[Paper Review] Morphological classification of radio sources for galaxy evolution and cosmology with SKA-MID

Sphesihle Makhathini, O. Smirnov+2 more|arXiv (Cornell University)|Dec 18, 2014

Radio Astronomy Observations and Technology1 references3 citations

TL;DR

This paper uses MeqTrees-based simulations to forecast the morphological classification of high-redshift radio sources (FRI, FRII, star-forming galaxies) with the SKA-MID telescope. It finds that SKA1-MID can reliably distinguish source morphologies up to redshift z ≈ 8 after 10 hours of integration, with resolution—not sensitivity—being the limiting factor, and that a 120 km maximum baseline performs as well as longer baselines, enabling robust galaxy evolution and cosmology studies via morphological tracers of dark matter and AGN feedback.

ABSTRACT

Morphologically classifying radio sources in continuum images with the SKA has the potential to address some of the key questions in cosmology and galaxy evolution. In particular, we may use different classes of radio sources as independent tracers of the dark-matter density field, and thus overcome cosmic variance in measuring large-scale structure, while on the galaxy evolution side we could measure the mechanical feedback from FRII and FRI jets. This work makes use of a exttt{MeqTrees}-based simulations framework to forecast the ability of the SKA to recover true source morphologies at high redshifts. A suite of high resolution images containing realistic continuum source distributions with different morphologies (FRI, FRII, starburst galaxies) is fed through an SKA Phase 1 simulator, then analysed to determine the sensitivity limits at which the morphologies can still be distinguished. We also explore how changing the antenna distribution affects these results.

Motivation & Objective

• To assess the feasibility of morphologically classifying high-redshift radio sources using SKA1-MID, focusing on instrumental limitations such as sensitivity, resolution, and imaging fidelity.
• To determine the redshift limits at which FRI, FRII, and star-forming galaxies (SFGs) can be reliably distinguished in deep SKA1-MID continuum surveys.
• To evaluate whether a 120 km maximum baseline layout (W9) performs as well as longer baselines (e.g., 170 km) in resolving source morphologies.
• To explore how changes in sensitivity (50%, 70%, 100% of SKA1-MID) affect signal-to-noise ratios and morphological classification limits.
• To assess the potential of model-fitting techniques to super-resolve sub-PSF structures in high-SNR SKA1-MID images.

Proposed method

• Simulated high-resolution continuum images of FRI, FRII, and SFGs at various redshifts (z = 1 to z = 8.5) were generated using realistic source distributions.
• These sky models were passed through an SKA Phase 1 simulator using two antenna layouts: REF2 (baseline design) and W9 (120 km max baseline).
• Imaging simulations were performed with 10-hour integration times to assess source detectability and morphological fidelity.
• A pyBDSM source finder with a 5σ detection threshold was applied to identify sources, with false positives quantified using techniques from Serra et al. (2012).
• Signal-to-noise ratios (SNR) were computed as a function of redshift for different sensitivity levels (50%, 70%, 100% of SKA1-MID) to determine classification limits.
• Model-fitting techniques (e.g., Martí-Vidal et al. 2014) were evaluated for their ability to resolve sub-PSF structures, scaling as √SNR.

Experimental results

Research questions

• RQ1 Can SKA1-MID reliably distinguish FRI, FRII, and star-forming galaxy morphologies at high redshifts (z > 5)?
• RQ2 What is the maximum redshift at which morphological classification remains feasible with 10 hours of integration?
• RQ3 Does a 120 km maximum baseline layout (W9) perform as well as the longer 170 km baseline (REF2) in resolving source morphologies?
• RQ4 How does sensitivity (50%, 70%, 100% of SKA1-MID) affect the signal-to-noise ratio and thus morphological classification capability?
• RQ5 To what extent can model-fitting techniques super-resolve sub-PSF features in high-SNR SKA1-MID images?

Key findings

• Morphological classification of FRI and FRII sources is feasible up to redshift z ≈ 8 after 10 hours of integration with SKA1-MID.
• Star-forming galaxies (SFGs) can be reliably classified up to z ≈ 5 under the same integration time, with detection possible below 5σ if false positives are quantified.
• The W9 layout (120 km max baseline) performs as well as or slightly better than the REF2 layout (170 km max baseline), indicating that longer baselines are not necessary for morphological classification.
• Resolution, not sensitivity, is the primary limiting factor for morphological classification, with SKA1-MID resolving structures down to ~5 kpc at cosmological distances.
• With SNR between 10 and 100 at z ≈ 1–4, model-fitting techniques can resolve features at 1/3 to 1/10 of the PSF size, enabling sub-PSF classification of kpc-scale structures.
• The full SKA-MID (10× SKA1-MID sensitivity) would extend classification limits to even higher redshifts, though resolution remains the key constraint.