StarXiv: a podcast discussing the latest astronomy papers
Jun 23, 2026 · 39 min · 8 segments
In this episode, Payel and Nicole discuss machine learning methods to to transform composite galaxy spectra and determine key features of galaxy evolution. They also examine new observations of the…
Payel DasHostNo entities detected.
So yeah, my two papers kind of are in a very kind of similar group, but I wo- I won't do any spoilers, but my first paper is called "A Novel Data-Driven Approach to Extract Stellar Population Properties from Galaxy Spectra Using Absorption Indices." So [laughs] it's, it's a very as-it-says title.
And the author for this is Zahra Shaabaf et al., from the Institute of Space Sciences in Spain, and this was submitted to Monthly Notices.
And so yeah, so as the, uh, title implies, you know, we're looking at galaxy spectra and what we can learn from it.
If we're looking at a whole spectrum, we want to know the history of a galaxy.
You kind of want to know how old its stars are, how metal-rich it is, and these properties are imprinted in the spectra, but they are very well hidden.
And so there have been all sorts of techniques to try and unearth all of these features, and one of the classic tools that you use is called the Lick Index system, which was something that I had to search up before because it's not really something that I'd encountered, I guess l- looking at...
Well, I guess stellar spectra and galaxy spectra fall into the same-
Yeah
Exactly.
But by using this system, instead of using the whole spectrum, you can measure the strength of kind of a, a handful of specific absorption features, and each are kind of sensitive to either, you know, age, metallicity, whatever it is you're interested in.
And then the idea is, is you compare those measurements against grids of stellar population models, and by doing that, you're able to read off the properties of the, of the galaxy.
But there are some problems with this.
So you've got the age-metallicity degeneracy.
So if you have an old metal-poor population and a younger metal-rich one, they can produce almost the same spectrum, and so there's a lot of degeneracy there, and there's a lot of tangling that people try to kind of pull apart.
And one of the ways that people try to do this is by using multivariate statistics, so looking at how your measurements, your different absorption features vary with one another to try and separate the underlying physical drivers.
And so that is exactly what this paper does, uh, but it uses a method, and Payel, you'll be familiar with this.
It's principal component analysis.
Yes.
[laughs] Um-
So yeah, my two papers kind of are in a very kind of similar group, but I wo- I won't do any spoilers, but my first paper is called "A Novel Data-Driven Approach to Extract Stellar Population Properties from Galaxy Spectra Using Absorption Indices." So [laughs] it's, it's a very as-it-says title.
And the author for this is Zahra Shaabaf et al., from the Institute of Space Sciences in Spain, and this was submitted to Monthly Notices.
And so yeah, so as the, uh, title implies, you know, we're looking at galaxy spectra and what we can learn from it.
If we're looking at a whole spectrum, we want to know the history of a galaxy.
You kind of want to know how old its stars are, how metal-rich it is, and these properties are imprinted in the spectra, but they are very well hidden.
And so there have been all sorts of techniques to try and unearth all of these features, and one of the classic tools that you use is called the Lick Index system, which was something that I had to search up before because it's not really something that I'd encountered, I guess l- looking at...
Well, I guess stellar spectra and galaxy spectra fall into the same-
Yeah
Exactly.
But by using this system, instead of using the whole spectrum, you can measure the strength of kind of a, a handful of specific absorption features, and each are kind of sensitive to either, you know, age, metallicity, whatever it is you're interested in.
And then the idea is, is you compare those measurements against grids of stellar population models, and by doing that, you're able to read off the properties of the, of the galaxy.
But there are some problems with this.
So you've got the age-metallicity degeneracy.
So if you have an old metal-poor population and a younger metal-rich one, they can produce almost the same spectrum, and so there's a lot of degeneracy there, and there's a lot of tangling that people try to kind of pull apart.
And one of the ways that people try to do this is by using multivariate statistics, so looking at how your measurements, your different absorption features vary with one another to try and separate the underlying physical drivers.
And so that is exactly what this paper does, uh, but it uses a method, and Payel, you'll be familiar with this.
It's principal component analysis.
Yes.
[laughs] Um-
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.
