The Pristine survey XXIX. Data Release 2: Photometric metallicities for 15 millions stars down to G about 20.5

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: I'm Vera.

Jocelyn: And I'm Jocelyn, here with Subrahmanyan.

Vera: We're looking at the paper: "The Pristine survey XXIX. Data Release 2".

Jocelyn: The Pristine survey Data Release 2 presents an expanded catalogue of photometric metallicities for 15 million stars down to G ∼ 20.5,

Vera: First, who's behind it and why it matters.

Title and authors: Vera: So we’re looking at The Pristine survey XXIX Data Release two: Photometric metallicities for fifteen millions stars down to G about twenty point five today and what that means for our understanding of the universe <ref:2610.20714#pg1,The Pristine survey XXIX Data Release 2: Photometric metallicities for 15 millions>.

Jocelyn: It's a big update because this release expands the catalogue significantly, giving us deeper access to metal-poor halo stars than we had before with the first release.

Subrahmanyan: That depth is important because it lets us test our models of how structure forms in the early universe, and getting more data points at those low metallicities helps constrain those models.

Vera: Exactly. We’re talking about fifteen million stars now, reaching down to G about twenty point five, which is a very faint depth that was really pushing the limits of what we could do with these Ca H and K observations.

Jocelyn: And the coverage itself has grown quite a bit too; it covers about eight thousand square degrees now, up from the previous six thousand five hundred square degrees in DR1.

Subrahmanyan: That increase in area means we’re sampling more of the Milky Way halo population that lives in those very metal-poor environments.

Vera: Right. And they've included all Pristine observations up until January two thousand twenty-five which is an additional one thousand five hundred square degrees of coverage compared to DR1.

Jocelyn: The main thing we want the listener to grasp is that this data has good agreement with what other large surveys like DESI are finding when they look at metallicities.

Subrahmanyan: That agreement between Pristine and DESI is a crucial check, because it validates our entire method for estimating chemical compositions across different observational techniques.

Vera: And when we talk about the data reduction, they used the Cambridge Astronomy Survey Unit pipeline to process those Ca H and K images.

Jocelyn: Then they used a code called PhotCalib to calibrate the photometry on an absolute scale, using Gaia DR3 XP information to anchor it all.

Subrahmanyan: That calibration step is where you get your quantifiable uncertainty bounds, which is vital because we can’t just give a single number; we have to show the range of possibilities for each star.

Vera: They report both a median value and percentile ranges, like the 16th and 84th percentiles of the metallicity distribution function for each star.

Title and authors: Jocelyn: That’s useful because it lets users characterize that probability distribution function, showing where the bulk of the stars are clustered.

Subrahmanyan: From a theoretical standpoint, knowing that you have these distributions helps us see if our assumptions about the initial conditions of star formation match what we actually see in these metal-poor regimes.

Vera: They built their model grid using a combination of spectroscopic metallicities from SDSS/SEGUE and APOGEE, plus some high-resolution samples.

Jocelyn: To specifically boost the training set for the very low metallicity stars, those with

Fe/H: less than minus two point zero, they complemented it with stars from the Boötes I dwarf galaxy and a revised LAMOST DR3 VMP sample.

Subrahmanyan: That inclusion of stars like those from Boötes I is directly targeting the regime where structure formation physics gets really interesting, because those are often relics of the earliest building blocks of our galaxy.

Vera: They also specifically selected Pristine stars confirmed to have

Fe/H: less than minus two point zero during various spectroscopic follow-up campaigns to strengthen that low-metallicity training set.

Jocelyn: For those metal-poor stars, they recommend using the model for giants because for most of them, we can’t know if they are dwarfs or giants, and since most metal-poor stars are in the Milky Way halo, the giant model is suggested.

Subrahmanyan: That choice is a practical one; it acknowledges that in that extreme metal-poor regime, the majority of those stars reside in the more distant halo structure.

Vera: The paper also flags potential systematic biases by identifying stars that fall below the model's defined range, which suggests if a star has a true metallicity above this limit and isn't correctly classified as a giant or dwarf, its catalog metallicity estimate will be erroneous.

Jocelyn: That’s a good safety check for users; they are telling you when their data might be suspect based on the model constraints.

Subrahmanyan: So, to summarize The Pristine survey XXIX Data Release two: Photometric metallicities for fifteen millions stars down to G about twenty point five, it’s a catalogue that significantly increases the sample size and depth for probing the metal-poor halo using narrow-band photometry centered on Ca H and K lines <ref:2610.20714#pg1,The Pristine survey XXIX Data Release 2: Photometric metallicities for 15 millions>.

Title and authors: Vera: It allows us to get calibrated metallicity estimates with quantifiable uncertainty bounds, providing median values and percentile ranges to characterize the probability distribution function of a star’s metallicity.

Jocelyn: The results show good agreement with independent DESI metallicities, particularly in the metal-poor regime where they are accurate at the zero point two to zero point three level of dex float.

Subrahmanyan: This consistency means that our theoretical models for chemical evolution within these low-metallicity environments are being tested against real observational constraints from different instruments across the sky.

Vera: They also showed that this system can distinguish between giant and dwarf stars in cross-matched samples by using a model choice based on Gaia parallax information.

Jocelyn: And that’s an improvement because it helps resolve some of the ambiguity about the stellar type, which is crucial when we’re trying to assign metallicity correctly.

Subrahmanyan: That distinction matters for structure formation because dwarfs and giants evolve differently; having a better handle on their initial chemical composition helps us trace the assembly history of those halos.

Vera: The authors noted that this system can identify systematic biases by flagging stars below the model's defined range, which points out potential errors if a star isn't correctly classified as a giant or dwarf.

Jocelyn: So, it’s not just about cataloging stars; it’s about building a robust framework for estimating stellar properties in these faint environments.

Subrahmanyan: The implication is that we can start using this larger, deeper sample to really constrain the physics of star formation at low metallicities, which is where the earliest chemical enrichment signals are strongest.

Vera: And moving forward, the next step involves waiting for Data Release three in two thousand twenty-eight which will come one year after the end of the CFHT Large Program <ref:2610.20714#pg1>.

Jocelyn: So that means we’ll be looking at even more data from this survey in a couple of years, hopefully with even deeper limits.

Subrahmanyan: And if the results hold up as expected, this sustained observational effort will provide better constraints on how dark matter halos and baryonic matter have evolved together over cosmic time.

Vera: We’ll keep watching those results from The Pristine survey XXIX Data Release two seeing how these metal-poor stars fit into our larger picture of galactic evolution <ref:2610.20714#pg1,The Pristine survey XXIX Data Release 2>.

The paper's summary: Subrahmanyan: It means we can sample more of the universe's oldest components, those very metal-poor environments where structure formation physics really matters.

Vera: Exactly. We’re getting a much larger candidate pool for those extremely metal-poor stars down to G about twenty point five, which is what we need to test our models of how the first small structures in the universe assembled.

Jocelyn: And they’ve done a lot of work on how they handle the data itself, using that Cambridge Astronomy Survey Unit pipeline and calibrating everything against Gaia DR3 information.

Subrahmanyan: That calibration is key because it lets us get these calibrated metallicity estimates with quantifiable uncertainty bounds, so we're not just guessing a number.

Vera: Right. They report both a median value and percentile ranges, like the 16th and 84th percentiles of the metallicity distribution function for each star.

Jocelyn: That percentile information is important because it lets us characterize that probability distribution function, showing where the bulk of the stars are clustered in terms of chemical composition.

Subrahmanyan: And they used a training set built from SDSS and APOGEE metallicities, but they specifically added stars from the Boötes I dwarf galaxy to boost those low-metallicity samples below minus two point zero.

Vera: That’s smart because that addition directly targets the very metal-poor regime where we have the least constraints on stellar physics.

Jocelyn: They also have a specific recommendation for those metal-poor stars, suggesting we use the model for giants because for most of them, it's hard to tell if they are dwarfs or giants.

Subrahmanyan: That choice makes sense because in that metal-poor regime, most of those stars live out in the more distant Milky Way halo structure.

Vera: And there’s a good warning on the paper too; they flag any star that falls below the model's defined range as potentially having an erroneous metallicity if it isn't correctly classified as a giant or dwarf.

Jocelyn: That’s a way for users to check their own data; it warns you when your results might be suspect based on the model constraints.

Subrahmanyan: Overall, this release really strengthens our ability to use this survey to constrain the physics of chemical evolution in those very early building blocks of the universe.

Vera: It’s about moving beyond just counting stars and starting to get robust estimates for their chemical makeup in these faint, metal-poor regions.

Jocelyn: And by comparing these Pristine metallicities with independent results from DESI, they show good agreement at the zero point two to zero point three level of dex float in the metal-poor regime.

Subrahmanyan: That cross-validation is pretty strong; it means our survey method is consistent with other large spectroscopic surveys out there.

Vera: So, this Data Release two is basically giving us a much deeper, more reliable catalogue for testing how chemical enrichment happened in the earliest structures we can see.

The paper's improvements: Subrahmanyan: We're now moving past just describing what they found and talking about how we can actually use this data better, which is where these suggestions get interesting for cosmology.

Vera: Right, so it’s not just a catalog; they are proposing specific ways to handle the uncertainty and classification issues we talked about earlier.

Jocelyn: They suggest using Gaia parallax information to make a smarter choice between giant and dwarf stars in cross-matched samples, which is way better than just picking one model blindly.

Subrahmanyan: That makes sense because if we get the stellar type wrong, it messes up how we interpret the metallicity for structure formation models.

Vera: And they also point out that by flagging stars that fall outside the model's defined range, we can identify systematic biases in those photometric metallicity models themselves.

Jocelyn: So if a star has a true metallicity above that limit, but it’s not correctly classified as a giant or dwarf, then our catalog estimate for it is going to be wrong.

Subrahmanyan: That's an important caveat because it tells us the limitations of the model grid we built using those SDSS and APOGEE samples.

Vera: It’s a way to make the data quality check more robust, ensuring that our derived values actually match what the underlying physics predicts.

Jocelyn: Plus, they emphasize using that percentile ranges again—the 16th and 84th percentiles—to give users a clearer picture of the probability distribution function of a star’s metallicity.

Subrahmanyan: That distribution function is crucial for us; it tells us if we are seeing a single population or something more complex in terms of chemical enrichment history.

Vera: So, they're not just giving us numbers; they're giving us tools to understand the statistical nature of these metal-poor star populations.

Jocelyn: It’s like moving from a simple measurement to understanding the whole statistical spread of those measurements in one go.

Subrahmanyan: This level of detail is what lets us test whether the initial conditions we set for structure formation actually match what we observe in these distant halos.

Vera: So, it’s about turning this survey into a tool for deep statistical analysis rather than just another large catalog of stars.

Jocelyn: And they also reinforced the idea that for those metal-poor halo stars, sticking to the giant model is the most practical approach given where most of them are expected to reside.

Subrahmanyan: That’s a pragmatic choice, acknowledging that in those extreme environments, we don't have enough information to perfectly separate dwarfs from giants.

Vera: So the main improvement is making the output more statistically rich and flagging potential systematic errors within the model itself.

Conclusion: Vera: So we’re wrapping up on The Pristine survey XXIX Data Release two: Photometric metallicities for fifteen millions stars down to G about twenty point five today, and what it means for how we see the earliest building blocks of galaxies.

Jocelyn: Basically, they gave us a much deeper look into those metal-poor halo stars than before, using this expanded coverage down to G about twenty point five.

Subrahmanyan: It really helps constrain the physics of structure formation because we are now sampling more of the very first chemical enrichment signals in the universe.

Vera: That’s right. They produced a catalogue with median values and percentile ranges, which lets us characterize that probability distribution function for each star's metallicity.

Jocelyn: And they showed good agreement with DESI metallicities, particularly at the zero point two to zero point three level of dex float in the metal-poor regime.

Subrahmanyan: That consistency is important; it means our survey method is actually sound when we look at these extreme environments.

Vera: It’s about using this larger, deeper sample to really test our theoretical models of how dark matter halos and baryonic matter have evolved together over cosmic time.

Jocelyn: The implication is that we can start building a more robust picture of the chemical evolution history of the Milky Way halo.

Subrahmanyan: This sustained observational effort will provide better constraints on how these halos built up their mass across different epochs in cosmic time.

Vera: So, this Data Release two is a big step forward for probing those faint, metal-poor regions of the sky with high-precision photometry.

Jocelyn: We’ll be keeping an eye on the next release from The Pristine survey and see how much deeper they can push those limits.

Subrahmanyan: And we'll wait to see if this observational data confirms or kills our current theories about the earliest star formation processes.