Exocomet orbital distribution around $β$ Pictoris

Abstract: The 23 Myr young star $\beta$ Pictoris is a laboratory for planet formation studies given its observed debris disk, its two directly imaged super-Jovian planets, and the evidence of transiting extrasolar comets. The most recent evidence of exocometary transits around $\beta$ Pic came from the TESS space mission. Previous analyses of these transits constrained the orbital distribution of the underlying exocomet population to range between about 0.03 and 1.3 AU assuming a fixed transit impact parameter. Here we examine the distribution of the observed transit durations (${\Delta}t$) to infer the orbital surface density distribution ($\delta$) of the underlying exocomet sample. We show that a narrow belt of exocomets around $\beta$ Pic, in which the transit impact parameters are randomized but the orbital semimajor axes are equal, results in a pile-up of long transit durations. This is contrary to observations, which reveal a pile-up of short transit durations (${\Delta}t \approx 0.1$ d) and a tail of only a few transits with ${\Delta}t > 0.4$ d. A flat density distribution of exocomets between about 0.03 and 2.5 AU results in a better match between the resulting ${\Delta}t$ distribution and the observations but the slope of the predicted ${\Delta}t$ histogram is not sufficiently steep. An even better match to the observations can be produced with a $\delta \propto a^{\beta}$ power law. Our modeling reveals a best fit between the observed and predicted ${\Delta}t$ distribution for $\beta = -0.15_{-0.10}^{+0.05}$. A more reasonable scenario in which the exocometary trajectories are modeled as hyperbolic orbits can also reproduce the observed ${\Delta}t$ distribution to some extent. Our results imply that cometary material exists on highly eccentric orbits with a more extended range of semimajor axes than suggested by previous spectroscopic observations.