[Paper Review] The HARPS search for southern extra-solar planets XXXV. The interesting case of HD41248: stellar activity, no planets?

N. C. Santos, Annelies Mortier+22 more|HAL (Le Centre pour la Communication Scientifique Directe)|Apr 24, 2014

Stellar, planetary, and galactic studiesPhysics and Astronomy75 references81 citations

TL;DR

This study reanalyzes radial velocity data for the metal-poor star HD 41248, originally claimed to host a resonant pair of super-Earth planets. Using 162 new HARPS measurements and activity diagnostics, the authors find no evidence for planetary signals; instead, the previously reported 25-day and 18-day signals are attributed to stellar activity, particularly evolving active regions on a differentially rotating star.

ABSTRACT

The search for planets orbiting metal-poor stars is of uttermost importance for our understanding of the planet formation models. However, no dedicated searches have been conducted so far for very low mass planets orbiting such objects. Only a few cases of low mass planets orbiting metal-poor stars are thus known. Amongst these, HD41248 is a metal-poor, solar-type star on which a resonant pair of super-Earth like planets has In the present paper we present a new planet search program that is using the HARPS spectrograph to search for Neptunes and Super-Earths orbiting a sample of metal-poor FGK dwarfs. We then present a detailed analysis of an additional 162 radial velocity measurements of HD41248, obtained within this program, with the goal of confirming the existence of the proposed planetary system. We analyzed the precise radial velocities, obtained with the HARPS spectrograph, together with several stellar activity diagnostics and line profile indicators. A careful analysis shows no evidence for the planetary system previously announced. One of the signals, with a period of about 25 days, is shown to be related to the rotational period of the star, and is clearly seen in some of the activity proxies. The remaining signal (P~18 days) could not be convincingly retrieved in the new data set. We discuss possible causes for the complex (evolving) signals observed in the data of HD41248, proposing that they may be explained by the appearance and disappearance of active regions on the surface of a star with strong differential rotation, or by a combination of the sparse data sampling and active region evolution.

Motivation & Objective

• To confirm the existence of a proposed resonant pair of super-Earths around the metal-poor star HD 41248 using new radial velocity data.
• To investigate whether the previously detected 25-day and 18-day signals are caused by planetary companions or stellar activity.
• To assess the impact of stellar activity on radial velocity measurements in low-mass planet searches, especially in metal-poor stars.
• To evaluate the reliability of Keplerian modeling in the presence of complex, time-varying activity signals.
• To demonstrate the importance of long-term, densely sampled data for disentangling planetary signals from stellar noise.

Proposed method

• Acquired 162 new radial velocity measurements of HD 41248 using the HARPS spectrograph at the ESO 3.6-m telescope.
• Analyzed radial velocities in conjunction with stellar activity indicators, including bisector span, cross-correlation function (CCF) bisector, and Hα line profile measurements.
• Used Bayesian model comparison to assess the significance of Keplerian signals versus activity-induced variations.
• Evaluated the rotational period of the star using activity proxies and compared it to the reported planetary periods.
• Modeled the evolution of active regions on a differentially rotating star to explain the complex, time-varying signals.
• Assessed data sampling effects by comparing results from sparsely sampled data sets (#1 and #2) with a densely sampled set (#3).

Figure 1: Metallicity (top) and effective temperature (bottom) distributions for the stars in our sample.

Experimental results

Research questions

• RQ1 Do the previously reported radial velocity signals at ~25 and ~18 days in HD 41248 correspond to planetary companions or stellar activity?
• RQ2 Can the complex, evolving signal structure in the radial velocity data be explained by the appearance and disappearance of active regions on a differentially rotating star?
• RQ3 To what extent does sparse data sampling contribute to the misinterpretation of stellar activity as planetary signals?
• RQ4 How do activity diagnostics like bisector span and CCF profile shape correlate with the observed radial velocity variations?
• RQ5 Can a physical model of evolving active regions explain the observed signal pattern without invoking planetary companions?

Key findings

• The 25-day signal, previously interpreted as a planetary orbital period, is shown to match the star's rotational period and is strongly correlated with activity indicators.
• The 18-day signal, previously reported as a planetary signal, is not consistently retrieved in the new data set and lacks robust support in the Bayesian analysis.
• The complex signal structure in the radial velocity data is best explained by the evolution of active regions on a star with strong differential rotation.
• The data sampling in the original study (62 measurements over ~10 years) was insufficient to disentangle activity signals from potential planetary signals.
• A densely sampled data set (#3) was critical in identifying the true origin of the signals, demonstrating that long-term, high-cadence monitoring is essential for reliable planet detection.
• The study highlights the risk of false-positive detections in radial velocity surveys when stellar activity is not properly modeled, even with Bayesian analysis.

Figure 2: Time series of the radial velocity, stellar activity, FWHM, and BIS.