🔭 astrophysics

Revealing α-Element's Past with Subaru/IRD: Oxygen Abundance of 35 Very Metal-Poor Stars from Near-IR OH lines

This study determines oxygen abundances for 35 very metal-poor stars using Subaru/IRD near-infrared OH lines, identifies a temperature-dependent systematic offset compared to optical [OI] lines, and derives an empirical calibration to align the more precise OH measurements with chemical evolution models.

Original authors: Bakuh Danang Setyo Budi, Wako Aoki, Nicholas Storm, Miho Ishigaki, Tadafumi Matsuno, Teruyuki Hirano, Masayuki Kuzuhara, Jun Nishikawa, Masashi Omiya, Takayuki Kotani, Tomoyuki Kudo, Sebastien Vievard

Published 2026-07-09

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Bakuh Danang Setyo Budi, Wako Aoki, Nicholas Storm, Miho Ishigaki, Tadafumi Matsuno, Teruyuki Hirano, Masayuki Kuzuhara, Jun Nishikawa, Masashi Omiya, Takayuki Kotani, Tomoyuki Kudo, Sebastien Vievard, Motohide Tamura

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the universe as a giant, ancient kitchen. For billions of years, massive stars have been the chefs, cooking up heavy elements like oxygen and iron and tossing them into the cosmic soup when they explode. To understand the very first recipes, astronomers look at "very metal-poor" stars—ancient leftovers from the early universe that still hold the original flavors in their atmospheres.

The big question this paper tackles is: How much oxygen is actually in these ancient stars?

Oxygen is the third most common element in the universe, but measuring it in these dim, ancient stars is like trying to hear a whisper in a hurricane. For decades, astronomers have had two main ways to listen for this whisper. One way uses a "forbidden" light line in the visible spectrum (the [OI] line), which is like a super-reliable, steady drumbeat. The other way uses "OH" molecules in the near-infrared (NIR) spectrum, which is like a whole choir of singers. The problem? The choir (OH lines) has been singing a slightly different tune than the drumbeat (the [OI] line), and nobody was sure who was right.

The Main Discovery: A Temperature-Dependent Mix-Up

The authors, using the Subaru Telescope's IRD instrument, decided to test the choir. They looked at 35 very old stars and measured the oxygen using both the drumbeat and the choir.

They found a fascinating pattern that depends entirely on how hot the star is:

For the warmer red giants (stars with temperatures around 4,600 K or higher), the choir (OH lines) sang a louder note, suggesting more oxygen than the drumbeat. This matches what we expected from previous theories about how 3D effects in hot stars work.
For the cooler red giants (stars with temperatures below 4,600 K), the choir suddenly sang a quieter note. In these cool stars, the oxygen measured by the OH lines was systematically 0.05 to 0.25 dex lower than the reliable drumbeat.

This is a big deal because it flips the script. Usually, we think the molecular lines are overestimating the oxygen. Here, in the cool stars, they are underestimating it.

Why is the choir singing off-key?

The paper suggests that the cool stars have a hidden "heater" in their deep atmosphere layers. While standard models (1D models) assume the atmosphere cools down smoothly as you go up, real 3D simulations suggest that in these cool, metal-poor stars, the deeper layers where the OH lines form are actually 100 to 200 K hotter than we thought.

Think of it like a campfire. If you stand too close (the deep layers), you feel a blast of heat that the standard model didn't predict. This extra heat breaks up the OH molecules, making the "choir" sing quieter. To match the quiet sound we actually hear, the standard model has to guess that there is less oxygen than there really is.

The paper also rules out a few other suspects:

It's not the carbon: They checked if the amount of carbon was messing up the oxygen count. For most of these stars, the answer is "no." The oxygen measurement from the OH lines is surprisingly independent of how much carbon is there.
It's not a measurement error: The "drumbeat" (the [OI] line) is so stable that it barely changes even if you tweak the star's temperature or gravity. The "choir" (OH lines), however, is extremely sensitive. A tiny change in temperature of just 100 K can shift the oxygen measurement by 0.25 dex. This sensitivity is why the choir gets confused by the hidden heat in the cool stars.

The Fix: A New Recipe Card

Since the choir is so sensitive but also has a huge advantage—it has many more lines to measure than the single, weak drumbeat—the authors didn't throw the choir out. Instead, they created a "calibration recipe."

They used the reliable drumbeat ([OI] line) as the truth and figured out a mathematical formula to correct the choir's singing. This formula takes into account the star's temperature, surface gravity, metallicity, and carbon-to-iron ratio.

When they applied this correction:

The scatter (the messy spread of data points) in the oxygen measurements dropped significantly.
The trend of oxygen abundance across different metal levels flattened out, matching the predictions of how the galaxy's chemical evolution should look.

How Sure Are They?

The authors are very confident in the data they collected. They measured 35 stars with high-quality spectra and found a clear, systematic difference between the two methods. They are also confident that the [OI] line is the "gold standard" baseline because it is insensitive to the tricky 3D effects and non-LTE physics that confuse other lines.

However, they are suggesting the cause of the discrepancy (the 3D radiative heating) rather than proving it with a full 3D simulation of every single star. They note that a full 3D/NLTE (non-local thermodynamic equilibrium) analysis for both oxygen and iron is still needed to get the final, perfect picture. But for now, they have shown that with a careful "calibration," the near-infrared OH lines are a reliable tool for exploring the oxygen history of the oldest stars in the galaxy.

In short: The ancient stars were singing a different song than we thought, but not because they were wrong—because our sheet music (the models) was missing a few notes about how hot the deep atmosphere really gets. With a little tuning, the music finally makes sense.

Technical Summary: Revealing $\alpha$ -Element's Past with Subaru/IRD: Oxygen Abundance of 35 Very Metal-Poor Stars from Near-IR OH lines

Problem Statement
Oxygen (O) is a critical tracer for understanding the nucleosynthesis of early massive stars and the chemical evolution of the Galaxy. In very metal-poor (VMP) and extremely metal-poor (EMP) stars ( $-4.0 < \text{[Fe/H]} < -1.5$ ), the oxygen-to-iron ratio ([O/Fe]) serves as a primary indicator of the relative contributions of core-collapse supernovae (CCSNe) versus Type Ia supernovae. However, reliably measuring oxygen abundance in this metallicity regime remains a significant observational challenge.

The standard optical tracer, the forbidden [OI] 6300 Å line, is theoretically robust (insensitive to non-LTE and 3D effects) but becomes excessively weak in V/EMP stars, leading to large random errors and potential telluric contamination. Conversely, near-ultraviolet (NUV) OH lines are strong but suffer from severe 3D hydrodynamical effects and a strong dependence on carbon abundance. Near-infrared (NIR) OH vibro-rotational lines in the H-band (1.5–1.7 $\mu$ m) offer a promising alternative due to their strength and the availability of numerous measurable lines, yet their reliability as an oxygen tracer in the extremely metal-poor regime requires rigorous validation against the [OI] baseline.

Methodology
The authors present a homogeneous analysis of 35 V/EMP stars (34 red giants and 1 dwarf) using high-resolution spectra from the Subaru/IRD (InfraRed Doppler) spectrograph. The study employs the following methodological framework:

Stellar Parameters: Effective temperature ( $T_{\text{eff}}$ ), surface gravity ( $\log g$ ), and microturbulence ( $\xi_t$ ) were uniformly rederived using Gaia DR3 parallaxes and 2MASS photometry, employing empirical color- $T_{\text{eff}}$ relations and Monte Carlo simulations to propagate uncertainties.
Iron Abundance: 1D/NLTE iron abundances were derived from optical Fe I and Fe II equivalent widths from literature, applying non-LTE corrections via the INASAN grid.
Oxygen Abundance Determination:
- NIR OH Lines: Oxygen abundances were derived via 1D/LTE spectral synthesis of H-band OH lines using the Turbospectrum code and MARCS atmospheric models.
- Optical [OI] Line: For 23 stars, oxygen abundances were also derived from the [OI] 6300 Å line using archival high-resolution optical spectra (Subaru/HDS, VLT/UVES, Keck/HIRES, HARPS, ESPRESSO).
Validation and Sensitivity: The study conducted extensive sensitivity analyses on both tracers regarding variations in $T_{\text{eff}}$ , $\log g$ , metallicity, and carbon abundance. Additionally, 3D/LTE and 1D/NLTE corrections were tested for the [OI] line to confirm its stability as a baseline.

Key Results

Sensitivity Discrepancy: A sensitivity analysis confirms that NIR OH lines are highly sensitive to effective temperature ( $\Delta \log \epsilon(\text{O}) / \Delta T \sim \pm 0.25$ dex per 100 K), whereas the forbidden [OI] line is remarkably stable ( $\Delta \text{A(O)} < 0.1$ dex).
Systematic Offset: A temperature-dependent discrepancy exists between the two tracers. In warmer red giants ( $T_{\text{eff}} \gtrsim 4600$ K), OH-based abundances are higher than [OI]-based abundances, consistent with expected 3D effects. However, in cooler red giants ( $T_{\text{eff}} \lesssim 4600$ K), OH-based abundances are systematically lower than [OI]-based abundances by 0.05 to 0.25 dex.
Error Budget: While the [OI] line is physically stable, its weakness in V/EMP stars results in large random errors (often exceeding 0.3 dex). In contrast, the numerous measurable NIR OH lines yield significantly smaller random abundance errors, making the OH tracer statistically more precise despite its systematic sensitivity to stellar parameters.
Carbon Dependence: The study confirms that for the sample (which generally has low [C/O] ratios), the NIR OH abundance is relatively independent of the adopted carbon abundance, unlike NUV OH lines.
Empirical Calibration: To reconcile the systematic offset, the authors derived an empirical linear calibration function based on $T_{\text{eff}}$ , $\log g$ , [Fe/H], and [C/Fe]. This calibration aligns the 1D/LTE OH abundances onto the [OI] scale.

Significance and Claims
The paper claims that while NIR OH lines are subject to systematic temperature-dependent offsets in 1D/LTE analysis, they offer superior statistical precision due to the availability of multiple lines. By applying the derived empirical calibration, the authors successfully reduce the scatter and temperature dependence in the [O/Fe] versus [Fe/H] plane.

The corrected OH-based [O/Fe] trend exhibits a flatter slope and reduced scatter, bringing the observational data into fair agreement with Galactic chemical evolution models (specifically those by Kobayashi et al. 2020). The authors conclude that, with careful calibration, NIR OH lines from Subaru/IRD are a reliable and useful tool for exploring oxygen abundance distributions in the oldest stars, providing a viable alternative when the optical [OI] line is too weak to measure accurately.

The study does not claim to have solved the full 3D/NLTE physics of OH formation but demonstrates that an empirical correction based on a high-quality sample can effectively bridge the gap between the statistically precise but systemically biased OH tracer and the physically robust but statistically noisy [OI] tracer.

More like this