🔭 astrophysics

SIMTERFERE: An optical interferometry simulator for quantifying the coherent flux stability of VLTI/GRAVITY+. Reaching per mill stability: Application to exoplanet spectroscopy

The paper introduces SIMTERFERE, a data-driven simulation tool that analyzes seven-hour observations of beta Pictoris to demonstrate that the VLTI/GRAVITY+ instrument achieves the per-mill coherent flux stability required for high-fidelity exoplanet spectroscopy by effectively correcting instrumental throughput variations and telluric effects.

Original authors: J. R. Sauter, A. von Stauffenberg, G. Bourdarot, W. Brandner, F. Eisenhauer, L. Kreidberg, L. Labadie, S. Scheithauer, D. Trevascus, R. van Boekel

Published 2026-03-17

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: J. R. Sauter, A. von Stauffenberg, G. Bourdarot, W. Brandner, F. Eisenhauer, L. Kreidberg, L. Labadie, S. Scheithauer, D. Trevascus, R. van Boekel

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

The Big Picture: Catching a Firefly in a Hurricane

Imagine you are trying to take a photograph of a tiny, glowing firefly (an exoplanet) sitting right next to a blindingly bright spotlight (its host star). The firefly is millions of times dimmer than the spotlight. If you just point a camera at them, the spotlight will wash out the firefly completely.

To solve this, astronomers use a super-powerful tool called an interferometer (specifically, the GRAVITY+ instrument at the Very Large Telescope). Think of this instrument as a giant, high-tech pair of binoculars that combines light from four massive telescopes to create a "super-eye." This super-eye is so sharp it can separate the firefly from the spotlight.

However, there's a catch. To see the firefly's colors (its spectrum) and learn what its atmosphere is made of, you need to stare at it for hours. During those hours, the "camera" (the telescope) and the "air" (Earth's atmosphere) can wiggle, shake, and change brightness. If the camera's sensitivity drifts even a tiny bit, you might mistake a flicker of the camera for a change in the firefly's atmosphere.

The Goal: The authors wanted to prove that the new GRAVITY+ camera is stable enough to take these long, high-quality photos without the camera itself messing up the data.

The Problem: The "Jittery" Fiber Optic Cable

The team observed a very bright, stable star called Beta Pictoris for seven hours straight. They expected the light to be steady. Instead, they saw the brightness of the light fluctuating by about 10%.

The Analogy: Imagine you are trying to pour water from a bucket into a very narrow drinking straw (the fiber optic cable) while standing on a boat in a choppy sea (the atmosphere).

If the boat rocks (atmospheric turbulence), the water misses the straw.
If the straw moves slightly (fiber position offsets), the water misses.
If the wind blows the water away (atmospheric dispersion), the water misses.

The paper found that the main reason the light was fluctuating wasn't because the star was changing, but because the "water" (starlight) was struggling to get into the "straw" (fiber optic cable) due to these tiny movements and shifts. This caused a smooth, wavy change in brightness across the colors of light.

The Solution: SIMTERFERE (The "Digital Twin")

To figure out exactly what was causing the jitter, the authors built a simulation tool called SIMTERFERE.

The Analogy: Think of SIMTERFERE as a flight simulator for the telescope.

Instead of building a new telescope, they built a virtual one inside a computer.
They fed the simulator all the real-time data from the night of the observation: the wind speed, the temperature, the exact position of the fiber cables, and the wavefronts of the light.
The simulator then "re-ran" the night, predicting what the light should have looked like if it were just the instrument and the atmosphere acting alone.

When they compared the Simulator's Prediction with the Real Data, they matched almost perfectly. This proved that the 10% fluctuations were indeed caused by the "jittery straw" (fiber coupling) and not by the star itself or some unknown mystery.

The Fix: Smoothing Out the Wiggles

Once they knew the cause, they found a way to fix it.

The Polynomial Correction: The fluctuations looked like a smooth curve (like a gentle hill). The team realized they could use a simple math formula (a second-order polynomial) to draw a line through that hill and flatten it out.
- Result: This removed the big 10% wiggles, leaving only tiny ripples.
The Airmass Interpolation: The tiny ripples that were left over were caused by Earth's atmosphere (water vapor, CO2) blocking some light as the star moved across the sky.
- The Old Way: The team used to just take the average of the star's brightness before and after looking at the planet.
- The New Way: They realized the atmosphere changes linearly as the star moves higher or lower in the sky (changing airmass). By using a weighted linear interpolation (a fancy way of drawing a straight line between two points based on the star's height), they could predict the star's brightness much more accurately.

The Result: Reaching "Per-Mill" Stability

After applying these fixes, the team achieved something incredible: 0.1% stability.

The Analogy: Before, the camera was like a shaky hand trying to write a letter. After the fix, the hand was steady enough to write with a fountain pen on a piece of paper without the ink bleeding.

They proved that GRAVITY+ is now stable enough to detect the faint chemical signatures of exoplanet atmospheres. They can now distinguish between a real change in a planet's atmosphere and a glitch in the telescope.

Why This Matters

This paper is a "user manual" for the future of exoplanet hunting. It tells us:

We found the bug: The instability was just fiber coupling issues, not a broken telescope.
We fixed the bug: We have a mathematical recipe to smooth out the data.
We are ready: The instrument is now stable enough to study the atmospheres of young, hot gas giants (like Beta Pictoris b) in incredible detail, potentially revealing if they have clouds, water, or even signs of formation.

In short, the authors built a digital twin to understand their telescope's "jitters," fixed the math to smooth them out, and cleared the path for humanity to finally read the "atmospheric books" of distant worlds.

1. Problem Statement

The GRAVITY+ upgrade to the Very Large Telescope Interferometer (VLTI) introduces a new Adaptive Optics (AO) system (GPAO) designed to significantly improve sensitivity and stability for observing directly imaged exoplanets. However, the spectrometric stability of the instrument over multi-hour timescales had not been fully characterized.

The Challenge: Detecting exoplanet emission requires overcoming high contrast ratios ( $10^{-4}$ to $10^{-6}$ ). This relies on the precise measurement of the coherent flux ( $\Gamma$ ).
The Issue: Previous observations with GRAVITY showed flux variations of $\sim10\%$ between epochs, attributed to poorly understood instrumental effects. These variations bias the derived planet-to-star contrast spectra, limiting the reliability of atmospheric characterization.
The Goal: To quantify the short- and long-term stability of GRAVITY+, identify the physical origins of flux variations, and develop mitigation strategies to reach the photon-noise limit ( $\sim0.1\%$ precision) required for high-fidelity exoplanet spectroscopy.

2. Methodology

The authors employed a data-driven simulation approach combined with a dedicated observation of the bright, stable star $\beta$ Pictoris.

A. Observation Strategy

Target: $\beta$ Pictoris (A6V star, $K=3.48$ mag), chosen for its high brightness and negligible intrinsic variability ( $\ll 1\%$ in K-band).
Setup: A continuous 7-hour observation on December 20, 2024, using the four VLTI Unit Telescopes (UTs) in high-spectral-resolution mode ( $R \approx 4000$ ).
Conditions: Observations covered airmasses from 1.1 to 1.5. Notably, wind speeds were exceptionally low ( $<3$ m/s), inducing the Low Wind Effect (LWE), which degrades Point Spread Function (PSF) stability.

B. The SIMTERFERE Simulator

The core of the study is SIMTERFERE, a publicly available Python-based simulator that reproduces GRAVITY+ on-star observations using ancillary telemetry data.

Data-Driven Approach: Instead of a full end-to-end physical simulation, SIMTERFERE injects real-time measured parameters (telemetry, wavefronts, fiber positions, atmospheric conditions) into a linear model of the instrument.
Workflow:
1. Input: User-defined stellar spectrum (PHOENIX model for $\beta$ Pic) and ancillary data (V2PM, fiber offsets, wavefronts, atmospheric transmission).
2. Propagation: Simulates light propagation through the atmosphere (dispersion/transmission), telescope apertures, and fiber coupling (using Gaussian profiles and measured wavefronts).
3. Interferometry: Combines beams using the Visibility-to-Pixel-Matrix (V2PM) derived from calibration data, applying residual piston errors from the Fringe Tracker (FT).
4. Output: Generates synthetic coherent fluxes ( $\Gamma$ ) for direct comparison with measured data.
Key Feature: This allows the isolation of specific systematic effects (e.g., fiber misalignment vs. atmospheric dispersion) by toggling specific inputs.

3. Key Contributions

Development of SIMTERFERE: A modular, publicly available tool that isolates instrumental processes to diagnose systematic flux variations in interferometric data.
Quantification of Stability: A comprehensive characterization of GRAVITY+ performance over a 7-hour window, establishing a benchmark for future exoplanet observations.
Identification of Systematics: Pinpointing the specific physical mechanisms causing the observed $\sim10\%$ flux variations.
Mitigation Strategies: Validating a two-step correction method (polynomial throughput correction + airmass interpolation) to reduce flux errors to the photon-noise limit.

4. Key Results

A. Sources of Flux Variations

The study decomposed the total flux variation into two distinct components:

Instrumental/Throughput Variations ( $\sim10\%$ ):
- Cause: Dominated by variable fiber coupling efficiency. This is driven by the interplay of wavefront stability (affected by the Low Wind Effect), atmospheric dispersion, and residual fiber offsets.
- Evidence: SIMTERFERE showed that even sub-pixel fiber offsets ( $\approx 5$ mas) significantly alter the throughput. The "Low Wind Effect" caused phase discontinuities between spider vanes, increasing light leakage and reducing coupling stability.
- Character: These variations appear as smooth, low-order continuum changes across the wavelength spectrum.
Telluric Variations ( $\sim1\%$ ):
- Cause: Changes in atmospheric transmission (water vapor, CO $_2$ , CH $_4$ ) correlated with airmass.
- Character: Follows the Beer-Lambert law; linearly dependent on airmass changes.

B. Mitigation and Correction

The authors demonstrated a robust pipeline to correct these variations:

Polynomial Correction: The $\sim10\%$ $\sim 10%$ throughput variations were effectively modeled and removed using second-order polynomials ( $Q_{\lambda,2}$ $Q_{λ, 2}$ ) fitted to the wavelength dependence of the flux.
- Result: Reduced RMS variations from $\sim6.5\%$ to $\sim0.21\%$ .
Telluric Correction (Airmass Interpolation):
- Instead of the standard exogravity approach of simply averaging the star observations before and after the planet exposure, the authors proposed a linear interpolation weighted by airmass change.
- Result: This accounts for the non-linear evolution of atmospheric transmission, reducing residuals to the photon-noise limit.

C. Final Stability Metrics

Precision: After corrections, the coherent flux prediction precision reached $\sim0.1\%$ (specifically $0.117\%$ for baseline-averaged fluxes).
Stability: No systematic drift was observed over the full 7-hour duration.
Comparison: The interpolation method improved precision by $\sim10\%$ for individual baselines and $\sim30\%$ for baseline averages compared to simple averaging.

5. Significance

Validation of GRAVITY+: The study confirms that the GRAVITY+ instrument is highly stable. Once fiber-coupling systematics are corrected, the instrument's performance is limited only by photon noise and telluric variations, not instrumental drift.
Enabling Exoplanet Science: Reaching the 0.1% precision level is critical for detecting subtle spectral features in exoplanet atmospheres (e.g., isotopologues, cloud structures, and variability).
Methodological Advance: The SIMTERFERE framework provides a new standard for diagnosing interferometric data, allowing the community to separate instrumental artifacts from astrophysical signals without needing complex, computationally expensive end-to-end simulations.
Future Impact: The developed correction strategies (polynomial + airmass interpolation) will be integrated into the exogravity data reduction pipeline, directly enhancing the scientific return of future direct imaging campaigns.

In conclusion, the paper demonstrates that the GRAVITY+ instrument, when coupled with the SIMTERFERE diagnostic framework and the proposed correction algorithms, is uniquely capable of high-fidelity, long-duration spectroscopic characterization of directly imaged exoplanets.