Estimates of the planet yield from ground-based high-contrast imaging observations as a function of stellar mass

Abstract

We use Monte Carlo simulations to estimate the number of extrasolar planets that are directly detectable in the solar neighborhood using current and forthcoming high-contrast imaging instruments. Our calculations take into consideration the important factors that govern the likelihood for imaging a planet, including the statistical properties of stars in the solar neighborhood, correlations between star and planet properties, observational effects, and selection criteria. We consider several different ground-based surveys, both biased and unbiased, and express the resulting planet yields as a function of stellar mass. Selecting targets based on their youth and visual brightness, we find that strong correlations between star mass and planet properties are required to reproduce high-contrast imaging results to date (i.e., HR 8799, β Pic). Using the most recent empirical findings for the occurrence rate of gas-giant planets from radial velocity (RV) surveys, our simulations indicate that naive extrapolation of the Doppler planet population to semimajor axes accessible to high-contrast instruments provides an excellent agreement between simulations and observations using present-day contrast levels. In addition to being intrinsically young and sufficiently bright to serve as their own beacon for adaptive optics correction, A-stars have a high planet occurrence rate and propensity to form massive planets in wide orbits, making them ideal targets. The same effects responsible for creating a multitude of detectable planets around massive stars conspire to reduce the number orbiting low-mass stars. However, in the case of a young stellar cluster, where targets are approximately the same age and situated at roughly the same distance, MK-stars can easily dominate the number of detections because of an observational bias related to small number statistics. The degree to which low-mass stars produce the most planet detections in this special case depends upon whether multiple formation mechanisms are at work. Upon relaxing our assumption that planets in ultra-wide (a > 100 AU) orbits resemble the RV sample, our simulations suggest that the companions found orbiting late-type stars (AB Pic, 1RXSJ1609, GSC 06214, etc.) are consistent with a formation channel distinct from that of RV planets. These calculations explain why planets have thus far been imaged preferentially around A-stars and K-, M-stars, but no spectral types in between, despite concerted efforts targeting F-, G-stars. © 2011. The American Astronomical Society. All rights reserved.