A Python-Based Peeling Framework for Radio Interferometry: Application to uGMRT 650MHz Imaging

TL;DR

A Python-based direction-dependent calibration and peeling framework that efficiently subtracts bright-source models and suppresses their associated direction-dependent artifacts, producing significantly flattened backgrounds and improving image fidelity and faint-source detectability is presented.

Abstract

Modern radio interferometric arrays offer high sensitivity, wide fields of view, and broad frequency coverage, but also pose significant data calibration challenges. Standard direction-independent calibration is insufficient to correct direction-dependent effects, such as ionospheric phase distortions and primary beam variations, which produce strong artifacts around bright sources and limit achievable image dynamic range. Built on standard CASA tasks, we present a Python-based direction-dependent calibration and peeling framework, demonstrated using radio continuum imaging data from the upgraded Giant Metrewave Radio Telescope (uGMRT). The framework efficiently subtracts bright-source models and suppresses their associated direction-dependent artifacts, producing significantly flattened backgrounds and improving image fidelity and faint-source detectability. We further introduce an optimized ``model-restoration'' strategy that mitigates direction-dependent artifacts while preserving the flux densities and morphologies of bright sources that are themselves of scientific interest. For fields containing multiple bright sources, sequential application of the framework systematically reduces background noise, thereby increasing sensitivity and faint-source detectability. The framework is Python-based, CASA-compatible, and can be readily applied to other mid- and low-frequency interferometric arrays. The code is publicly released with this paper.

Figures (8)

• Figure 1: The uv-coverage of the J0210-3 observations.
• Figure 2: Comparison illustrating the effect of the peeling procedure on the uGMRT 650 MHz imaging data. The white dashed box outlines the peeling mask region. The left panel shows a sub-region of the image obtained after the eighth (final) round of self-calibration, representing the optimized result of direction-independent calibration. The right panel presents the corresponding image after applying direction-dependent calibration and model subtraction to the target bright source, in which artifacts associated with the bright source are substantially suppressed, resulting in a significantly flatter background and enabling neighboring faint sources to be more clearly discerned.
• Figure 3: Schematic overview of the direction-dependent calibration framework adopted in this work. Solid boxes represent the standard peeling procedure, in which the model of a problematic bright source is subtracted from the visibility data after direction-dependent calibration. Dashed boxes highlight the additional steps introduced in our optimized model-restoration strategy, designed to preserve the flux density and morphology of bright sources of scientific interest. All TempX.ms are deleted immediately after use to minimize storage usage.
• Figure 4: Comparison of radio source detections in the vicinity of the target bright source before (left panel) and after (middle panel) the peeling procedure, with independent validation from MeerKAT 1.3 GHz data (right panel). In the post-peeling image (middle panel), two faint radio sources (solid green ellipses) are detected at significance levels exceeding $3\,\sigma$. These sources are severely obscured by strong direction-dependent artifacts in the direction-independent calibrated image (left panel) and are therefore missed by the source detection algorithm prior to peeling. Their positions, as detected after peeling, are indicated by dashed green ellipses in the left panel. The right panel shows the corresponding MeerKAT 1.3 GHz reference image, in which the same sources (solid cyan ellipses) are detected at $>5\,\sigma$, thereby independently confirming the reality of these detections. The synthesized beam (indicated by the white line in bottom-left corner) is $5.4" \times 4.2"$ for the uGMRT 650 MHz image and $8.0" \times 7.5"$ for the MeerKAT 1.3 GHz image.
• Figure 5: Comparison of source flux densities before and after peeling. Red open circles denote sources located within 5 arcmin of the target bright source, while blue open squares indicate sources at radial distance of 5--10 arcmin. The median ratios of pre-peeling to post-peeling flux densities are $0.98^{+0.49}_{-0.19}$ and $1.00^{+0.27}_{-0.21}$ for the inner and outer regions, respectively. The shaded region bounded by dashed lines marks the 16th--84th percentile range (corresponding to the 1$\sigma$ interval).