🔭 astrophysics

A Candidate Innermost Fifth Planet In the HR 8799 System Revealed By JWST NIRISS Aperture Masking Interferometry

Using a new JWST/NIRISS Aperture Masking Interferometry pipeline that overcomes previous sensitivity limitations, researchers have detected a candidate fifth planet in the HR 8799 system, which, if confirmed, would make it the first directly imaged five-planet system and suggest a stable 3:1 mean motion resonance with the known planet e.

Original authors: Jayke S. Nguyen, Louis Desdoigts, Alexandra Z. Greenbaum, Benjamin J. S. Pope, Max Charles, Quinn M. Konopacky, Kaitlyn Hessel, Dori Blakely, Matthew De Furio, Clarissa R. Do Ó, René Doyon, Doug Johns

Published 2026-09-10

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Jayke S. Nguyen, Louis Desdoigts, Alexandra Z. Greenbaum, Benjamin J. S. Pope, Max Charles, Quinn M. Konopacky, Kaitlyn Hessel, Dori Blakely, Matthew De Furio, Clarissa R. Do Ó, René Doyon, Doug Johnstone, Jens Kammerer, David Lafrenière, Bruce A. Macintosh, Michael R. Meyer, Eric L. Nielsen, Anne E. Peck, William Roberson, Anand Sivaramakrishnan, Peter Tuthill, Thomas Vandal, Marie Ygouf

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

For decades, astronomers have been captivated by the idea of finding other worlds like our own, but the most difficult ones to spot are often the closest to their stars. This is because the glare of the star itself is so overwhelming that it washes out the faint light of any nearby planets, much like trying to see a firefly hovering next to a searchlight. To solve this, scientists have developed techniques to block the star's light or to use the wave-like nature of light itself to sharpen their view, allowing them to peer into the crowded inner regions of distant solar systems. One of the most promising systems for this kind of study is HR 8799, a young star located about 130 light-years away. It is already famous for hosting four massive gas giant planets that orbit far from their sun, a configuration that has challenged our understanding of how planetary systems form and survive. However, for years, astronomers have suspected that a fifth, smaller planet might be hiding much closer to the star, tucked away in the glare where previous telescopes could not look.

A team of researchers has now turned a powerful new eye toward this system using the James Webb Space Telescope, and they have found a strong candidate for that missing inner world. By using a specialized instrument on the telescope called the Near Infrared Imager and Slitless Spectrograph, the team employed a technique known as aperture masking interferometry. This method involves placing a mask with seven small holes over the telescope's mirror, which splits the incoming light into multiple beams. When these beams recombine, they create an interference pattern that reveals details far smaller than the telescope could normally see. The researchers were looking for a faint companion just a tiny fraction of a degree away from the star, a distance so small that it corresponds to a physical separation of only about seven times the distance between the Earth and the Sun.

The team's search was made possible by a new computer program they developed called Amigo, which acts as a highly sophisticated simulator of the telescope's entire optical system. Previous attempts to analyze data from this instrument were limited by subtle electronic glitches in the detector that distorted the delicate interference patterns, making it impossible to see faint objects near bright stars. The new software models these glitches and corrects for them, effectively cleaning up the image to reveal what was previously hidden. When the researchers applied this new method to the data, they found a distinct signal just north of the star. This signal appeared at a contrast level of about two parts in ten thousand, meaning the object is roughly 50,000 times fainter than the star itself. The object sits at a projected separation of about 150 milliarcseconds, which translates to a physical distance of roughly seven astronomical units from the star.

The researchers interpret this signal as a candidate planet, which they have named HR 8799 f. Based on its brightness and the known age of the system, this object likely has a mass somewhere between a few and several times the mass of Jupiter. While the detection is statistically significant, appearing clearly above the noise level of the instrument, the team describes it as tentative because it relies on a single observation. The object's position is consistent with the movement of the star as measured by other space missions, suggesting it is indeed a bound companion and not a background star. Furthermore, its location fits neatly into a theoretical model where the five planets would be locked in a stable orbital dance, with the new planet orbiting in a specific resonance with the innermost of the four known giants. This resonance would act as a gravitational anchor, preventing the planets from colliding or being ejected from the system over billions of years.

Despite the excitement, the authors are careful to note that this is not yet a confirmed discovery. The signal lies just above the threshold of what the instrument can reliably detect, and there are other faint ripples in the data that could potentially be false alarms, though the team's analysis suggests these are likely artifacts of the instrument rather than real objects. The team also compared their findings with previous searches from ground-based telescopes, which had failed to find this planet. They suggest that those earlier instruments simply lacked the sensitivity to see an object this close to the star, or that the planet's brightness varies in a way that made it invisible at the wavelengths those telescopes used. To turn this candidate into a confirmed discovery, the researchers plan to observe the system again with the same instrument, using multiple filters and slightly shifting the telescope's aim to separate the planet's light from any remaining instrumental noise. If confirmed, HR 8799 f would become the first directly imaged five-planet system, offering a rare glimpse into the complex architecture of a young solar system and providing a crucial test for theories of planetary formation.

Technical Summary: A Candidate Innermost Fifth Planet in the HR 8799 System

Problem and Motivation
The HR 8799 system is a benchmark directly imaged multi-planet system known to harbor four giant exoplanets (b, c, d, e) orbiting in a likely Laplace mean motion resonance (MMR). Theoretical models and dynamical stability analyses have long speculated the existence of a fifth, innermost planet (HR 8799 f) to complete a resonant chain (e.g., a 3:1 or 2:1 resonance with planet e). However, previous ground-based observations and earlier analyses of James Webb Space Telescope (JWST) data have failed to confirm this candidate. The primary challenge lies in the extreme contrast requirements ( $\sim 10^{-4}$ ) needed to detect a few-to-several Jupiter-mass object at a projected separation of $\sim 7$ au ( $\sim 150$ mas), which falls within the inner working angle of conventional direct imaging and is obscured by detector systematics in Aperture Masking Interferometry (AMI).

Methodology
The authors present a detection of a candidate fifth planet using JWST/NIRISS AMI data from the Guaranteed Time Observation (GTO) program 1200, taken on August 3, 2023, in the F380M filter. The core methodological advancement enabling this detection is the application of Amigo (Aperture Masking Interferometry Generative Observations), a new end-to-end forward-modeling data processing pipeline.

Forward Modeling: Unlike traditional pipelines that extract closure phases and suffer from Brighter-Fatter Effect (BFE) systematics, Amigo models the complete optical and electronic system from photons to pixels. It utilizes a differentiable framework ( $\partial$ Lux) to simultaneously fit optical aberrations (parameterized by Zernike modes) and detector response (including non-linear gains and BFE dynamics via a neural network).
Data Reduction: The pipeline operates on raw Stage 0 data, inferring the telescope Optical Path Difference (OPD) state and astrophysical source properties (position, flux, complex visibility) jointly. It employs Bayesian inference and Hamiltonian Monte Carlo (HMC) to derive posterior distributions.
Observables: The analysis uses DISCO (Delay-Insensitive Subspace of Calibrated Observables) self-calibrated combinations of visibility amplitudes and phases to suppress first-order optical aberrations while retaining astrophysical information.
Orbit Fitting: To validate the candidate, the authors combined the new JWST relative astrometry with absolute astrometry from Gaia and Hipparcos, as well as historical relative astrometry, using the Octofitter package. They tested consistency with a stable five-planet resonant model (Va solution) from previous literature.

Key Results

Detection of HR 8799 f: The analysis reveals a statistically significant signal ( $\Delta \ln \mathcal{L} \approx 8$ ) at a projected separation of $\sim 150$ mas ( $\sim 7$ au) from the host star. The candidate lies just above the $3\sigma$ contrast curve with a contrast of $\sim 2 \times 10^{-4}$ ( $\Delta mag \approx 9.14$ ) in the F380M filter.
Astrometric Consistency: The recovered position of the candidate is consistent with absolute proper motion constraints from Gaia and Hipparcos. When fitting a five-planet system, the candidate's position aligns with a stable orbital solution in a 3:1 MMR with planet e, though the authors note this alignment is circumstantial rather than definitive proof of resonance.
Mass and Photometry: The orbital fit yields a median mass for the candidate of $\sim 1.7 M_{Jup}$ (with a 68% credible interval extending to $\sim 3 M_{Jup}$ ), which is lower than the $7-9 M_{Jup}$ predicted by some stability models but consistent with a $3\sigma$ upper limit from Gaia-Hipparcos analysis. The photometry is consistent with atmospheric models for a young, low-mass planet, though limited by single-band data.
Systematic Mitigation: The detection is attributed to the Amigo pipeline's ability to reach photon-noise-limited performance by correcting for BFE and other systematics that limited previous AMI analyses to contrasts $> 10^{-3}$ .

Significance and Claims
The paper claims that if confirmed, HR 8799 f would represent the first directly imaged five-planet system, providing critical insights into the orbital dynamics and formation of systems with widely separated gas giants. The authors emphasize that the detection is tentative and statistically significant but not yet definitive confirmation.

Caveats: The authors explicitly state that the candidate's position is near a resonant solution but do not claim the object is definitively in resonance. They acknowledge discrepancies in the astrometry of other planets (d and e) which may reflect residual detector calibration uncertainties due to the single-dither observation strategy.
Future Work: The paper asserts that confirmation requires future observations, specifically multi-filter NIRISS AMI data with subpixel dithering to constrain the spectral energy distribution and common proper motion. The authors note that the object is currently too faint for NIRSpec or NIRCam but may be accessible to ground-based GRAVITY+.
Technological Impact: The study demonstrates that differentiable forward modeling can extend AMI sensitivity to previously inaccessible parameter spaces (high contrast, small separation), potentially revolutionizing the search for inner companions in directly imaged systems.

The authors conclude that the detection of this candidate is only possible through the combination of JWST/NIRISS AMI capabilities and the novel Amigo forward-modeling approach, marking a step toward probing the innermost regions of planetary systems at solar-system scales.

More like this