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

Abstract

This code release accompanies the study “A Python-Based Peeling Framework for Radio Interferometry: Application to uGMRT 650 MHz Imaging”. The study presents a modular Python pipeline for implementing direction-dependent calibration and bright-source peeling to suppress artifacts and improve image fidelity in upgraded Giant Metrewave Radio Telescope (uGMRT) Band 4 (550–850 MHz) continuum data. The repository is structured around two complementary codebases: A Standard Peeling Pipeline: Implements the core direction-dependent gain calibration and model subtraction to remove bright sources and flatten the imaging background. An Optimized Model-Restoration Peeling Pipeline: An optimized module that accurately restores the flux and morphology of scientifically interesting bright sources after peeling, mitigating artifacts while preserving key scientific signals. The provided Jupyter notebooks integrate these components to guide users through the complete workflow, from initial measurement sets to science-ready images. This Python-based, CASA-compatible framework is readily extensible to other mid- to low-frequency interferometric arrays. Modify only the "basic parameter settings" cell to run the entire pipeline. This is the main configuration cell. Modify these parameters to control the entire pipeline: 1. Target Field Settings: field = '3+4' -- Field name, e.g., '3+4' for combined observations 3 and 4. peeling_source_number = 3 -- Index of the source to be peeled, starting from the brightest one, 3 for the third brightest. output_number = peeling_source_number + 1 -- Index of the next source to be peeled. 2. TCLEAN Parameters: image_size = 4800 -- Image size in pixels, width = height. peeling_size = 128 -- Peeling size in pixels. cell_size='1arcsec' -- Pixel scale. threshold='0.05mJy' -- Stopping threshold for TCLEAN, ~3σ noise level. initial_niter=10000 -- Maximum iterations for initial imaging. modeling_niter = 20 -- Maximum iterations for source modeling. 3. Mask Parameters -- Target Bright Source Coordinates (Pixels): x = 3522 -- The X coordinate within the image_size × image_size pixel image. y = 1660 -- The Y coordinate within the image_size × image_size pixel image. 4. Phase Centers: initial_phasecenter = "J2000 02h08m06s +01d00m00s" -- Initial phase center. peeling_phasecenter = "J2000 02h06m51.1s +00d47m26.0s" -- Target bright source coordinates for peeling.