Dynamical characterization of galaxies up to z ∼ 7

Abstract

Context. The characterization of the dynamical state of galaxies up to z ∼ 7 is crucial for constraining the mechanisms that drive the mass assembly in the early Universe. However, it is unclear whether the data quality of typical observations obtained with current and future facilities is sufficient to perform a solid dynamical analysis at these redshifts.

Aims. This paper defines the angular resolution and signal-to-noise ratio (S/N) required for a robust characterization of the dynamical state of galaxies up to the Epoch of Reionization. The final aim is to help design future spatially resolved surveys targeting emission lines of primeval galaxies.

Methods. We investigate the [C  II]-158 μm emission from six z ∼ 6 − 7 Lyman break galaxies at three different inclinations from the S ERRA zoom-in cosmological simulation suite. The S ERRA galaxies cover a range of dynamical states: from isolated disks to major mergers. We create 102 mock observations with various data quality and apply the kinematic classification methods commonly used in the literature. These tests allow us to quantify the performances of the classification methods as a function of angular resolution and S/N.

Results. We find that barely resolved observations, typical of line detection surveys, do not allow the correct characterization of the dynamical stage of a galaxy, resulting in the misclassification of disks and mergers in our sample by 100 and 50%, respectively. However, even when using spatially resolved observations with data quality typical of high- z galaxies ( S/ N ∼ 10, and ∼3 independent resolution elements along the major axis), the success rates in the merger identification of the standard kinematic classification methods, based on the analysis of the moment maps, range between 50 and 70%. The high angular resolution and S/N needed to correctly classify disks with these standard methods can only be achieved with current instrumentation for a select number 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 mock data created from S ERRA galaxies, and show that PVsplit can predict whether a galaxy is a disk or a merger provided that S/ N ≳ 10, and the major axis is covered by ≳3 independent resolution elements.