🔭 astrophysics

A New Strategy for Using Spectroscopic Phase Curves to Characterize Non-Transiting Planets

This paper introduces the Variable Planetary Infrared Excess (VPIE) method, a novel strategy that uses empirical stellar modeling to extract spectroscopic phase curve information from non-transiting exoplanets around M-dwarfs, enabling the characterization of their atmospheric circulation and radii without requiring transits or stellar spectral models.

Original authors: Ted M. Johnson, Avi M. Mandell

Published 2026-08-25

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ted M. Johnson, Avi M. Mandell

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 two decades, astronomers have been able to study the atmospheres of distant worlds, but only if those worlds happen to pass directly in front of their host stars from our perspective. When a planet transits, it blocks a tiny fraction of the star's light, allowing scientists to analyze what that light reveals about the planet's air. However, this cosmic alignment is rare. Most planets orbiting the nearest stars do not cross in front of them, leaving them invisible to these traditional methods. For the vast majority of nearby worlds, particularly those orbiting small, cool red stars, we have been unable to tell if they possess an atmosphere, let alone what that atmosphere might be made of. This limitation has left a huge gap in our understanding of the most common type of planet in our galaxy.

To bridge this gap, researchers Ted M. Johnson and Avi M. Mandell have proposed a new strategy that does not rely on a planet passing in front of its star. Instead of waiting for a shadow, they aim to listen for the faint, rhythmic glow of the planet itself as it orbits. Every planet emits heat, and as it spins, the side facing its star gets hot while the side facing away cools down. This creates a changing signal in the combined light of the star and planet. The challenge is that the star is overwhelmingly bright and often flickers on its own due to surface spots and storms, drowning out the tiny planetary signal. The new method, called the Variable Planetary Infrared Excess, or VPIE, is designed to filter out the star's noise to reveal the planet's whisper.

The researchers tested this idea by creating detailed computer simulations of observations that could be made with the James Webb Space Telescope. They focused on three specific planets orbiting nearby red stars: a warm giant planet named TOI-519 b, a smaller warm world called GJ 876 d, and the potentially habitable Earth-sized planet Proxima Centauri b. In their simulations, they fed the telescope data a mix of light from the star and the planet, complete with the kind of stellar flickering that occurs in reality. The core of their new technique is a mathematical process that treats the star's light as a pattern. By analyzing the star's behavior at shorter wavelengths of light—where the planet is too cold to emit any signal—the method builds a model of how the star is changing over time. It then uses this model to predict what the star should look like at longer, infrared wavelengths.

When the researchers subtracted this predicted stellar light from the actual data, they were left with a residual signal. In their simulations, this leftover signal contained the distinct imprint of the planet's changing temperature as it orbited. For the warm giant planet TOI-519 b, the method successfully distinguished between different atmospheric scenarios. It could tell the difference between a planet that efficiently moves heat from its day side to its night side and one that does not. It also provided a way to estimate the planet's size, even without a transit. The results showed that for larger, hotter planets, the technique is robust enough to rule out extreme scenarios, such as a planet with no atmosphere at all or one with a perfectly uniform temperature.

The test became more difficult when the researchers applied the method to the smaller, cooler worlds. For the sub-Neptune GJ 876 d, the simulation showed that the technique could still separate the planet's signal from the star's noise, allowing them to determine if the planet had a thick atmosphere or was a bare rock. However, when they simulated observations of Proxima Centauri b, a temperate Earth-sized world, the current capabilities of the James Webb Space Telescope proved insufficient. The planet is so cold that its heat is emitted at wavelengths where the telescope's instruments are not sensitive enough to catch the faint signal against the background noise. The simulations indicated that while the method works in theory, the current hardware cannot yet detect the atmosphere of such a cool, small world.

The authors conclude that while the VPIE method is a powerful new tool for studying non-transiting planets, its success depends heavily on the instruments used. For the warmest and largest planets, the James Webb Space Telescope could begin to reveal the climates of worlds that were previously inaccessible. For the cooler, Earth-like planets that are the primary targets in the search for life, the method suggests that a new generation of telescopes is needed. The researchers propose that a future mission equipped with a specialized instrument capable of observing a wider range of infrared wavelengths simultaneously would be required to unlock the secrets of these temperate worlds. Their work provides a clear path forward: a mathematical strategy that can isolate the light of a hidden planet, waiting for the technology to catch up and make the invisible visible.

Technical Summary: A New Strategy for Using Spectroscopic Phase Curves to Characterize Non-Transiting Planets

Problem Statement
Current methods for characterizing exoplanet atmospheres, such as secondary eclipse and traditional phase curve analysis, rely on a transit geometry where the planet passes in front of or behind its host star. However, the vast majority of known planets orbiting nearby M-dwarfs—particularly those in the habitable zone—are non-transiting. The probability of a transit for a planet in the habitable zone of an M-dwarf is approximately 1%, leaving a significant population of potentially habitable worlds inaccessible to current atmospheric characterization techniques. Furthermore, existing methods for detecting non-transiting planets via thermal emission, such as the Planetary Infrared Excess (PIE) technique, rely heavily on physics-based stellar models to subtract the stellar flux. Recent studies indicate that even the best stellar models are only accurate to within ~1%, introducing systematic errors that exceed the planetary signal (often $10^{-5}$ to $10^{-4}$ of the stellar flux). Additionally, stellar variability (spots, faculae, and flares) introduces noise orders of magnitude larger than the planetary signal, complicating the extraction of planetary data.

Methodology: The VPIE Approach
The authors introduce the Variable Planetary Infrared Excess (VPIE) method, a time-series analysis strategy designed to extract planetary flux from combined star-planet light without requiring a transit or a physics-based stellar model.

The core of VPIE relies on the spectral separation of the observation into two regions:

Short Wavelength (SW) Region: Wavelengths where the planetary flux is negligible, and the signal is dominated by the star.
Long Wavelength (LW) Region: Wavelengths where the planetary thermal emission rises above the photon noise.

The method operates on the principle that stellar variability arises from "component mixing" (e.g., the changing visibility of star spots or granulation as the star rotates). Consequently, any observed stellar spectrum can be described as a linear combination of other spectra from the same time series. VPIE utilizes a small set of individual spectra (basis spectra) from the SW region to construct an empirical model of the stellar contribution at every time step.

The procedure involves:

Basis Selection: Selecting a set of basis spectra from the time series that minimizes the Bayesian Information Criterion (BIC) or Akaike Information Criterion (AIC).
Normalization: Using weighted least-squares fitting on the SW region to determine coefficients that reconstruct the stellar spectrum for each time step.
Residual Extraction: Applying these coefficients to the full spectrum (SW and LW) to generate a reconstructed stellar model ( $\tilde{f}$ ). Subtracting this model from the observed data ( $f$ ) yields residuals ( $f - \tilde{f}$ ). Because the fitting was performed only on SW data, the model is "agnostic" to the planet; thus, the LW residuals contain the imprint of the planetary flux ( $\delta$ ) plus noise ( $\epsilon$ ).
Inversion: Comparing the observed residuals against a grid of forward planetary models (varying radius and heat redistribution efficiency) to infer planetary properties using a $\chi^2_{red}$ metric.

Key Contributions and Simulated Results
The authors demonstrate the effectiveness of VPIE using simulated James Webb Space Telescope (JWST) observations of three distinct targets:

TOI-519 b (Warm Giant): A transiting warm giant around an M-dwarf.
- Result: VPIE successfully distinguishes between full, moderate, and zero heat redistribution scenarios. Even when the secondary eclipse is masked (simulating a non-transiting scenario), the method constrains the planet's radius and heat redistribution, ruling out scenarios with $R \gtrsim 0.5 R_{Jup}$ and $T_{night}/T_{day} < 0.8$ .
GJ 876 d (Warm Sub-Neptune): A known non-transiting planet.
- Result: The method distinguishes between a bare rock (zero heat redistribution) and a planet with an atmosphere (moderate redistribution). It breaks the degeneracy between small planetary radii and efficient heat redistribution, allowing constraints on the planet's interior composition (e.g., distinguishing rock/water-dominated planets from H2-rich atmospheres).
Proxima Centauri b (Temperate Rocky Planet): A potentially habitable, non-transiting Earth-sized planet.
- Result: Using JWST/MIRI LRS (5–11 $\mu$ m), VPIE provides limited constraints for Earth-sized planets due to low signal-to-noise ratios at these wavelengths. While volatile-rich (large radius) atmospheres are somewhat constrained, Earth-like compositions are difficult to distinguish from each other. The authors note that current JWST capabilities are insufficient for robust characterization of such cool targets.

Significance and Claims
The paper claims that VPIE offers a "promising new framework" for characterizing the population of non-transiting planets around nearby M-stars, which are otherwise inaccessible.

Model Independence: By using an empirical, data-driven approach to model stellar variability, VPIE avoids the systematic errors inherent in physics-based stellar models, which are particularly unreliable for M-dwarfs.
Scalability: The method is shown to be effective for giant planets and sub-Neptunes with current instrumentation (JWST).
Future Potential: While current performance for cooler, Earth-sized targets is limited by JWST's wavelength coverage and sensitivity, the authors suggest that future instrumentation (such as the proposed MIRECLE mission or a Far-IR Surveyor) with broader, simultaneous wavelength coverage (extending to 18 $\mu$ m or beyond) could enable the characterization of potentially habitable planets.

The authors conclude that while challenges remain regarding stellar complexity (flares, heterogeneous spots) and instrument saturation, VPIE represents a significant step toward a comprehensive census of planetary climates in the solar neighborhood, expanding the accessible parameter space beyond transiting systems.

More like this