The likelihood of not detecting cavity-carving companions in transition discs -- A statistical approach

Abstract

Protoplanetary discs with cavities, also known as ``transition discs'', constitute ~10% of protoplanetary discs at sub-mm wavelengths. Among several explanations, one hypothesis suggests these cavities are carved by undetected stellar or planetary companions. We present a novel approach to quantify the likelihood that a cavity-carving companion goes undetected because it is either too close to the star (i.e., has a small projected separation) or too faint to be resolved. We generate two independent samples of stellar and planetary companions with random sky orientations, assuming distributions in eccentricity, mass ratio, and time-weighted orbital phases, to study the statistical properties of the cavities they produce. We calculate the likelihood that a companion appears with a projected separation $d$ relative to its semi-major axis $a_{bin}$ ($d/a_{bin}$). Then, using a disc truncation model, we calculate the likelihood that companions carve a cavity with size $a_{cav}$ relative to its semi-major axis $a_{bin}$ and projected separation $d$, deriving distributions of $a_{bin}/a_{cav}$ and $d/a_{cav}$. We find that stellar companions carve cavities with median sizes ~3 times larger than their projected separation $d$ ($a_{cav}\sim3d$, $a_{cav}\sim1.7 d$ for planets), but with a statistically significant tail towards larger values ($a_{cav}\gg 3d$). We estimate the likelihood that cavity-carving companions go undetected due to projection effects when the system is observed with spatial resolution $R$, $P(d< R)$. Considering observational constraints on companion masses, we apply this framework to 13 well-known transition discs. We find that undetected stellar companions are unlikely in 8 out of 13 systems we considered, with 5 notable exceptions: ABAur, MWC758, HD135344B, CQTau and HD169142. The presence of undetected planets cannot be excluded in any of the transition discs considered.