Dynamical characterization of galaxies up to $z \sim 7$

TL;DR

This study assesses the data quality needed to accurately determine the dynamical states of high-redshift galaxies, revealing limitations of current methods and proposing a new classification approach using position-velocity diagrams.

Contribution

It introduces PVsplit, a novel classification method that improves dynamical state determination of early Universe galaxies from simulated observations.

Findings

Standard kinematic methods often misclassify galaxy dynamics at high redshift.

High angular resolution and S/N are required for reliable classification.

PVsplit effectively distinguishes disks from mergers with sufficient data quality.

Abstract

The characterization of the dynamical state of galaxies up to z~7 is crucial for constraining the mechanisms driving the mass assembly in the early Universe. However, it is unclear whether the data quality of current and future observations is sufficient to perform a solid dynamical analysis. This paper defines the angular resolution and S/N required for a robust characterization of the dynamical state of galaxies up to the EoR. The final aim is to help design spatially-resolved surveys targeting emission lines of primeval galaxies. We investigate the [CII]-158um emission from z~6-7 LBGs from the SERRA cosmological simulation, covering a range of dynamical states: from disks to major mergers. We create ALMA mock observations with various data quality and apply the kinematic classification methods used in the literature. These tests allow us to quantify the performances of such methods as a function of angular resolution and S/N. We find that barely-resolved observations do not allow the correct dynamical characterization of a galaxy, resulting in the misclassification of all disks in our sample. However, even when using spatially-resolved observations with data quality typical of high-z galaxies, the standard kinematic classification methods, based on the analysis of the moment maps, fail to distinguish a merger from a disk. The high angular resolution and S/N needed to apply these standard methods successfully can be achieved with current data only for a handful of bright galaxies. We propose a new classification method, called PVsplit, that quantifies the asymmetries and morphological features in position-velocity diagrams using three empirical parameters. We test PVsplit on our mock data concluding that it can predict whether a galaxy is a disk or a merger provided that S/N $\gtrsim10$, and the major axis is covered by $\gtrsim3$ independent resolution elements.