🔭 astrophysics

Radial velocity and atmospheric parameter calculations for the GaiaNIR spectrograph

This study identifies the 1926–1968 nm K-band region as the optimal spectral range for the proposed GaiaNIR near-infrared spectrograph, as it maximizes radial velocity precision and atmospheric parameter accuracy for mapping dust-obscured regions of the Milky Way while minimizing interstellar extinction.

Original authors: Szabolcs Mészáros, David Hobbs, Anna Liptrott, David Katz, Ricardo Schiavon, Viktória Pap, Anthony G. A. Brown, George Seabroke, Joss Bland-Hawthorn, Ronny Blomme, Nicholas A. Walton

Published 2026-07-20

📖 4 min read☕ Coffee break read

CC0 1.0

Original authors: Szabolcs Mészáros, David Hobbs, Anna Liptrott, David Katz, Ricardo Schiavon, Viktória Pap, Anthony G. A. Brown, George Seabroke, Joss Bland-Hawthorn, Ronny Blomme, Nicholas A. Walton

Original paper dedicated to the public domain under CC0 1.0 (http://creativecommons.org/publicdomain/zero/1.0/). ✨ 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 Milky Way as a giant, swirling city of stars, but with a catch: most of it is hidden behind thick, cosmic curtains of dust. For decades, our best space telescope, Gaia, has been mapping the stars we can see, but it's like trying to map a city while only looking through a few clear windows in a foggy night. It sees the neighborhood around us perfectly, but the dense, interesting parts of the galaxy—the center, the spiral arms, the far side—are completely blocked by this dust. To see through the fog, astronomers need to switch from looking with "optical" eyes (like human vision) to "near-infrared" eyes. Infrared light is like a superpower that can slip through the dust clouds that stop visible light, revealing the hidden streets of our galactic city. But here's the tricky part: building a telescope that can see in this new light is hard, and if you want to measure how fast stars are moving (their "radial velocity") or what they are made of (their "atmospheric parameters"), you need to pick the exact right slice of the rainbow to look at. If you pick the wrong slice, the stars look blurry, and the data is useless.

This paper is essentially a massive "test drive" for a future mission called GaiaNIR. The team wanted to figure out which specific color of infrared light would give the sharpest, most precise measurements for the new spectrograph they are planning to build. They didn't just guess; they created a virtual universe of 10,000 fake stars with different temperatures, sizes, and compositions. They simulated looking at these stars through different "windows" of the infrared spectrum, ranging from 800 to 2300 nanometers, and tested how well they could measure the stars' speeds and chemical makeup at various levels of sharpness (resolution). They were looking for the "Goldilocks zone": a wavelength that isn't just good at seeing through the dust, but also packed with enough distinct lines to act as a fingerprint for every type of star, from cool red dwarfs to hot giants.

The researchers found that not all infrared windows are created equal. They tested several candidates, including some in the "H-band" (around 1500 nm) and others in the "K-band" (around 2000 nm). They ruled out the shorter wavelengths (closer to 800 nm) because, even though they work well, they don't offer enough advantage in piercing the dust compared to the older Gaia telescope. They also found that the H-band, while decent, requires extremely high resolution to get good results, which might be too difficult to build and manage with the limited space on a detector.

The clear winner, according to their simulations, is a specific window in the K-band, specifically between 1926 and 1968 nanometers. The paper suggests that if the GaiaNIR spectrograph focuses on this narrow slice of light with a resolution between 16,100 and 20,100, it could measure the speed of the brightest stars with a precision of about 160 to 260 meters per second. That is incredibly precise—imagine trying to measure the speed of a car from miles away and knowing if it's moving at 60 mph or 60.1 mph. While this specific window is slightly less precise for measuring the surface gravity of stars compared to other options, it offers the best overall balance. It allows astronomers to measure the speed and chemical composition of ten different elements (like Oxygen, Sodium, Magnesium, and Calcium) for most stars.

However, the paper also points out a few limitations. For the very coolest stars (M-dwarfs), the K-band is tricky because water vapor in their atmospheres creates a "blanket" that hides most chemical details, leaving only Calcium visible. Additionally, for very hot stars (OBA types), there are almost no lines to measure in this window, making them harder to study than with the older Gaia telescope. The authors also identified a "runner-up" window around 1158 to 1202 nanometers. This spot is great for seeing a wider variety of star temperatures, but it doesn't cut through the dust as well as the K-band, which is the main reason GaiaNIR exists in the first place.

Ultimately, the paper concludes that the 1926–1968 nm window is the strategic choice. It offers a precision that rivals the current Gaia mission but unlocks the ability to see the dust-obscured regions of the Milky Way that have been invisible until now. The authors emphasize that while their simulations look very promising, the final design will need to account for real-world challenges, like how crowded the star fields are in the galactic center, which might force them to adjust the resolution or the size of the window they use. But for now, this study lights the path forward: to map the hidden galaxy, we need to tune our instruments to the K-band.

Technical Summary: Radial velocity and atmospheric parameter calculations for the GaiaNIR spectrograph

Problem Statement
The Gaia mission has revolutionized astrometry and spectroscopy in the optical regime but is severely limited by interstellar dust extinction, preventing a comprehensive mapping of the inner Galaxy, the bulge, and the far side of the Milky Way disk. The proposed GaiaNIR mission aims to overcome this by operating in the near-infrared (NIR, 800–2300 nm), where extinction is significantly lower. While GaiaNIR plans to perform astrometry and photometry, its scientific return could be exponentially enhanced by adding a moderately high-resolution spectrograph ( $R > 10,000$ ) to measure radial velocities ( $v_{rad}$ ), atmospheric parameters ( $T_{eff}$ , $\log g$ , $[M/H]$ ), and chemical abundances for hundreds of millions of stars. However, technical constraints (detector size, telemetry, and cooling requirements) prevent a single spectrograph from covering the entire 800–2300 nm range at high resolution. Consequently, the critical challenge is identifying the optimal wavelength window that balances high radial velocity precision with the ability to derive accurate atmospheric parameters and abundances, while maximizing the benefit of reduced interstellar extinction.

Methodology
The authors conducted a systematic simulation study to identify optimal spectral windows without relying on a comprehensive empirical library of high-resolution NIR spectra, which currently does not exist.

Synthetic Data Generation: The study utilized the BOSZ synthetic spectral library, generating 10,000 mock spectra across a wide range of stellar parameters ( $T_{eff}$ : 2800–10,000 K, $\log g$ : 0–5.5 dex, $[M/H]$ : -1.5 to 0.5 dex). The models incorporated LTE approximations using MARCS and ATLAS9 atmospheres, updated continuous opacities, and 23 molecular line lists.
Simulation Parameters: Spectra were generated at four resolutions ( $R = 5,000, 10,000, 15,000, 20,000$ ) and subjected to random Gaussian noise to simulate Signal-to-Noise ratios (S/N) between 5 and 100.
Radial Velocity Analysis: Mock spectra were Doppler-shifted and cross-correlated with ideal templates using the crosscorrRV function. The analysis scanned 100 nm windows across the 850–2250 nm range in 5 nm increments. Performance was evaluated using the scatter ( $\sigma$ ), median absolute deviation (MAD), and 95th percentile (P95) of velocity residuals.
Atmospheric Parameter Derivation: The FERRE code was employed to fit $T_{eff}$ , $\log g$ , and $[M/H]$ to the mock spectra using $\chi^2$ minimization against precomputed grids.
Sample Stratification: The analysis was stratified into five subsamples based on temperature and surface gravity (Red Dwarfs, Red Giants, FGK dwarfs/giants, and Hot A-type stars) to account for the varying dominance of atomic versus molecular lines.
Validation: The methodology was validated by comparing derived precisions against Gaia Data Release 3 (DR3) empirical performance in the optical Ca II triplet region, confirming that the simulation approach yields realistic uncertainty estimates for high S/N targets.

Key Results
The study identified six candidate windows for further refinement, ultimately narrowing the focus to two primary regions:

The K-band (1926–2000 nm):
- Performance: This region offers the highest radial velocity precision for cool stars (FGKM types), with estimated precisions of 160–260 m s $^{-1}$ at $R \approx 16,000–20,000$ for the brightest targets.
- Atmospheric Parameters: It provides $T_{eff}$ precision better than 100 K and metallicity precision around 0.06–0.08 dex, comparable to Gaia's current capabilities. However, $\log g$ precision is somewhat lower than optical results.
- Extinction: The K-band offers a critical advantage with interstellar reddening approximately 7.4 times lower than in the optical, enabling the mapping of dust-obscured regions.
- Abundances: Ten species can be derived (O, Na, Mg, Al, Si, Ca, Ti, V, Cr, Mn, Ni), though M-dwarfs are limited primarily to Ca due to H $_2$ O blanketing.
- Limitations: Performance degrades significantly for stars hotter than 7000–8000 K due to a lack of metallic lines, though the Br- $\delta$ Hydrogen line offers limited utility.
The J/H-band Transition (1158–1202 nm):
- Performance: This region provides slightly lower radial velocity precision ( $\approx 340$ m s $^{-1}$ ) compared to the K-band but offers a wider effective temperature range, extending up to 12,000–16,000 K.
- Atmospheric Parameters: $T_{eff}$ and metallicity precisions are competitive, though $\log g$ determination remains challenging.
- Extinction: The reduction in extinction is marginal compared to the optical, limiting its strategic value for probing the most obscured galactic regions.
- Abundances: Ten species are available (Mg, Si, K, Ca, Ti, V, Cr, Mn, Co, Ni). A distinct advantage is the ability to measure abundances for M-dwarfs below 4000 K (Mg, Si, K, Ca, Ti, Cr) where H $_2$ O lines are less dominant than in the K-band.

Conclusions and Significance
The paper concludes that the 1926–1968 nm window (centered at $\lambda_C \approx 1947$ nm) represents the optimal strategic compromise for the GaiaNIR spectrograph. While the narrower 1984–2000 nm window offers the highest radial velocity precision, it sacrifices atmospheric parameter accuracy. The 1926–1968 nm region, operating at resolutions up to $R \approx 20,000$ , delivers radial velocity precisions of 160–260 m s $^{-1}$ and atmospheric parameter precisions comparable to Gaia, all while leveraging the K-band's superior transmission through dust.

The authors emphasize that this selection enables GaiaNIR to fulfill its primary mandate: mapping the dust-obscured inner Galaxy and the far side of the disk. However, the paper notes significant caveats:

Crowding: The K-band may suffer from severe spectral crowding in the Galactic plane, potentially necessitating lower resolutions or smaller window sizes, which requires further study.
M-dwarfs: The inability to derive robust abundances for the coolest M-dwarfs (due to H $_2$ O) is a noted disadvantage, as these stars constitute a large fraction of the target population.
Future Work: The authors state that creating a new, high-quality line list for the K-band is essential, as current Earth-based observations are heavily contaminated by telluric lines. Future studies must address the trade-offs between resolution, window size, and stellar crowding to finalize the spectrograph design.

The study does not claim to have solved all design challenges but provides a data-driven framework for selecting the wavelength region that maximizes the scientific return of the GaiaNIR mission within its technical constraints.

More like this