A long-period planet orbiting a nearby Sun-like star

TL;DR

This study refines the understanding of the HD 134987 system by confirming a second planet through improved radial velocity analysis and photometry, highlighting the importance of considering stellar activity in exoplanet detection.

Contribution

The paper presents a double Keplerian fit for HD 134987, revealing a second planet and addressing stellar activity effects, improving the accuracy of exoplanet characterization.

Findings

Detection of a second planet with 0.82 Mjup mass

Orbital periods of 258 and 5000 days for the planets

Photometry shows no stellar variability supporting planetary origin

Abstract

The Doppler wobble induced by the extra-solar planet HD 134987b was first detected by data from the Keck Telescope nearly a decade ago, and was subsequently confirmed by data from the Anglo-Australian Telescope. However, as more data have been acquired for this star over the years since, the quality of a single Keplerian fit to that data has been getting steadily worse. The best fit single Keplerian to the 138 Keck and AAT observations now in hand has an root-mean-square (RMS) scatter of 6.6 m/s. This is significantly in excess of both the instrumental precision achieved by both the Keck and Anglo-Australian Planet Searches for stars of this magnitude, and of the jitter expected for a star with the properties of HD134987. However, a double Keplerian (i.e. dual planet) fit delivers a significantly reduced RMS of 3.3 m/s. The best-fit double planet solution has minimum planet masses of 1.59 and 0.82Mjup, orbital periods of 258 and 5000d, and eccentricities of 0.23 and 0.12 respectively. We find evidence that activity-induced jitter is a significant factor in our fits and do not find evidence for asteroseismological p-modes. We also present seven years of photometry at a typical precision of 0.003mag with the T8 0.8m automatic photometric telescope at Fairborn observatory. These observations do not detect photometric variability and support the inference that the detected radial-velocity periods are due to planetary mass companions rather than due to photospheric spots and plages.