be in phase with each other, their brightness peaks would be randomly distributed, and their periods, quoting Wikipedia on RR Lyrae stars, are "typically less than one day, sometimes ranging down to seven hours."

So the creators of this simulation must have applied a simple sine function with the same period for each of these variable stars.

Bruce

[Cousin Ricky]

Why do the pulsations appear to be somewhat in phase? Here is a plot of average image brightness for each of the 10 frames.

Nice data analysis pferkul. There is no way such a smooth brightness curve could be natural. These variable stars wouldn't be in phase with each other, their brightness peaks would be randomly distributed, and their periods, quoting Wikipedia on RR Lyrae stars, are "typically less than one day, sometimes ranging down to seven hours."

So the creators of this simulation must have applied a simple sine function with the same period for each of these variable stars.

This was plainly obvious to me just by looking at the image.

[Chris Peterson]

Why do the pulsations appear to be somewhat in phase? Here is a plot of average image brightness for each of the 10 frames.

Nice data analysis pferkul. There is no way such a smooth brightness curve could be natural. These variable stars wouldn't be in phase with each other, their brightness peaks would be randomly distributed, and their periods, quoting Wikipedia on RR Lyrae stars, are "typically less than one day, sometimes ranging down to seven hours."

So the creators of this simulation must have applied a simple sine function with the same period for each of these variable stars.

This was plainly obvious to me just by looking at the image.

I assumed just from the caption that the twinkling was applied as a visualization tool unrelated to the actual periodicity of the individual stars, even before I saw the image.

[Nitpicker]

I would assume that Gaia recorded enough samples from each RR Lyrae star, to determine the average period of each. So, I suppose they could have produced a more realistic gif, but for what purpose, I don't know. An ordinary luminosity vs time graph, for each variable star, would convey that more clearly to most people. But the APOD is prettier.

[Cousin Ricky]

I would assume that Gaia recorded enough samples from each RR Lyrae star, to determine the average period of each. So, I suppose they could have produced a more realistic gif, but for what purpose, I don't know. An ordinary luminosity vs time graph, for each variable star, would convey that more clearly to most people. But the APOD is prettier.

You’re talking about synchronizing dozens of different periods such that they can repeat in a loop. That would require an impressive least common multiple, and a truly massive GIF. An idea that wouldn’t break the memory bank would be to stagger the phases while keeping the same period for each star. It wouldn’t be more realistic, but it would look more realistic.

[Chris Peterson]

I would assume that Gaia recorded enough samples from each RR Lyrae star, to determine the average period of each. So, I suppose they could have produced a more realistic gif, but for what purpose, I don't know. An ordinary luminosity vs time graph, for each variable star, would convey that more clearly to most people. But the APOD is prettier.

You’re talking about synchronizing dozens of different periods such that they can repeat in a loop. That would require an impressive least common multiple, and a truly massive GIF. An idea that wouldn’t break the memory bank would be to stagger the phases while keeping the same period for each star. It wouldn’t be more realistic, but it would look more realistic.

It wouldn't be at all impractical to create an animation with the actual periods and phases, possibly even with accurate light curves (which are not sinusoidal, and do not all have the same shape). But an animated GIF would not be the right format for such an animation.

[geckzilla]

just speaking from experience trying to make these... if you only have 2 or 3 frames originally and you add frames between to make a longer, smoother animation (known as tweening), it makes them look like they're all in sync.

[Nitpicker]

I don't think it would require a much bigger animation to give a good rough impression of all the periods.

[Chris Peterson]

I don't think it would require a much bigger animation to give a good rough impression of all the periods.

The problem is doing it as a GIF. Not very efficient for that type of animation.

[Nitpicker]

I don't think it would require a much bigger animation to give a good rough impression of all the periods.

The problem is doing it as a GIF. Not very efficient for that type of animation.

I should have said I don't think it would require many more frames, regardless of format.

But I don't think it would be any better, practically, at conveying the real data, nor would it make it look more like a real image of this GC.

[Tara_Li]

Actually, the GIF wouldn't need to be that much larger, really. Remember that most of the image data is identical between frames. So the only areas that need to vary between frames would be the variables - which are a very small segment of the image. Using difference overlaying, and limiting the area to be overlayed to just the section with variables in it, you wouldn't add a whole lot to add each additional frame.\

I came in, at first, to see what time acceleration this was supposed to represent, but ... well, as it's apparently completely synthetic, for the most part, that's a meaningless question.

[paulhummerman]

What's the time-scale of this animation? Visible wavelengths?