Does HD 3167 Have Planets with Perpendicular Orbits?

arXiv:2609.19277 · astro-ph.EP · Submitted 2026-09-16 · Read on arXiv

Listen

Transcript

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: Next we'll be talking about the paper "Does HD 3167 Have Planets with Perpendicular Orbits?".

Jocelyn: The paper was written by Joshua N. Winn, Jack Lubin, Guðmundur Stefánsson, Haochuan Yu, Erik Petigura et al. from Department of Astrophysical Sciences, Princeton University and Department of Physics & Astronomy, University of California Los Angeles and Anton Pannekoek Institute for Astronomy, University of Amsterdam and Astrophysics & Space Center, Schmidt Sciences and Department of Physics, University of Oxford and Department of Astronomy, University of California at Berkeley and Department of Astronomy, California Institute of Technology and Institute for Astronomy, University of Hawaii.

Vera: Stay tuned as we take you through the paper and discuss its implications.

Jocelyn: We also have Subrahmanyan with us today — guest researcher.

Vera: Alright, let's get started.

Paper discussion segment 1: Vera: We’ve just pulled up a fascinating new preprint titled "Does HD three thousand one hundred sixty-seven Have Planets with Perpendicular Orbits?" from Joshua Winn and his large team of collaborators. It's a direct challenge to a previous claim that this system contains planets orbiting at ninety-degree angles to one another.

Jocelyn: That sounds like something straight out of science fiction, Vera. If they actually had perpendicular orbits, wouldn't that be a massive anomaly in how we think solar systems organize themselves?

Vera: It would be incredibly rare because most systems we see are relatively flat, like our own. The authors from Princeton and UCLA are essentially re-examining the data to see if that "perpendicular" claim holds water.

Jocelyn: So, instead of accepting the previous finding as gospel, they went out and gathered more data to check the math?

Vera: Exactly, they used the Keck Planet Finder to get a much clearer look at how these planets move across their star.

Subrahmanyan: You have to understand why this matters for my side of things, because a perpendicular system would break our standard models of how planets form from a flat disk of dust. If you see orbits like that, it implies some violent dynamical history, like one planet getting kicked into a weird angle by another.

Jocelyn: That's what I was wondering; does this paper actually suggest such a violent history happened there?

Subrahmanyan: Well, the authors aren't saying it did; they are actually casting doubt on whether that extreme geometry even exists. They found that the data is much more consistent with coplanar orbits.

Vera: It’s a classic case of needing more photons to settle a debate, and they certainly brought the heavy machinery to do it.

Jocelyn: It makes you wonder if the original detection was just an error caused by having too little data to work with.

Paper discussion segment 2: Vera: Now that we know they're questioning the architecture, let's look at what they actually found in their analysis of the Rossiter-McLaughlin effect. They looked at planet b, which is a super-Earth about one point seven times the size of Earth.

Jocelyn: How did they use that effect to determine if it was misaligned?

Vera: They look at how the planet blocks different parts of the rotating star during a transit, which creates a specific shift in the radial velocity. The previous study suggested planet b was aligned with the equator, but this new analysis favors a projected obliquity of negative sixty-six degrees, plus or minus about fifteen degrees.

Jocelyn: Wait, so they aren't seeing that low obliquity anymore?

Vera: No, they don't confirm it; instead, they find a much larger misalignment than the previous paper claimed.

Subrahmanyan: But even this new result is a bit of a mess because the best-fit rotation velocity for the star came out higher than expected based on other measurements. It’s not a clean "everything is flat" answer; it's more like "the geometry of the system is still unsettled."

Jocelyn: If the rotation speed doesn't match, does that mean their model for the planet's orbit might be flawed too?

Subrahmanyan: It certainly suggests there’s a tension in the data. When they left out one specific dataset—the most discrepant one—the uncertainty on the angle grew so much that you couldn't really say anything for sure.

Vera: It’s frustrating from an observational standpoint because you want a clean, definitive result, but the signal for this star is incredibly tiny.

Jocelyn: So they’ve essentially moved us from "definitely perpendicular" to "we actually have no idea"?

Paper discussion segment 3: Vera: This uncertainty is really stemming from how difficult it is to measure the rotation of a slow-moving star like HD three thousand one hundred sixty-seven. The team spent a lot of time comparing the line widths they saw in the KPF spectra to other stars to try and pin down that velocity.

Jocelyn: How do you even distinguish between the star spinning and other things, like turbulence in its atmosphere?

Vera: That’s the big challenge; they had to compare HD three thousand one hundred sixty-seven's cross-correlation functions against stars with known rotation speeds to get a handle on it. Their data suggests the rotation velocity is likely between zero point nine and one point four kilometers per second.

Jocelyn: And I'm guessing that doesn't match what the previous researchers found?

Vera: Right, Bourrier et al. had suggested a much higher rotation velocity of about two point four kilometers per second to make their "perpendicular" model work.

Subrahmanyan: This is where the theoretical implications get interesting because if the star is actually spinning slowly, then the previous model was likely forced into a weird solution to explain the data. The authors suggest that a line-profile analysis might be necessary to finally settle this, rather than just looking at radial velocity shifts.

Jocelyn: So, what is the actual path forward for someone trying to solve this mystery?

Vera: They're suggesting that if we want to find these truly weird perpendicular systems, we should probably look at stars that rotate much faster.

Subrahmanyan: That’s a very practical point; high rotation speeds make these signals much easier to disentangle from the background noise of the star.

Jocelyn: It sounds like they've essentially reset the clock on this system, leaving us with more questions than answers for now.

Conclusion: Vera: We've covered a lot of ground, starting with the massive claim of perpendicular orbits in "Does HD three thousand one hundred sixty-seven Have Planets with Perpendicular Orbits?" and ending up in a place of scientific caution. The new data doesn't support that extreme geometry, but it hasn't provided a perfect replacement measurement either.

Jocelyn: It’s a reminder that even when we think we have a discovery, more data can completely flip the script.

Vera: It really does; the system remains unsettled, with the possibility of coplanar orbits being much more likely than we once thought.

Subrahmanyan: I think this paper serves as an important lesson in the complexity of multi-planet systems and the limits of our current spectroscopic techniques. We need to be careful when interpreting these tiny signals in slow rotators.

Jocelyn: It's definitely not a closed case, but at least we're moving toward a more accurate picture of what's actually out there.

Vera: Exactly, and that’s why we keep watching the data. Thanks for joining us, everyone; we'll see you next time!

Subrahmanyan: Looking forward to the next one!

Jocelyn: Goodbye for now!

Vera: Bye! --- END OF SCRIPT ---

Department of Astrophysical Sciences, Princeton University · Department of Physics & Astronomy, University of California Los Angeles · Anton Pannekoek Institute for Astronomy, University of Amsterdam · Astrophysics & Space Center, Schmidt Sciences · Department of Physics, University of Oxford · Department of Astronomy, University of California at Berkeley · Department of Astronomy, California Institute of Technology · Institute for Astronomy, University of Hawaii

astro-ph.EP

Submitted: 2026-09-16

Updated: 2026-09-16

Comments: To appear in ApJ Letters (7 pages, 3 figures)

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 85/100

The gist: The authors investigated whether "the two transiting planets of HD 3167 were reported to follow nearly perpendicular paths, with the inner orbit aligned with the stellar equator and the outer orbit

Terminology

Keck Planet Finder — 5× in this paper
Accuracy — 1× in this paper · explained in 1 episode
Alignment — 1× in this paper · explained in 1 episode
Trace — 1× in this paper · explained in 2 episodes

Summary

The authors investigated whether the two transiting planets of HD 3167 were reported to follow nearly perpendicular paths, with the inner orbit aligned with the stellar equator and the outer orbit nearly polar. This interpretation previously depends critically on a challenging single-transit detection of the Rossiter-McLaughlin effect of the inner planet. To re-evaluate this, we observed three additional transits of the inner planet with the Keck Planet Finder and analyzed the new data together with two archival ESPRESSO transit datasets.

Regarding their findings, We do not confirm the previously reported low obliquity. Our analysis favors a projected obliquity of −66+14−12 deg, consistent with coplanar orbits. The researchers note that "the best-fit projected rotation velocity is higher than expected, and because the obliquity uncertainty grows substantially when the most discrepant of the five transit datasets is omitted, we regard the geometry of the HD 3167 system as still unsettled."

The paper further details that "the full dataset favors a large projected misalignment for planet b, closer to the measurement for planet c than the low-obliquity solution reported by Bourrier et al. (2021). This result is compatible with coplanar orbits. However, our analysis does not support the claim of perpendicular orbits, [nor] does it rule out a large mutual inclination. Ultimately, the most conservative conclusion is that the mutual inclination between the two transiting planets in the HD 3167 system cannot be robustly determined with the available data. Consequently, coplanar orbits are a reasonable default assumption."

Improvements for AI systems

To improve AI systems—specifically those used in automated scientific discovery, signal processing, and astrophysical modeling—based on the findings of this paper, I propose the following technical improvements:

Incorporate Model-Agnostic Sensitivity Analysis into Bayesian Inference Engines.
Implement Discrepancy-Aware Outlier Weighting in time-series regression models.
Develop Multi-Instrument Cross-Calibration Layers for heterogeneous data fusion.

Incorporate Model-Agnostic Sensitivity Analysis into Bayesian Inference Engines.

The paper demonstrates that a specific modeling choice (modeling CCF distortions vs. simple radial velocity) can lead to vastly different physical conclusions (perpendicular vs. coplanar orbits).

The improved AI system would not just output a posterior distribution, but would automatically run parallel simulations using alternative mathematical frameworks (e.g., switching from centroid-shift models to full line-profile models) to quantify model uncertainty.
This allows the AI to flag when a scientific conclusion is highly dependent on the chosen mathematical abstraction rather than the raw data.

Implement Discrepancy-Aware Outlier Weighting in time-series regression models.

The researchers found that omitting a single dataset (the second KPF transit) significantly altered the inferred stellar rotation velocity and obliquity, suggesting that outliers in scientific datasets are often not random noise but contain critical information about model inadequacy.

The improved AI system would utilize a Leave-One-Out (LOO) automated loop during training/fitting to identify which specific data points drive the most significant shifts in parameter space.
Instead of simply discarding these points as noise, the AI would categorize them as High-Influence Discrepancies, alerting researchers that the current physical model is likely missing a fundamental component (such as differential rotation or local line contrast).

Develop Multi-Instrument Cross-Calibration Layers for heterogeneous data fusion.

The paper highlights how combining datasets from different spectrographs (KPF and ESPRESSO) introduces systematic tensions due to varying instrumental profiles, resolutions, and zero-point offsets.

The improved AI system would implement a deep learning layer designed specifically for instrumental signature decoupling. This layer would learn the unique transfer functions of different sensors (e.g., the specific resolution and stability characteristics of KPF vs. ESPRESSO) to prevent instrumental artifacts from being misinterpreted as physical phenomena (like the Rossiter-McLaughlin effect).
This enables much more robust automated discovery when merging massive, heterogeneous datasets from global observatories.

Abstract

The two transiting planets of HD 3167 were reported to follow nearly perpendicular paths, with the inner orbit aligned with the stellar equator and the outer orbit nearly polar. This interpretation depends critically on a challenging single-transit detection of the Rossiter-McLaughlin effect of the inner planet. We observed three additional transits of the inner planet with the Keck Planet Finder and analyzed the new data together with two archival ESPRESSO transit datasets. We do not confirm the previously reported low obliquity. Our analysis favors a projected obliquity of-66+14-12 degrees, consistent with coplanar orbits. However, because the best-fit projected rotation velocity is higher than expected, and because the obliquity uncertainty grows substantially when the most discrepant of the five transit datasets is omitted, we regard the geometry of the HD 3167 system as still unsettled.