Presentation #102.117 in the session Poster Session.
We aim to leverage different kind of observations to constrain the initial properties of protoplanetary disks for synthetic populations. For this purpose, we first run a parameter study using a 1D model that includes the viscous evolution of a gas disk, the concurrent growth and drift of dust and pebbles using a two-population model, flux-limited planetesimal formation coupled with an emission model to computed the mm flux observable with ALMA. This is used to train a machine learning surrogate model that computes the relevant quantity for comparison with observations in seconds. We use that surrogate model to find the distributions of initial conditions that best reproduce disk observations together: lifetimes, stellar accretion rates, and emitted flux.
We find that to best reproduce observations, the initial masses should be larger by a factor of ~3 compared to observed dust masses. This indicates that not all the dust contributes to the emitted flux. We also find that disc radii have to be smaller so that they can maintain stellar accretion rates. Otherwise, gas remains in the outer regions without being accreted onto the star. The optimum values of the alpha viscosity parameter lie between 3×10-3 and 3×10-2.
We also investigate the effects of internal (X-ray) and external (using the FRIED grid) photoevaporations. We find that external photoevaporation lead to a strong dependence of lifetimes with stellar mass, which is not retrieved in observation. Internal photoevaporation, due to correlation between X-ray luminosity and stellar mass, has a much weaker dependence. This suggests that internal photoevaporation is better suited to reproduce disk lifetime observations than external photoevaporation.