Rossiter–McLaughlin Measurements

Updated 26 February 2026

Rossiter–McLaughlin measurements are spectroscopic techniques that detect transient Doppler anomalies during transits to infer spin–orbit alignments.
They leverage high-cadence time-series spectroscopy and methods like cross-correlation, iodine-cell modeling, and Doppler tomography to extract precise radial velocities.
Advanced approaches such as the reloaded RM method and tomographic techniques enhance the measurement of differential rotation and system dynamics even in active stellar environments.

The Rossiter–McLaughlin (RM) effect is a spectroscopic phenomenon observed when an eclipsing object, such as an exoplanet or a stellar companion, transits a rotating star and selectively blocks Doppler-shifted regions of the stellar photosphere. This transient perturbation in the disk-integrated spectral line profile provides a direct measurement of the sky-projected angle (λ) between the orbital axis of the companion and the rotation axis of the primary, thus serving as a powerful empirical probe of spin–orbit alignment, system architecture, migration history, and tidal evolution. RM measurements have become a cornerstone technique not only in exoplanetary astrophysics but also in binary-star studies and stellar rotation diagnostics.

1. Physical Basis and Mathematical Formulation

During a transit, the companion (planet or star) occults successive regions of its host’s rotating surface. A rotating star possesses a Doppler-broadened line profile, with one limb blueshifted and the other redshifted. The selective occultation removes a locus of velocity from the disk-integrated light, momentarily shifting the centroid of the stellar absorption line and creating the RM anomaly: a characteristic “bump” or “wiggle” in the radial velocity (RV) curve during occultation.

A general, limb-darkened analytic expression for the instantaneous radial-velocity anomaly is: $\Delta v(t) = -\left(\frac{R_p}{R_*}\right)^2 (v\sin i_*) \frac{\int_{\rm occulted} I(\mathbf{r})\,(\mathbf{r} \cdot \hat{x})\,d^2r}{\int_{\rm disk} I(\mathbf{r})\,d^2r}$ Here, $R_p$ and $R_*$ are the planet and stellar radii, $v\sin i_*$ is the sky-projected equatorial rotational velocity, $I(\mathbf{r})$ the surface brightness (including limb darkening), and $\hat{x}$ the projected rotation axis direction (Jenkins et al., 2010, Knudstrup et al., 2024). The shape and amplitude of $\Delta v(t)$ encode both the projected spin–orbit angle (λ) and system geometry.

For a linear or quadratic limb-darkening law ( $I(\mu) = 1-u_1(1-\mu)-u_2(1-\mu)^2$ ), closed-form analytic models (e.g., Ohta, Taruya & Suto 2005; Hirano et al. 2011) are widely used, and are generally robust over a large range of rotation rates, impact parameters, and planet-to-star size ratios (Jenkins et al., 2010, Boué et al., 2012, Knudstrup et al., 2024).

2. Instrumental Techniques and Data Reduction

Modern RM measurements are fundamentally high-cadence, high-S/N spectroscopic time-series obtained during transits or eclipses. Key observing approaches include:

Cross-correlation Function (CCF) RVs: As implemented in HARPS/ESPRESSO/HARPS-N, RVs are extracted by cross-correlation against a numerical mask (Knudstrup et al., 2024, Grouffal et al., 2022).
Iodine-cell Forward Modeling: Used in Keck/HIRES, spectral regions containing the iodine cell are forward modeled to derive precise RVs (Boué et al., 2012, Albrecht et al., 2011).
Tomographic Techniques: Doppler tomography directly maps the time-resolved line profile (“Doppler shadow”) induced by the transiting body, bypassing centroid-based RV extraction and minimizing parameter covariance (Miller et al., 2010, Brown et al., 2016, Knudstrup et al., 2024).
Reloaded RM: By subtracting appropriately scaled in-transit spectra from a master out-of-transit template, spatially localized line profiles from the occulted region of the star can be recovered, enabling direct measurement of local velocity fields, differential rotation, and spot crossing anomalies (Bourrier et al., 2016, Cegla et al., 2016).

Instrument-specific corrections for instrumental broadening, macroturbulence, and spectrograph stability are critical, and the choice of analytic formula for RM modeling must match the data reduction strategy to avoid biases in both $v\sin i_*$ and λ (Boué et al., 2012, Brown et al., 2016).

3. Statistical Inference and Degeneracies

Joint inference of system parameters employs simultaneous fitting of the photometric transit and spectroscopic RM anomaly, typically via Markov Chain Monte Carlo (MCMC). Parameters commonly varied include orbital period (P), mid-transit time (T₀), $R_p/R_*$ , a/R_*, inclination (i), λ, and $v\sin i_*$ , with limb-darkening coefficients either fixed or weakly constrained by priors (Knudstrup et al., 2024, Brown et al., 2012). Marginalized posteriors provide robust confidence intervals.

Degeneracies between λ and $v\sin i_*$ are particularly severe at low impact parameter (central transits), where only the product $v\sin i_* \cos\lambda$ is tightly constrained (Narita et al., 2010, Albrecht et al., 2011). Use of informative spectroscopic priors on $v\sin i_*$ helps, but may lead to bias if spectroscopic and RM-inferred values differ due to differential rotation or convective effects.

The inclusion of convective blueshift and explicit modeling of spot crossings (via photometric monitoring or iterative spectral cleaning) are essential for robust λ estimates on active stars, as RM waveforms can vary by up to 40° transit-to-transit due to surface activity (Oshagh et al., 2018, Palle et al., 2020).

4. Expanded Methodological Capabilities: Tomography, Reloaded RM, and Differential Rotation

Doppler tomography and the reloaded RM method have advanced the field by enabling:

Measurement of both λ and the true 3D obliquity (ψ), when the stellar inclination (i_*) is constrained (e.g., by rotation period, asteroseismology, or MHD modeling) (Cegla et al., 2016, Bourrier et al., 2016, Knudstrup et al., 2024).
Direct detection or constraints on differential rotation via mapping of the transit chord velocity residuals (Bourrier et al., 2016, Cegla et al., 2016).
Unbiased λ and $v\sin i_*$ even in the presence of strong contrast/FWHM variations across the transit chord (Bourrier et al., 2016).
Systematic comparison of analytic models (Hirano/Boué) and tomographic approaches, highlighting systematic underestimation of $v\sin i_*$ in some CCF-based models but robust λ recovery across techniques (Brown et al., 2016).

These methods are favored in rapidly rotating systems and yield the tightest formal uncertainties for both λ and $v\sin i_*$ .

5. Astrophysical Results and Trends from RM Measurements

As of 2024, over 200 RM measurements have been published across transiting exoplanetary and binary-star systems (Knudstrup et al., 2024). Ensemble studies reveal:

Hot Jupiters around cool stars ( $T_{\rm eff} \lesssim 6100\,K$ ): Intrinsic λ dispersion of only $1.4 \pm 0.7^\circ$ , much less than the Solar obliquity, consistent with efficient tidal realignment (Knudstrup et al., 2024, Mancini et al., 2018).
Hot Jupiters around hot stars ( $T_{\rm eff} \gtrsim 6250\,K$ ): Broad range of λ including retrograde/polar orbits, with λ–separation and λ–eccentricity correlations; tidal damping timescales are orders of magnitude longer, preserving primordial misalignments (Knudstrup et al., 2024).
Sub-Saturn and polar-orbit populations: Tentative but not statistically robust peaks at |λ| ≈ 90°, especially for specific mass/stellar-type regimes (Knudstrup et al., 2024).
Multi-planet systems: Low-λ systems dominate, supporting disk-driven alignment; rare misaligned multiples are associated with strong perturbations by outer companions (Knudstrup et al., 2024).
Long-period and low-mass planets: Extension of the RM technique to Neptune-mass and long-period planets (e.g., HIP 41378 d, λ ≈ 57°) demonstrates feasibility for PLATO-era targets, pending moderate stellar rotation and sufficient SNR (Grouffal et al., 2022).
Binary stars: Eclipsing binary RM surveys indicate general spin–orbit alignment in circular, short-period systems, with strong misalignments linked to perturbations from tertiary companions (Lidov–Kozai cycles) or preserved primordial orientation in wide/eccentric binaries (Sybilski et al., 2018).

6. Limitations, Special Cases, and Future Prospects

Stellar activity: Starspots and plages introduce significant systematic errors in RM modeling, necessitating multi-epoch or contemporaneous photometry to disentangle true spin–orbit signals from activity-induced anomalies (Oshagh et al., 2018, Palle et al., 2020).
Gravitational microlensing: For massive, distant planets and brown dwarfs, microlensing by the transiting body can partially “refill” the occulted region, reducing the RM anomaly amplitude by ≲1 m/s; this is negligible for hot Jupiters but relevant for sub-m/s precision in extreme systems (Oshagh et al., 2013).
Secondary eclipse RM (RMse): The RMse effect during secondary eclipse provides direct measurement of planetary rotation and obliquity, extending velocity mapping to planetary surfaces. Requires extreme SNR, large-aperture telescopes, and is most feasible in the NIR for highly irradiated planets (Nikolov et al., 2015).
Chromatic RM: Wavelength dependence of the RM amplitude enables atmospheric absorption studies; in the regime of rapidly rotating late-M dwarfs, atmospheric features in the RM effect can exceed 5–10 m/s and be isolated from white-light RM baselines (Cloutier et al., 2016).

With new high-precision spectrographs (e.g., ESPRESSO, future ELTs) and continuous input from missions like TESS and PLATO, the RM technique is poised to measure spin–orbit architectures for smaller, cooler planets and sub-Saturns, achieving comprehensive mapping of system obliquity distributions and enabling powerful constraints on disk–star coupling, migration pathways, and long-term dynamical evolution (Knudstrup et al., 2024, Grouffal et al., 2022).

References

Key technical and population-level results are found in (Jenkins et al., 2010, Knudstrup et al., 2024, Bourrier et al., 2016, Oshagh et al., 2018, Boué et al., 2012, Miller et al., 2010, Albrecht et al., 2011, Brown et al., 2016, Grouffal et al., 2022, Palle et al., 2020, Oshagh et al., 2013, Sybilski et al., 2018), and (Nikolov et al., 2015), among others.