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

Abstract

The implementation of the GRAVITY+ Adaptive Optics (GPAO) system at VLTI enables unprecedented sensitivity and stability in optical interferometry. This allows high-precision characterization of directly imaged exoplanets at medium spectral resolution, providing a new pathway for studying planetary atmospheres. We aim to quantify and characterize the short- and long-term stability of GRAVITY+ through a consecutive seven-hour observation of the bright and stable star beta Pictoris, providing a benchmark for future exoplanet observations. We developed SIMTERFERE, a data-driven simulation tool that reproduces GRAVITY+ on-star observations using ancillary instrument and telemetry data. By comparing the simulations with the measured coherent fluxes, we traced the origins of systematic flux variations and assessed their impact on exoplanet contrast measurements. We find that the approximately 10% variations are dominated by throughput changes driven by variable fiber coupling, which depends on wavefront stability, atmospheric dispersion, and residual fiber offsets. These variations appear as smooth continuum changes across wavelength and can be effectively mitigated using second-order polynomial corrections. After removing these instrumental effects, the remaining approximately 1% variations are almost purely of telluric origin, which we can reliably correct down to the photon-noise limit (0.1% precision) using a contrast spectrum approach with linear airmass interpolation. The GRAVITY+ inferometric instrument is highly stable: low-order continuum and telluric variations can be corrected with high precision, making it uniquely capable of high-fidelity characterization of directly imaged exoplanets.

Figures (13)

• Figure 1: Flow chart of the SIMTERFERE simulation framework. Simulated coherent fluxes (orange) are generated by injecting atmospheric and instrumental effects (blue), derived from the ancillary telemetry of the GRAVITY observations (dark gray). The overall configuration of the simulation and the subcomponents is controlled through user-defined hyperparameters (light gray).
• Figure 2: Top: Example frame (exposure 1/18, integration 11/16) showing measured (left) and simulated (right) absolute SC coherent flux per wavelength bin for the six baselines (43, ..., 12) and their baseline averages $\langle \mathrm{SC} \rangle_b$ (different colors), taken at high airmass. Absolute coherent fluxes are normalized to the maximum flux in the exposure. Bottom: Relative absolute coherent flux differences with respect to the time-averaged value for each baseline. Second order polynomial fits are indicated by dashed white lines. The systematic differences between the measured and simulated coherent fluxes are attributed to unknown fiber offsets, which are discussed in more detail in Sect. \ref{['subsection: fiber offsets']}. The associated movie is available online.
• Figure 3: Left: Standard Deviation (STD) of the measured (solid lines) and simulated (dotted lines) relative SC (1628 spectral bins) absolute coherent flux $|\Gamma_{\mathrm{SC43}}|$, the combined SC flux $\sqrt{F_{\mathrm{SC}4}F_{\mathrm{SC}3}}$, and the FT (6 spectral bins) absolute coherent flux $|\Gamma_{\mathrm{FT43}}|$ (different shades) for an examples exposure at high airmass on baseline 43. Right: Same as the left panel, but for an exposure at low airmass.
• Figure 4: Standard deviation (STD) of the simulated SC2 relative fluxes for different applied fiber offsets, either along or orthogonal to the dispersion direction. Black lines represent STDs with manual fiber tracking and no applied offsets, while green lines show STDs with applied offsets. The dotted lines indicate the respective wavelength average. An offset parallel to the dispersion direction shifts the characteristic "V-shape" toward positive wavelengths, and an antiparallel shift produces the opposite effect. Offsets orthogonal to the dispersion direction increase the overall STDs. The observed asymmetries between opposite offset directions arise from the asymmetry of the PSFs.
• Figure 5: Left: Standard deviation (STD) of the measured relative absolute coherent SC fluxes for all exposures during the observation, shown for the six baselines (43, ..., 21) and the baseline average $\langle \cdot \rangle_{b}$ (different colors). Middle: Same as the left panel, but showing the respective simulation results. Right: Same as the middle panel, but with manual fiber tracking applied during the simulations.