The use of genetic algorithm to model protoplanetary discs

TL;DR

This paper introduces a genetic algorithm-based method to optimize parameters in a flared disc model for fitting the spectral energy distributions of protoplanetary discs around young stars, improving analysis efficiency.

Contribution

It develops and applies a genetic algorithm approach for parameter estimation in flared disc models, enhancing the fitting process for diverse SED shapes of young stellar objects.

Findings

GA provides good quality fits for Herbig stars' SEDs.

The method reproduces canonical flared disc models successfully.

Limitations include potential unrealistic parameters and poor fits for embedded and debris discs.

Abstract

The protoplanetary discs of T Tauri and Herbig Ae/Be stars have been studied by using geometric disc models to fit their spectral energy distribution (SED). The simulations provide means to reproduce the signatures of different circumstellar structures, which are related to different levels of infrared excess. Aiming to improve our previous model that assumed a simple flat disc configuration, in the present work we adopt a reprocessing flared disc model that considers hydrostatic, radiative equilibrium (Dullemond et al. 2001). We developed a method to optimise the parameters estimation based on genetic algorithms (GA). This paper is dedicated to describe the implementation of the new code, which has been applied for Herbig stars from the Pico dos Dias Survey catalogue, in order to illustrate the quality of the fitting for a variety of SED shapes. The star AB Aur was used as a test of the GA parameters estimation, demonstrating that the new code reproduces successfully a canonical example of the flared disc model. The GA method gives good quality of fittings, but the range of input parameters must be chosen with caution, since unrealistic disc parameters can be derived. The flared disc model is confirmed to fit the flattened SEDs typical from Herbig stars, however the embedded objects (increasing SED slope) and debris discs (steeper SED) are not well fitted with this configuration. Even considering the limitation of the derived parameters, the automatic process of SED fitting provides an interesting tool for the statistical analysis of the circumstellar luminosity of large samples of young stars.