RT info:eu-repo/semantics/masterThesis
T1 A probabilistic deep learning model to distinguish cusps and cores in dwarf galaxies
A1 Expósito Márquez, Julen
A2 Máster Universitario en Astrofísica
AB Numerical simulations within a cold dark matter (DM) cosmology form halos with a characteristicdensity profile with a logarithmic inner slope of -1. Various methods, such as Jeans and Schwarzschildmodelling, have been used in an attempt to determine the inner density of observed dwarf galaxies,in order to test this theoretical prediction. Here, we develop a convolutional mixture density neuralnetwork (CMDNN) to derive a posterior distribution of the inner density slopes of DM halos. We trainthe CMDNN on a suite of simulated galaxies from the NIHAO and AURIGA projects, inputting line-ofsight velocities and 2D spatial information of the stars within simulated galaxies. The output of theCMDNN is a probability density function representing the posterior probability of a certain slope to bethe correct one, thus producing accurate and complex information on the uncertainty of the predictions.The model recovers accurately the correct inner slope of dwarfs: ∼82%of the galaxies have a derivedinner slope within ±0.1 of their true value, while ∼98% within ±0.3. We then apply our model tofour Local Group dwarf spheroidal galaxies and find similar results to those obtained with the Jeansmodelling based code GravSphere. Fornax dSph has a strong indication of possessing a central DMcore, Carina and Sextans have cusps (although the latter with a large uncertainty), while Sculptor showsa double peaked PDF indicating that a cusp is preferred, but a core can not be ruled out.Our results show that simulation-based inference with neural networks provide a innovative andcomplementary method for the determination of the inner DM profiles in galaxies.Keywords: galaxies: dwarf - evolution - formation - haloes - dark matter - simulations - NIHAO - slope- NFWmachine learning: convolutional - neural - network - posterior - probability - distribution
YR 2022
FD 2022
LK http://riull.ull.es/xmlui/handle/915/31656
UL http://riull.ull.es/xmlui/handle/915/31656
LA en
DS Repositorio institucional de la Universidad de La Laguna
RD 19-may-2024