Masses of Potentially Habitable Planets Characterized by the Habitable Worlds Observatory

Abstract

Constraints on the masses of exoplanets directly imaged and characterized by the Habitable Worlds Observatory (HWO) are crucial for categorizing these planets and interpreting their spectra. In particular, achieving a mass measurement with a precision of approximately 10% or better may be necessary to identify the dominant gaseous species in the atmospheres of Earth-like planets. This is essential for assessing their habitability and interpreting potential biosignatures (arXiv:2502.01513). Space-based astrometry will be required to measure the masses of planets in face-on systems, or planets orbiting hot and rapidly rotating or highly active stars. Astrometric uncertainties are dominated by the number and magnitude of background reference stars needed to precisely measure the astrometric wobble of the target star induced by the planet. To that end, we propose a program to measure the masses of Earth analogs orbiting HWO target stars with ultra-high-precision astrometry obtained with the HWO high-resolution instrument. We assess the photon-noise error budget for these observations. We find that, for a field of view spanning a few square arcminutes, the astrometric uncertainty due to the number and brightness of reference stars dominates the photon-noise error budget, particularly for targets near the Galactic poles. We explored the impact of filter choice and location in the sky on the photon-noise astrometric uncertainties by simulating the magnitude distribution of reference stars across different filters at a range of galactic longitudes and latitudes. We find that a ~ 200-day survey in the Gaia G band consisting of 100 epochs per target star distributed over the 5-year prime mission with a 6m aperture HWO equipped with a 6' x 6' field-of-view would be required to achieve the photon-noise sensitivity to measure the masses of the ~ 40 Earth-mass habitable-zone planets to ~10%.

Figures (5)

• Figure 1: A prior on the mass of a directly-imaged Earthlike planet is needed to accurately interpret the reflected-light spectrum. Each panel shows the posterior probability disribution for the parameters used in the retrieval of a simulated reflected-light spectrum of a modern-day Earthlike planet. A prior on the planetary mass of 10% is needed to correctly identify the dominant background atmospheric gas. Figure from Damiano:25, used with permission.
• Figure 2: The astrometric and RV signals of Earth analogs orbiting nearby stars are small and below the current state of the art, while the astrometric signals of such planets are positively correlated with their angular separation but anti- correlated with their contrast and RV signal. (left) The amplitude of the astrometric reflex signal versus the radial velocity reflex signal for Earth-mass planets receiving an Earth-like instellation orbiting a sample of nearby stars that are likely to be targets for HWO direct imaging. Credit: Eric Mamajek, used with permission. (right) Contrast versus astrometric signal for a similar sample of stars hosting Earth-like planets. The size of the circles are proportional to the angular separation of the planet and the colors of the circles are proportional to the RV amplitude.
• Figure 3: Roughly 30% of the LUVOIR-B (and thus HWO) target sample will be of spectral type A or F, and thus be too hot and/or rapidly rotating for very high precision RV measurements. (left) The colored points show the stellar luminosity versus distance for the 158 nearby stars the LUVOIR-B target sample, which will be similar to the HWO sample. The points color coded by habitable zone (HZ) completeness (right) Distribution of spectral types for the LUVOIR-B sample. Figure from LUVOIR, used with permission.
• Figure 4: Number of stars per square degree per magntitude in the Gaia $G$ band as function of $G$ magnitude for a line-of-sight toward Galactic latitude $l=90^\circ$ and Galactic longitude $b=0^\circ$ (left), and $b=80^\circ$ (right)
• Figure 5: (left) Cumulative number of reference stars with magnitude $m_X\le20$ as a function of filter for $\ell = 90^\circ$ and various Galactic latitudes $b$ for a detector FOV of $\Omega=36~{\rm arcmin^2}$. (right) The contribution of the astrometric uncertainty due to reference stars as a function of filters for $\ell = 90^\circ$ and various Galactic latitudes $b$. Here we have assumed a diffraction-limited PSF, $t_{\rm exp}=30~{\rm min}$, $D=6~{\rm m}$, $\Omega_{\rm det}=36~{\rm arcmin^2}$, and an overall throughput of $\epsilon=0.25$.