The magnetic field and multiple planets of the young dwarf AU Mic

Abstract

In this paper we present an analysis of near-infrared spectropolarimetric and velocimetric data of the young M dwarf AU Mic, collected with SPIRou at the Canada-France-Hawaii telescope from 2019 to 2022, mostly within the SPIRou Legacy Survey. With these data, we study the large- and small-scale magnetic field of AU Mic, detected through the unpolarized and circularly-polarized Zeeman signatures of spectral lines. We find that both are modulated with the stellar rotation period (4.86 d), and evolve on a timescale of months under differential rotation and intrinsic variability. The small-scale field, estimated from the broadening of spectral lines, reaches 2.61 ± 0.05 kG. The large-scale field, inferred with Zeeman-Doppler imaging from Least-Squares Deconvolved profiles of circularly-polarized and unpolarized spectral lines, is mostly poloidal and axisymmetric, with an average intensity of 550 ± 30 G. We also find that surface differential rotation, as derived from the large-scale field, is ≃30% weaker than that of the Sun. We detect the radial velocity (RV) signatures of transiting planets b and c, although dwarfed by activity, and put an upper limit on that of candidate planet d, putatively causing the transit-timing variations of b and c. We also report the detection of the RV signature of a new candidate planet (e) orbiting further out with a period of 33.39 ± 0.10 d, i.e., near the 4:1 resonance with b. The RV signature of e is detected at 6.5σ while those of b and c show up at ≃ 4σ, yielding masses of $10.2^{+3.9}_{-2.7}$ and $14.2^{+4.8}_{-3.5}$ M⊕ for b and c, and a minimum mass of $35.2^{+6.7}_{-5.4}$ M⊕ for e.

Validating AU Microscopii d with Transit Timing Variations

“…We determined the super-period of AU Mic b's TTVs, which we then used to estimate the orbital period of AU Mic d. We modeled the high-eccentricity and loweccentricity two-planet configurations. The high-eccentricity configurations' high eccentricities (e > 0.2) and very small inferred masses (K < 0.07 m s −1 ) are in strong disagreement with those from RV models (Cale et al 2021;Donati et al 2023) and a secondary-eclipse search (Kevin et al 2023, in preparation). While this configuration does model a TTV super-period, it is clearly not dynamically settled given the apparent orbital migrations of AU Mic b and c and the extreme fluctuations in both planets' eccentricities and inclinations in the rebound simulations.…”

“…So, in theory, planet d may not exist if the system is currently dynamically unstable. However, future TTVs will clearly rule out this case, and current RVs from Cale et al (2021) and Donati et al (2023), while challenging to constrain the eccentricity, do not give any clear indication that this system is highly eccentric. Additionally, the secondary-eclipse model of AU Mic b by K. I.…”

“…In Section 4.3, we presented a high-eccentricity two-planet TTV model that does exhibit the super-period (Figure 4), making it seems a potentially good fit for a two-planet system with a stable but unlikely eccentric-aligned configuration with abnormally small masses inconsistent with the RV masses from Cale et al (2021) and Donati et al (2023). After having the very small masses replaced with those RV masses, the higheccentricity two-planet model provides an example of an unstable eccentric system that does model the TTVs and does persist for 20 Myr, as seen in Figure 6, but it is clearly not dynamically settled given the apparent orbital migrations of AU Mic b and c and the extreme fluctuations in both planets' eccentricities and inclinations in the rebound simulations.…”

“…Wittrock et al 2022;Donati et al 2023), that the RV models fromCale et al (2021) andDonati et al (2023) place both…”

“…Additionally, the detrending parameters flare (Spitzer), sky (Spitzer and PEST), Sky/Pixel_T1 (LCOGT), and SKY (ASTEP) were set as additive while the remaining detrending parameters were set as multiplicative. a Multiplicative detrending coefficient.from RV literature (e.g.,Cale et al 2021;Donati et al 2023). Moreover, the angular momentum deficit (AMD;Laskar 1997Laskar , 2000Laskar & Petit 2017) criteria built within the Exo-Striker package indicated that this model is unstable.…”