🔭 astrophysics

The Impact of Planetary Phase Functions on Exo-Earth Detectability with EXOSIMS

This study utilizes the EXOSIMS simulator to demonstrate that incorporating a realistic, non-Lambertian Earth phase function and varying coronagraph inner working angles significantly alters exo-Earth yield estimates and detection distributions for the proposed Habitable Worlds Observatory, highlighting the necessity of accurate physical modeling in mission design.

Original authors: Searra Foote, Tyler D. Robinson, Rhonda Morgan, Dmitry Savransky, Mario Damiano, Armen Tokadjian

Published 2026-07-10

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Searra Foote, Tyler D. Robinson, Rhonda Morgan, Dmitry Savransky, Mario Damiano, Armen Tokadjian

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

Imagine you are a space detective trying to find a needle in a haystack, but the needle is a tiny, glowing Earth orbiting a blindingly bright star. To find it, you need a special pair of sunglasses (a coronagraph) that blocks the star's glare. But here's the tricky part: the "needle" doesn't shine with a steady light. It waxes and wanes like the Moon, changing its brightness depending on how much of its day side you can see. This changing brightness is called a phase function.

For a long time, scientists designing the future Habitable Worlds Observatory (HWO)—a massive space telescope planned for the 2040s—have been using a very simple, old-fashioned rule to guess how bright these planets are. They used a model called the Lambert phase function. Think of this like assuming every planet is a perfect, matte-white billiard ball that scatters light evenly in all directions. It's a handy, easy-to-use shortcut, but as this new study by Searra Foote and her team suggests, it's not quite right for a real Earth.

The "Glint" in the Crescent

Real Earth isn't a matte ball. It has oceans that act like mirrors and clouds that scatter light in specific directions. The researchers took a high-fidelity, realistic model of Earth's actual brightness (based on data from the Virtual Planetary Laboratory) and plugged it into EXOSIMS, a powerful computer simulator that predicts how many planets a mission might find.

They ran a simulation comparing the old "billiard ball" model against the "real Earth" model. Here is what they found:

The Old Model Missed the Glow: The simple Lambert model underestimated how bright Earth gets when it's in a "crescent" shape (when we see mostly its night side, but the sun is glinting off the oceans and clouds). In reality, Earth can be more than twice as bright at these angles than the old model predicted.
The "Where" Changed, Not Just the "How Many": Because the real Earth is brighter at these crescent angles, the simulation showed that the telescope would spot these planets at different times in their orbits. Instead of mostly finding them when they are at a "quarter" phase (like a half-moon), the realistic model suggested the telescope would find them more often at medium to high phase angles (closer to the crescent shape).
The Total Count Stayed Roughly the Same: This is the most important twist. Even though the timing and location of the detections shifted, the total number of planets found didn't skyrocket. In their simulation of 100 different mission runs, the old model found an average of 20 ± 2 planets, while the realistic model found 20 ± 1.5. The paper suggests that while the distribution of where we find them changes, the overall "yield" (the total count) only sees a modest change.

The Size of the Sunglasses Matters

The study also played with the size of the telescope's "sunglasses," known as the Inner Working Angle (IWA). This is the smallest distance from the star where the telescope can see a planet.

Big Sunglasses (Large IWA): If the telescope can't look too close to the star, it misses the inner parts of the orbit. In this case, the difference between the "billiard ball" and "real Earth" models is less noticeable because the telescope can't see the angles where the real Earth glows the brightest.
Small Sunglasses (Small IWA): If the telescope has a sharper eye and can look closer to the star, it opens up a wider range of orbital phases. The paper found that with a smaller IWA, the realistic Earth model makes a bigger difference, allowing the telescope to access a broader range of angles where the planet is intrinsically brighter.

What This Means for the Future

The authors are careful to note that this is a simulation, not a final discovery. They didn't find a new planet; they just updated the math used to predict how many we might find.

They also ruled out the idea that every Earth-like planet is exactly like our modern Earth. In their simulation, they treated all targets as "Earth twins" to isolate the effect of the brightness model. In reality, the universe is likely full of diverse worlds with different clouds, oceans, and atmospheres, which would make the picture even more complex.

However, the study strongly suggests that if we want to design the best possible mission for the 2040s, we can't rely on the simple "billiard ball" math anymore. We need to use realistic models that account for the glint of oceans and the scatter of clouds. This won't necessarily double the number of planets we find, but it will tell us when and where to look, ensuring we don't miss a glowing crescent Earth just because our math assumed it was dim.

In short: The universe is more complex and colorful than our old, simple models let on, and updating our "flashlight" math helps us plan a better hunt for our cosmic neighbors.

Technical Summary: The Impact of Planetary Phase Functions on Exo-Earth Detectability with EXOSIMS

Problem Statement
The development of the Habitable Worlds Observatory (HWO) requires accurate yield modeling to predict the number of Earth-like exoplanets a mission can detect and characterize. A critical component of these models is the planetary phase function, which describes how a planet's reflected brightness varies with the phase angle (the star-planet-observer angle). Current yield tools, such as the Exoplanet Open-Source Imaging Mission Simulator (EXOSIMS), typically rely on a simplified Lambertian phase function. This model assumes isotropic scattering and fails to capture the complex, anisotropic scattering behaviors of real planetary atmospheres and surfaces. Specifically, the Lambertian model underestimates Earth's brightness at phase angles beyond 90 degrees (crescent phases), where specular reflection from oceans and forward scattering by clouds significantly enhance reflectivity. This discrepancy raises concerns that standard yield estimates may inaccurately represent the detectability and optimal observation windows for Earth-like worlds.

Methodology
To address these limitations, the authors integrated a high-fidelity, observationally derived Earth phase function into EXOSIMS. The new phase function data was sourced from the Virtual Planetary Laboratory Three-Dimensional Spectral Earth Model (Robinson et al. 2010), which accounts for wavelength-dependent atmospheric scattering, cloud coverage, and surface specular reflection. The study employed a controlled experimental design:

Simulation Setup: Two sets of Monte Carlo simulations were run using EXOSIMS with identical mission parameters (6-meter aperture, 5-year mission life, specific star catalogs, and planet populations) but differing only in the phase function used: one using the default Lambertian model and one using the realistic, V-band integrated Earth phase function.
IWA Variation: The authors further investigated the interplay between phase functions and the coronagraph Inner Working Angle (IWA). Simulations were conducted across multiple IWA values (2, 2.5, 3, and 3.5 $\lambda/D$ ) to determine how instrument constraints affect the accessibility of different orbital phases.
Analysis: The study compared detection counts, phase-angle distributions at the time of detection, and probability distributions of total yield between the two phase function models.

Key Results

Redistribution of Detections: The primary finding is that the realistic Earth phase function systematically redistributes detections toward medium and high phase angles (beyond 90 degrees). In these regions, the realistic Earth model is intrinsically brighter than the Lambertian approximation. The simulations showed an ensemble-averaged increase of approximately 20% in detection events occurring beyond 90 degrees when using the realistic model.
Modest Impact on Total Yield: Despite the significant shift in where detections occur, the total number of detected exo-Earths remained relatively stable. The median yield for the Lambertian case was $20 \pm 2$ detections per mission, compared to $20 \pm 1.5$ for the realistic case. The authors attribute this to the non-linear relationship between brightness and exposure time; while brightness differences are large at extreme phases, the fractional changes in the planet-star flux ratio in the detectable regime are often small enough that the overall mission yield does not change dramatically.
Role of Inner Working Angle (IWA): The study demonstrated that IWA and phase function assumptions jointly determine detectability. Smaller IWAs expand the accessible range of projected separations, allowing the telescope to observe planets at a broader range of orbital phases, including those closer to the star. This expanded access allows the brightness advantages of the realistic phase function at crescent phases to influence a larger portion of the detection sample. Conversely, larger IWAs restrict observations to wider separations, suppressing the impact of the realistic brightness model at high phase angles.

Significance and Claims
The paper argues that while the total yield estimate for an HWO-like mission may not drastically change with the adoption of realistic phase functions, the distribution of detections is significantly altered. This has practical implications for mission planning and follow-up characterization:

Observation Timing: Realistic phase functions shift detections toward phases where planets are intrinsically brighter but may be geometrically more challenging (e.g., crescent phases). This affects the scheduling of observations and the planning of revisit epochs for spectroscopic characterization.
Mission Design: The results highlight the necessity of incorporating physically motivated phase functions into yield tools to accurately predict not just how many planets will be found, but when and where they will be detectable.
Future Refinement: The authors emphasize that while their study uses "Earth twins" as a simplification, the methodology establishes a framework for incorporating diverse planetary environments. They conclude that realistic phase functions and precise IWA analysis are essential steps toward making yield models more physically grounded for future direct imaging missions.

The "Glint" in the Crescent

The Size of the Sunglasses Matters

What This Means for the Future

More like this