Astrometric Reconnaissance of Exoplanetary Systems (ARES). I. Methodology validation with HST point-source images of Proxima Centauri

TL;DR

The paper presents an astrometric methodology validation for exoplanet reconnaissance (ARES) using multi-epoch HST point-source imaging of Proxima Centauri, anchored to Gaia DR3. By refining WFC3/UVIS geometric-distortion corrections and employing epoch propagation with Gaia, the authors obtain Proxima’s position, proper motion, and parallax with sub-mas precision, consistent with Gaia within ~1σ. They assess the presence of Proxima c via proper-motion anomaly (PMa), deriving a provisional mass near a few Earth masses under simplifying orbital assumptions, though current uncertainties limit a definitive detection. The work establishes a foundation for future HST spatial-scanning observations to reach tens of microarcseconds and enable direct searches for low-mass companions, leveraging Gaia–HST synergy for robust astrometric benchmarking.

Abstract

We present the first results of the Astrometric Reconnaissance of Exoplanetary Systems (ARES) project, aimed at validating and characterizing candidate exoplanets around the nearest systems using multi-epoch Hubble Space Telescope (HST) data. In this first paper, we focus on Proxima Centauri, leveraging archival and recent HST observations in point-source imaging mode. We refine the geometric-distortion calibration of the HST detector used, and develop a robust methodology to derive high-precision astrometric parameters by combining HST measurements with the Gaia DR3 catalog. We determine Proxima's position, proper motion, and parallax with uncertainties at the $\sim$0.4-mas, 50-$μ$as yr$^{-1}$, and 0.2-mas level, respectively, achieving consistent results with what measured by Gaia within $\sim$1$σ$. We further investigate the presence of the candidate exoplanet Proxima c by analyzing the proper-motion anomaly derived from combining long-term HST-based and short-term Gaia astrometry. Under the assumption of a circular, face-on orbit, we obtain an estimated mass of $m_c = 3.4^{+5.2}_{-3.4}$ $M_\odot$, broadly consistent with radial-velocity constraints but limited by our current uncertainties. These results establish the foundation for the next phase of ARES, which will exploit HST spatial-scanning observations to achieve astrometric precisions of a few tens of $μ$as and enable a direct search for astrometric signatures of low-mass companions.

Figures (6)

• Figure 1: Composite image of Proxima obtained by co-adding all HST exposures, properly normalized, from October 2012 to February 2025. The colored line represents the motion of Proxima predicted by our astrometric fit (see Sect. \ref{['sec:astromodel']}), with the color changing from blue to white to red as the epoch increases (see also labels in the plot). North is up and East is to the left.
• Figure 2: Positional residuals from the model (O$-$C) as a function of $x_{\rm raw}$ and $y_{\rm raw}$, respectively from top to bottom. Black points refer to the $x$ residuals, while red points represent the $y$ residuals. The gray dashed line in each plot is set at 0 for reference. The left panels show the O$-$C at the first iteration of the astrometric fit, while the right panels display the residuals at the last iteration of the fit (less points are visible because of the iterative rejection of outliers, see the text for details). In the top panels, we report the standard deviation ($\sigma_{\rm O-C}$ and the absolute maximum scatter of the points ($|\Delta_{\rm O-C}|$) as a reference.
• Figure 3: Series of HST observations of Proxima. The main panel shows the path in the $(\alpha\cos\delta,\delta)$ plane of Proxima. The filled black dots are the HST measurements used in the last iteration of the astrometric fit, while the open dots refer to the rejected measurements. The gray, solid line is the expected linear motion (i.e., without the annual-parallax contribution) of Proxima, whereas the colored line represents the astrometric solution of Proxima. The line color changes from blue to white to red as the epoch increases (from the bottom left to the top left of the plot). The inset in the main panel shows the residuals from the linear motion, which highlight the annual-parallax ellipse. The residuals of the astrometric fit as a function of $\alpha\cos\delta$/$\delta$ coordinates are displayed in the side panels. The gray, dashed line is set to 0 as a reference. In each of these side panels, we report the standard deviation ($\sigma_{\rm O-C}$ and the absolute maximum scatter of the points ($|\Delta_{\rm O-C}|$).
• Figure 4: Expected mass of Proxima c as a function of its orbital radius derived from the PMa analysis. The solid blue line shows the mass estimate obtained in this work from the combination of HST and Gaia-DR3 astrometry, with the shaded light-blue region representing the $1\sigma$ uncertainty. The relation obtained combining Hip-Gaia (just "Hipparcos" in the plot legend for clarity) and part of the HST data is shown as a solid green line. The dashed yellow line represents the PMa-based relation obtained by 2022KervellaPMa using Hipparcos and Gaia. The pink triangle marks the minimum mass derived of Proxima c from 2020SciA....6.7467D, while the red point shows the mass estimate from 2020KervellaProxima.
• Figure 5: One- (histograms) and two-dimensional (marginalized) projections of the posterior-probability distributions of the astrometric parameters of Proxima. The dashed, gray lines set the best-fit parameters (also reported, with the corresponding uncertainties, above each histogram). The marginalized projections of the posterior-probability distributions show 1, 2 and 3 $\sigma$ contour in different shades of blue.