[Paper Review] Extrasolar planets and brown dwarfs around A-F type stars. I. Performances of radial velocity measurements, first analyses of variations

F. Galland, A.‐M. Lagrange+6 more|ArXiv.org|Sep 6, 2005

Stellar, planetary, and galactic studiesPhysics and Astronomy14 references103 citations

TL;DR

This paper presents a novel Fourier-based radial velocity measurement method optimized for A-F type stars, which exhibit high rotational velocities that challenge traditional cross-correlation techniques. The method achieves radial velocity uncertainties as low as 0.03 m s⁻¹ with HARPS and 0.16 m s⁻¹ with ELODIE, enabling the detection of planets and brown dwarfs around these stars—previously considered observationally inaccessible—demonstrated through confirmed detections of known planets and binary systems.

ABSTRACT

We present the performances of a radial velocity measurement method that we developed for A-F type stars. These perfomances are evaluated through an extensive set of simulations, together with actual radial velocity observations of such stars using the ELODIE and HARPS spectrographs. We report the case of stars constant in radial velocity, the example of a binary detection on HD 48097 (an A2V star, with vsini equal to 90 km/s) and a confirmation of the existence of a 3.9 MJup planet orbiting around HD 120136 (Tau Boo). The instability strip problem is also discussed. We show that with this method, it is in principle possible to detect planets and brown dwarfs around A-F type stars, thus allowing further study of the impact of stellar masses on planetary system formation over a wider range of stellar masses than is currently done.

Motivation & Objective

• Overcome the challenge of low signal-to-noise and broad spectral lines in A-F type stars, which hinder traditional radial velocity methods.
• Develop a robust radial velocity measurement technique suitable for high-rotational-velocity stars (v sin i up to 200 km s⁻¹) using Fourier cross-spectrum analysis.
• Demonstrate the feasibility of detecting low-mass companions (planets and brown dwarfs) around A-F type main sequence stars, extending the radial velocity method beyond late-type stars.
• Assess the method's performance through simulations and real observations, validating precision and reliability on constant-velocity stars, binaries, and known exoplanet systems.
• Establish mass detection limits for different stellar types and instruments, enabling future surveys to target planetary systems around higher-mass stars.

Proposed method

• Apply a Fourier-domain cross-spectrum method to correlate a target star's spectrum with a high-S/N reference spectrum specific to that star, derived from co-added observations.
• Extract radial velocity from the phase of the cross-spectrum by minimizing the imaginary part of a complex phase-corrected cross-spectrum using a least-squares fit.
• Model radial velocity uncertainty (ε_RV) as a function of v sin i and signal-to-noise ratio (S/N), with empirical fits derived from simulations and real data: ε_RV = 0.16 × v sin i^1.54 × (200/S/N) for ELODIE and ε_RV = 0.032 × v sin i^1.50 × (400/S/N) for HARPS.
• Account for instrumental stability and photon noise limits, showing that uncertainty scales with v sin i^1.5 when lines are resolved and photon noise dominates.
• Use the radial velocity uncertainty to compute mass detection limits for circular orbits, assuming ±3ε_RV as the detection threshold.
• Compare performance between ELODIE (1.93m telescope) and HARPS (3.6m telescope), attributing the 5× improvement in precision to higher S/N and increased pixel sampling per spectral element.

Experimental results

Research questions

• RQ1 Can radial velocity measurements be reliably performed on A-F type stars despite their broad, weak spectral lines and high rotational velocities?
• RQ2 What is the achievable radial velocity precision for A-F type stars as a function of v sin i and signal-to-noise ratio?
• RQ3 Can this method detect known planets and binary systems around A-F stars, validating its reliability?
• RQ4 What are the mass detection limits for planets and brown dwarfs around A-F type stars using ELODIE and HARPS?
• RQ5 How does the method’s performance compare between ELODIE and HARPS, and what instrumental factors contribute to the difference?

Key findings

• The Fourier-based radial velocity method achieves radial velocity uncertainties of 0.16 m s⁻¹ with ELODIE and 0.032 m s⁻¹ with HARPS, scaling with v sin i^1.5 and S/N.
• For A-type stars with v sin i ≤ 100 km s⁻¹, the method enables detection of planets with orbital periods <10 days; for v sin i ≤ 40 km s⁻¹, periods up to 1000 days are accessible.
• With HARPS, the detection limit for a 10-day period planet around an A5V star with v sin i = 60 km s⁻¹ is reduced to 0.7 M_Jup, compared to 4 M_Jup with ELODIE.
• The method successfully confirmed the 3.9 M_Jup planet around HD 120136 (Tau Boo) and detected a binary system in HD 48097, validating its accuracy.
• The radial velocity uncertainty is dominated by photon noise when v sin i ≥ 15 km s⁻¹, while instrumental limits dominate at lower v sin i for ELODIE.
• The 5× improvement in radial velocity precision with HARPS over ELODIE arises from a 2× increase in S/N per pixel and a 4× increase in pixel sampling per spectral element, yielding a total factor of 4–5 improvement.