Discovery of a rapidly evolving global magnetic field in the M-dwarf YZ Cet and constraints on the magnetic field of its planet YZ Cet b

TL;DR

YZ Ceti shows rapid evolution of its large-scale magnetic field on timescales of a few rotation periods, as revealed by SPIRou spectropolarimetry across two epochs and ZDI reconstructions. The rotation period is $P_{ m rot} \approx 69.7$ days, with an evolution timescale $\theta_3 \approx 204$ days, while the mean surface field grows from $|B| \sim 219$ G to $|B| \sim 298$ G and the dipole fraction increases from ~53% to ~75% between 2023 and 2024. The dipole obliquity changes from $\sim 17.5^{\circ}$ to $\sim 51.5^{\circ}$ and axisymmetry drops from ~83% to <35%, indicating substantial topology reconfiguration and a possible polarity reversal. By combining magnetic maps with radio SPI observations, the study yields higher lower limits on YZ Cet b’s magnetic field, $B_{ m planet} \gtrsim 18$ G in 2023 and $\gtrsim 8$ G in 2024, though these inferences depend on dipolar assumptions and require simultaneous observations. Overall, the results point to a dynamo regime in slowly rotating M dwarfs that can produce rapid, large-scale magnetic changes, underscoring the need for coordinated spectropolarimetric and radio campaigns to interpret SPI signatures and characterize exoplanetary magnetospheres.

Abstract

We present a spectropolarimetric study of the nearby M4.5V exoplanet host star YZ Cet, based on near-infrared observations obtained with the SpectroPolarimètre InfraRouge (SPIRou) at the Canada--France--Hawaii Telescope. We detect striking changes in the large-scale magnetic field strength and geometry over the course of just a few stellar rotations, a level of short-term global magnetic field evolution rarely reported in M dwarfs. We modeled the temporal variation of the longitudinal magnetic field using a Gaussian regression process, which allowed us to robustly determine the stellar rotation period and quantify the evolution timescale of the magnetic field. Independent Zeeman Doppler Imaging reconstructions of the two epochs confirm a significant reconfiguration of the star's global magnetic strength and topology. The detection of a weaker, complex, axisymmetric magnetic field (mean $|B| \sim 201$~G), which changes into a stronger, non-axisymmetric, dipole-dominated field (mean $|B| \sim 276$~G) over a few rotation cycles, is in contrast to results from similar fully convective M-dwarf stars. YZ Cet is known to exhibit polarized radio bursts potentially driven by auroral radio emission from star--planet interaction (SPI). By combining our magnetic maps with recent radio observations, we refine the constraints on the magnetic field strength of the innermost planet, YZ Cet b. These results underscore the importance of monitoring stellar magnetic variability to interpret multi-wavelength SPI signatures and to characterize the magnetospheres of potentially habitable exoplanets.

Figures (11)

• Figure 1: Timeseries of LSD profiles of YZ Cet derived from the 2024 SPIRou observations. The three panels show (a) Stokes I, (b) Null N, and (c) Stokes V, respectively. The profiles are shifted vertically by an offset corresponding to their rotation phases according to the ephemeris given in the text. The vertical dashed line represent the mean radial velocity obtained from the LSD I profiles. For panel (b) and (c), the shaded regions represent the integration limits while determining the longitudinal fields.
• Figure 2: GRP fit: (a) Gaussian regression process fit to the SPIRou data. The red data in the top column correspond to the longitudinal field strength in gauss. The blue line represent the best fit, with the shaded region as the fit uncertainty. The bottom column represent the residual, defined as $\chi$ = residual/error. The fit has a reduced $\chi^2$ of $\approx1.1$. All data points are within $2\sigma$ limit of the GRP fit. (b) The posterior density distribution resulting from MCMC analysis of GRP model of $B_{\ell}$. The contours represent the $1\sigma$, $2\sigma$, and $3\sigma$ levels. The black solid lines mark the median values of the posterior distribution function (PDF). The black dashed lines represent the 16 per cent and 84 per cent percentiles of the PDF.
• Figure 3: Phase folded $B_{\ell}$ variation of the 2023 SPIRou observations (in blue), 2024 SPIRou observations (in red), ESPaDOnS observations (in orange), and NARVAL observation (in green). For the 2024 observations, we fit a single sinusoid with the rotational frequency, while for the 2023 SPIRou observations we fit an additional first harmonic. All data were folded with the same ephemeris, as described in the text.
• Figure 4: Comparison of SPIRou spectra between 2023 (blue) and 2024 (red) observations. (a) A small wavelength interval of Stokes I and Stokes V (shifted) spectra near rotation phase $\phi_{\rm rot} \sim0$ showing the change in Stokes V amplitude over time. Both spectra have comparable SNR. (b) Comparison of LSD Stokes I spectra of 2023 and 2024 observations. The difference between individual observed LSD Stokes I profiles and the median observed profile is shown (shifter by 0.88) to illustrate the lack of significant rotational variability, and to justify the assumption of uniform brightness for further ZDI analysis. The vertical dashed line represents the velocity range for equivalent width (EW) calculation. (c) Rotational variation of EW and radial velocity (RV) in the 2023 and 2024 observations. The mean EW for the 2023 and 2024 observations are shown as blue and red dashed horizontal dashed lines, respectively. In the bottom panel, the mean RV is shown with black dashed line.
• Figure 5: ZDI model fits (red curve) to the Stokes I (a) and Stokes V (b) LSD profiles (in black) from the 2024 SPIRou observations of YZ Cet. Rotation cycles for each observation are labeled on the right of respective profiles. In both cases, the profiles are shifted vertically for clarity.