Table 3.
Summary of the application of the classification techniques on our sample of mock data.
| Method | Testing dataset ( † ) | Outcomes and Recommendations |
|---|---|---|
| Velocity fields and V/σ (Sect. 6.1) | a), b), c), d) | This method’s ability to identify mergers strongly depends on the S/N and angular resolution of the data. 50% of mergers in our sample are misclassified as disks at low angular resolutions. Disk galaxies are always classified correctly. |
| Kinemetry based (Sect. 6.2) | a), b), c), d) | Up to 30% of our mergers are misclassified as disks at medium- and low-angular resolutions. This method is optimal for performing a kinematic classification only when high-angular resolution (NIRE ≳ 4) data are available. The success rate of this method in distinguishing disks and mergers weakly depends on the S/N of the data. |
| Qualitative (Sect. 6.3) | e) | This method can result in a substantial underestimation of the disk fraction: ≈100% of the disks in our sample are misclassified as mergers or dispersion-dominated galaxies. Instead, the fraction of mergers correctly classified is 50%. Barely resolved data with S/N ≲ 4, typical of line-detection surveys, should not be employed to extract galaxies’ dynamical states. |
| PVsplit (Sect. 7) | a), b), c), d) | Observations of galaxies with at least 3 independent resolution elements along the major axis and S/N ≳ 10 guarantee the correct identification of disks and mergers. |
Notes. ( † )a): Ideal data; b) ALMA high-angular resolution, high S/N data; c) ALMA medium- and low-angular resolution, high S/N; d) ALMA high- and medium-angular resolution, low S/N; e) ALMA barely resolved, low S/N. For further details about the definition of the datasets, see Sect. 4.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.