Self-Gravitating Scalar Field Configurations, Ultra Light Dark Matter and Galactic Scale Observations

TL;DR

This thesis investigates ultra light scalar field dark matter with self-interactions, demonstrating how galactic observations can constrain particle properties, and shows that self-interactions influence galaxy dynamics and satellite galaxy longevity.

Contribution

It introduces a detailed analysis of self-interacting ultra light dark matter using solutions of the Gross-Pitaevskii-Poisson equations and applies machine learning for parameter inference from galaxy data.

Findings

Galactic observations constrain self-interaction strengths to $ imes 10^{-96}$ to $ imes 10^{-95}$.

Self-interactions can explain rotation curves and soliton-halo relations in low surface brightness galaxies.

Attractive self-interactions extend satellite dwarf galaxy lifetimes, easing previous constraints.

Abstract

In this thesis, we investigate the possibility that dark matter consists of ultra light spin-zero particles with mass $m \sim 10^{-22}\ \text{eV}$. We focus on the role of self-interactions, assuming all other non-gravitational couplings to Standard Model particles are negligible. Such ultra light dark matter (ULDM) is expected to form stable self-gravitating scalar field configurations (solitons), whose properties depend on the particle mass and self-coupling $\lambda$. Using solutions of the Gross-Pitaevskii-Poisson equations, we explore how galactic-scale observations can constrain $m$ and $\lambda$. We show that observational upper limits on the mass enclosed in central galactic regions can probe both attractive and repulsive self-interactions with strengths $\lambda \sim \pm 10^{-96} - 10^{-95}$. We further demonstrate that self-interactions can allow ULDM to describe observed rotation curves as well as satisfy an empirical soliton-halo mass relation in low surface brightness galaxies for $m \sim 10^{-22}\ \text{eV}$ and $\lambda \gtrsim 10^{-90}$. We also study tidal effects in satellite dwarf galaxies and find that attractive self-interactions can extend their lifetimes over cosmological timescales, allowing ULDM to evade recent constraints derived for the non-interacting case. Finally, we explore machine learning based inference of dark matter and baryonic parameters from galaxy rotation curves, showing that neural networks can recover parameters consistent with observations.