Source-linked AI summary
ARES v2 - new features and improved performance
S. G. Sousa, N. C. Santos, V. Adibekyan, E. Delgado-Mena, G. Israelian
TL;DR
Precise equivalent-width analysis depends on continuum placement, which is difficult and subjective, while ARES's rejt parameter remains adjustable. ARES v2 estimates continuum noise in selected spectral regions to set rejt automatically, and its parallelized implementation improves performance while remaining compatible with the prior version.
Problem
Method
ARES v2 can estimate continuum noise in selected spectral regions and relate it to a suitable rejt value for automatic continuum setting.
Source: §3 · p. 2Results
Takeaways & Limitations
ARES's systematic measurements support differential analyses that require the same method in calibration and spectral analysis.
Source: §3.2 · p. 3ARES-derived errors are suitable for preliminary detection of poor measurements, but users should still check for outliers before further analysis.
Source: §5 · p. 4
Abstract
from arXiv · showhide
Aims: We present a new upgraded version of ARES. The new version includes a series of interesting new features such as automatic radial velocity correction, a fully automatic continuum determination, and an estimation of the errors for the equivalent widths. Methods: The automatic correction of the radial velocity is achieved with a simple cross-correlation function, and the automatic continuum determination, as well as the estimation of the errors, relies on a new approach to evaluating the spectral noise at the continuum level. Results: ARES v2 is totally compatible with its predecessor. We show that the fully automatic continuum determination is consistent with the previous methods applied for this task. It also presents a significant improvement on its performance thanks to the implementation of a parallel computation using the OpenMP library.
1. Introduction
Line-by-line spectral analysis relies on precise equivalent-width measurements of spectral-line strength. ARES automates these measurements, and its use has expanded across stellar studies and surveys.
1. Introduction
- Equivalent widths measure spectral-line strength, a quantity needed for most line-by-line analyses.The paper describes equivalent-width methods as having been used for decades in spectroscopy.Source: §1 · p. 1
- Automatic codes such as ARES aim to make equivalent-width measurements precise and systematic while reducing the burden of interactive routines.ARES follows steps similar to those an astronomer would take with interactive routines.Source: §1 · p. 1
- ARES has been used in stellar-characterization studies and large spectroscopic surveys, and its algorithms are being implemented in recipes for ESPRESSO.The cited examples include the GAIA-ESO Survey and ESPRESSO (Mégevand et al. 2012; Di Marcantonio et al. 2012; Pepe et al. 2014).Source: §1 · p. 1
2. New features
ARES v2 adds automatic radial-velocity correction, continuum-related options, EW error estimates, and support for ASCII spectra; it also parallelizes computation while retaining compatibility with the previous version.
2. New features
- Radial-velocity correction can use either a user-supplied stellar velocity or a rest-frame line mask for ARES to estimate the velocity.The previous version required input spectra to be fully corrected for radial velocity.Source: §2 · p. 1
- Users can request automatic rejt estimation by specifying spectral regions, whose noise ARES relates to a suitable rejt value.The regions should be selected carefully and avoid strong lines; the paper recommends line-free or nearly line-free regions for solar-type stars.Source: §2 · p. 2
- Setting rejt to −2 lets ARES interpolate wavelength-specific rejt values from a two-column wavelength-and-rejt file.ARES adjusts the continuum for each spectral region using the interpolated values.Source: §2 · p. 2
- ARES adds an efficient estimate of EW errors, addressing the original code's lack of error estimation or flags for problematic measurements.The new feature is particularly relevant when analyses rely on few spectral lines.Source: §2 · p. 2
- ARES can read ASCII spectra when no FITS file is provided, provided the data have wavelength and flux in two space-separated columns.Source: §2 · p. 2
- ARES scans spectra for zero-flux values and replaces them with unity to avoid unexpected crashes.Zero flux can occur in spectral-coverage gaps and cause problems for spectral normalization.Source: §2 · p. 2
- Performance improves from ∼60% for K stars to ∼250% for the F and G stars with OpenMP parallelization.The paper reports benchmark comparisons with the older version.Source: §2 · p. 2
- The new features remain compatible with the previous ARES version; users may choose or discard them except that errors always appear in the output.The paper also says performance improves on multicore machines.Source: §2 · p. 2
3. Automatic rejt
ARES v2 automates rejt estimation to reduce subjective continuum placement, using spectral noise from selected regions. In the comparison with recommended rejt values, most measured equivalent widths remain close to identity, though the automatic option requires care when selected regions contain contaminants.
3. Automatic rejt
- Because continuum placement is subjective, ARES adds automatic rejt estimation from user-selected regions that should avoid strong lines.An empirical S/N–rejt relation is recommended to homogenize measurements across similar spectra observed in the same spectroscopic configurations.Source: §3 · p. 2
3.1. Estimating the signal-to-noise ratio
- ARES estimates S/N by smoothing each selected region, subtracting the smoothed spectrum to isolate high-frequency noise, then taking the median of mean-flux-to-noise-standard-deviation ratios.Source: §3.1 · pp. 2–3
- For Alpha Cen A, the three-region median S/N estimate is 256.29, compatible with the spectrum header’s ~275.Source: §3.1 · p. 3
3.2. Auto-rejt and ARES local continuum determination
- Selected ranges may contain cosmic rays or telluric lines, so auto-rejt should be used cautiously.Users with prior S/N knowledge can enter it directly instead of estimating it from the ranges.Source: §3.2 · p. 3
- Against Sousa et al. (2008) rejt recommendations, auto-rejt yields ~0.7 mÅ mean offset, ~4.4 mÅ dispersion, and ~96% of lines near identity across seven spectra.The spectra span S/N ~70–~2000, with ~380 lines measured in each; most of the large spread comes from lower-S/N spectra.Source: Figure 2 · p. 3, §3.2 · p. 3
- Regardless of the selected option or rejt value, ARES measures systematically, which matters for differential analyses and line-ratio applications.Source: §3.2 · p. 3
4. EW error estimation
ARES estimates equivalent-width errors from its Gaussian line fits, using the fitted-parameter uncertainties and a noise estimate tied to the continuum. The errors include continuum-position uncertainty and part of the uncertainty associated with blended lines, and can help identify problematic measurements, especially when few lines are available.
4. EW error estimation
- The fit uses a per-pixel error based on spectral S/N, with equal weight assigned to every pixel.ARES requires only wavelength and flux inputs; its error estimate includes continuum-position uncertainty through σF.Source: §4 · p. 4
- ARES estimates EW errors by propagating uncertainties in the fitted line depth and width.The equivalent width is computed from the area of the fitted Gaussian profile.Source: §4 · p. 4
- The error estimate captures continuum-position uncertainty and part of the error from blended lines.Source: §4 · p. 4
- Per-line uncertainty estimates help flag problematic measurements, particularly when analyses rely on few lines.Na, Mg, Al, Sc, and Mn analyses typically use two to four lines per element.Source: §4 · p. 4
5. Summary
ARES v2 adds automatic radial-velocity correction, automatic continuum positioning, and EW error estimation while remaining compatible with the prior version. It runs faster with multicore parallel processing, but its error estimates are intended for preliminary screening, with outliers still requiring user review.
5. Summary
- The new version adds automatic radial-velocity correction, automatic continuum positioning, and EW error estimation.Source: §5 · p. 4
- ARES errors support preliminary detection of poor measurements, but users should check for outliers before further analysis.Source: §5 · p. 4
- ARES v2 remains fully compatible with the previous version and works faster through multicore parallel processing.Source: §5 · p. 4