pith. sign in

arxiv: 2604.17448 · v1 · submitted 2026-04-19 · 🌌 astro-ph.IM · astro-ph.GA

VASCO: A fully automated CASA pipeline for large volume VLBI data calibration

A. Kumar (1 , 2) , C. Casadio (1 , M. Janssen (3) , D. \'Alvarez-Ortega (1 , F. M. P\"otzl (1 , 2) ((1) Institute of Astrophysics , Foundation for Research
show 12 more authors
Technology - Hellas Heraklion Greece (2) Department of Physics University of Crete (3) Department of Astrophysics Institute for Mathematics Astrophysics Particle Physics (IMAPP) Radboud University Nijmegen the Netherlands)
This is my paper

Pith reviewed 2026-05-10 05:45 UTC · model grok-4.3

classification 🌌 astro-ph.IM astro-ph.GA
keywords VLBICASAautomationcalibrationVLBAarchival datapipelinefringe detection
0
0 comments X p. Extension

The pith

VASCO shows CASA can fully automate calibration of heterogeneous archival VLBI data.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces the VASCO pipeline to automate the calibration of large volumes of Very Long Baseline Interferometry data from the VLBA using the CASA software. Traditionally this process requires significant human intervention at every stage, but VASCO handles preprocessing of different data formats, selects calibrators and reference antennas automatically via FFT-based fringe detection, runs the full calibration, and tracks progress with ALFRD. It was tested on 1000 archival sources from 1995 to 2023 across multiple bands, successfully producing calibrated output for 97.8 percent of them. A sympathetic reader would care because projects needing to process thousands of sources, like the Search for Milli-Lenses, make manual calibration impractical, and this automation could unlock analysis of decades of existing data.

Core claim

The VASCO pipeline automates the calibration of archival VLBA data in CASA by extending the rPICARD framework with automated preprocessing for FITS-IDI and Measurement Set formats, FFT-based fringe detection for calibrator and reference antenna selection, execution of the full calibration workflow, and real-time progress tracking via ALFRD. Validation on 1000 NRAO archival sources spanning 1995-2023 across S, C, X, U, and K bands produced calibrated output for 978 sources with a mean execution time of about 30 minutes per source using MPI parallelization.

What carries the argument

The FFT-based fringe detection for automatic calibrator and reference antenna selection, which enables blind operation without manual overrides across heterogeneous data.

If this is right

  • The automated selection will be incorporated into a future rPICARD release, extending blind calibration to any supported array.
  • Large volumes of archival VLBI data can now be processed with minimal human effort.
  • Mean processing time of 30 minutes per source allows scaling to thousands of observations.
  • Open-source availability enables community use and further development.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Similar automation techniques could be adapted for other radio astronomy arrays or data formats not covered here.
  • Processing historical VLBI data might reveal new insights into variable sources or long-term monitoring that were previously too labor-intensive.
  • Success on 97.8% suggests the method is robust but the 2.2% failures from corrupted data highlight the need for data quality checks upstream.

Load-bearing premise

The FFT-based fringe detection can reliably identify suitable calibrators and reference antennas for all types of VLBA data without needing manual adjustments or causing errors in the results.

What would settle it

A dataset where the automated pipeline selects a poor calibrator leading to phase errors or low signal-to-noise in the calibrated images, while a manual selection would succeed.

Figures

Figures reproduced from arXiv: 2604.17448 by 2), 2) ((1) Institute of Astrophysics, (2) Department of Physics, (3) Department of Astrophysics, A. Kumar (1, Astrophysics, C. Casadio (1, D. \'Alvarez-Ortega (1, F. M. P\"otzl (1, Foundation for Research, Greece, Heraklion, Institute for Mathematics, M. Janssen (3), Nijmegen, Particle Physics (IMAPP), Radboud University, Technology - Hellas, the Netherlands), University of Crete.

Figure 1
Figure 1. Figure 1: Complete VASCO pipeline workflow. The preprocessing stage (top, purple) operates on the raw FITS-IDI and MS data, handling source extraction, format corrections, phase center shift, averaging, and the automated selection of calibrators and refer￾ence antennas. The calibration stage (bottom, orange) follows the rPICARD scheme, executing in sequence amplitude calibration, single-band fringe-fitting for instr… view at source ↗
Figure 2
Figure 2. Figure 2: VLA flux density at 8 GHz versus total observing time [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Distribution of fringe solution percentages for the band [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 5
Figure 5. Figure 5: Mean execution rate per pipeline step for the 1000-source [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
read the original abstract

Calibrating large volumes of Very Long Baseline Interferometry (VLBI) data traditionally requires significant human intervention at every stage. While the Common Astronomy Software Applications (CASA) package is the standard data reduction tool across major radio observatories, no existing CASA-based pipeline operates in a fully automated manner across the heterogeneous data formats produced by the Very Long Baseline Array (VLBA) over three decades of operations. The Search for Milli-Lenses (SMILE) project, requiring the calibration of ~5000 VLBA sources, makes such blind automation a practical necessity. We introduce the VLBI and SMILE-based CASA Optimizations (VASCO) pipeline, which automates the calibration of archival VLBA data. VASCO extends the CASA-based rPICARD framework by automating preprocessing of FITS-IDI and Measurement Set data formats, calibrator and reference antenna selection via FFT-based fringe detection, and execution of the full calibration workflow. Progress tracking is handled by ALFRD (Automated Logical Framework for executing Dynamic scripts), which orchestrates pipeline execution and records results in real time. VASCO was validated on 1000 NRAO archival sources spanning 1995-2023, covering 1372 band-separated observations across the S, C, X, U, and K bands. Calibrated output was produced for 978 sources (97.8%), with 22 failures due to corrupted or incomplete data. Mean per-source execution time was ~30 minutes using MPI parallelization with up to 20 cores. VASCO demonstrates that fully blind calibration of heterogeneous archival VLBA data is achievable with CASA. The automated calibrator and reference antenna selection will be incorporated into a future rPICARD release, extending blind calibration to any supported array. VASCO and ALFRD are available as open-source Python packages.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 1 minor

Summary. The manuscript presents VASCO, a CASA-based pipeline extending the rPICARD framework to enable fully automated, blind calibration of large volumes of heterogeneous archival VLBA VLBI data. It automates preprocessing of FITS-IDI and Measurement Set formats, uses FFT-based fringe detection for calibrator and reference antenna selection, executes the full calibration workflow, and employs ALFRD for orchestration and progress tracking. Validation on 1000 NRAO archival sources (1995–2023, S/C/X/U/K bands, 1372 observations) produced calibrated outputs for 978 sources (97.8% success), with failures only from corrupted data; mean runtime was ~30 minutes per source using MPI on up to 20 cores. The pipeline is released open-source, with plans to incorporate the automated selection into rPICARD.

Significance. If the automated outputs prove scientifically reliable, VASCO would substantially lower the barrier to processing thousands of VLBI sources for projects such as SMILE, shifting calibration from labor-intensive manual work to scalable, reproducible pipelines. The large-scale test on real, multi-decade, multi-band archival data supplies concrete empirical support for blind automation feasibility within CASA. Open-source release and planned rPICARD integration add immediate community value by providing reusable infrastructure for VLBI arrays.

major comments (1)
  1. [Abstract and Validation results] Validation results (as summarized in the Abstract): The 97.8% success rate is defined solely by production of output files for 978/1000 sources, with no reported quantitative metrics on calibration quality (residual phase/amplitude errors, dynamic range, rms noise) or direct comparisons against manual CASA reductions of the same datasets. This limits evaluation of whether the FFT-based fringe detection and automated workflow produce results equivalent to human calibration across heterogeneous formats and conditions.
minor comments (1)
  1. [Abstract] The abstract would be strengthened by briefly stating the precise success criteria (file production only) and noting the absence of quality metrics in the current validation.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their positive summary of the manuscript and constructive recommendation for minor revision. We address the major comment on the validation results below.

read point-by-point responses

  1. Referee: [Abstract and Validation results] Validation results (as summarized in the Abstract): The 97.8% success rate is defined solely by production of output files for 978/1000 sources, with no reported quantitative metrics on calibration quality (residual phase/amplitude errors, dynamic range, rms noise) or direct comparisons against manual CASA reductions of the same datasets. This limits evaluation of whether the FFT-based fringe detection and automated workflow produce results equivalent to human calibration across heterogeneous formats and conditions.

    Authors: We agree that the validation presented focuses on the pipeline's success in completing automated calibration and producing output files for 978 of 1000 sources, with the 22 failures explicitly attributed to corrupted or incomplete input data. This metric was selected to demonstrate the feasibility of fully blind, scalable processing of heterogeneous archival VLBA data spanning multiple decades and bands, which is the core contribution for projects such as SMILE. The FFT-based fringe detection and automated workflow are shown to handle the full CASA calibration sequence without manual intervention in the vast majority of cases. However, we acknowledge that the manuscript does not currently provide quantitative metrics on calibration quality (such as residual phase/amplitude errors, dynamic range, or rms noise) or direct comparisons to manual CASA reductions of the same datasets. In the revised manuscript, we will add a new subsection in the validation section that includes such metrics and comparisons for a representative subset of sources across bands and epochs. This will allow a direct assessment of whether the automated results are equivalent to human calibration. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical pipeline validation on external data

full rationale

The paper presents an implemented software pipeline (VASCO) for automated VLBI calibration and validates it by running on 1000 independent archival NRAO VLBA datasets spanning 1995-2023. Success is defined and measured as production of output files for 978 sources (97.8%), with failures attributed to input corruption. No equations, predictions, or first-principles derivations are claimed; there are no fitted parameters renamed as predictions, no self-definitional loops, and no load-bearing self-citations that reduce the central claim to prior author work. The FFT-based fringe detection and calibrator selection are algorithmic choices whose performance is assessed against external data rather than by construction from the pipeline's own outputs. This is a standard engineering validation paper whose results are falsifiable by re-running on the same public archives.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The pipeline rests on standard domain assumptions about CASA functionality and VLBI data formats rather than new free parameters or invented physical entities.

axioms (2)
  • domain assumption CASA provides all required calibration tasks and can be scripted for batch processing of VLBI data
    Invoked as the foundation for extending rPICARD; treated as established background in radio astronomy.
  • domain assumption FFT-based fringe detection can identify usable calibrators and reference antennas from the visibility data without human judgment
    Central to the automation claim; appears in the description of the new selection step.

pith-pipeline@v0.9.0 · 5732 in / 1501 out tokens · 54428 ms · 2026-05-10T05:45:23.739675+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

23 extracted references

  1. [1]

    M., Kumar, A., & Janssen, M

    Álvarez-Ortega, D., Casadio, C., Pötzl, F. M., Kumar, A., & Janssen, M. 2025, A&A, 703, A196 Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al. 2013, A&A, 558, A33

    2025

  2. [2]

    Benson, J. M. 1995, in Astronomical Society of the Pacific Conference Series, V ol. 82, Very Long Baseline Interferometry and the VLBA, ed. J. A. Zensus, P. J. Diamond, & P. J. Napier, 117

    1995

  3. [3]

    B., et al

    Blackburn, L., Chan, C.-k., Crew, G. B., et al. 2019, ApJ, 882, 23

    2019

  4. [4]

    Casadio, C., Blinov, D., Readhead, A. C. S., et al. 2021, MNRAS, 507, L6

    2021

  5. [5]

    T., Tingay, S

    Deller, A. T., Tingay, S. J., Bailes, M., & West, C. 2007, PASP, 119, 318

    2007

  6. [6]

    Greisen, E. W. 2003, in Astrophysics and Space Science Library, V ol. 285, In- formation Handling in Astronomy - Historical Vistas, ed. A. Heck, 109

    2003

  7. [7]

    P., Bregman, J

    Hamaker, J. P., Bregman, J. D., & Sault, R. J. 1996, A&AS, 117, 137

    1996

  8. [8]

    A., Lee, S.-S., Zhao, G.-Y ., et al

    Hodgson, J. A., Lee, S.-S., Zhao, G.-Y ., et al. 2016, Journal of Korean Astro- nomical Society, 49, 137

    2016

  9. [9]

    M., et al

    Janssen, M., Goddi, C., van Bemmel, I. M., et al. 2019, A&A, 626, A75 Article number, page 8 A. Kumar et al.: V ASCO: A fully automated CASA pipeline for large volume VLBI data calibration

    2019

  10. [10]

    F., & Wagner, J

    Janssen, M., Radcliffe, J. F., & Wagner, J. 2022, Universe, 8, 527

    2022

  11. [11]

    P., et al

    Kim, D.-W., Janssen, M., Krichbaum, T. P., et al. 2023, A&A, 680, L3 Martí-Vidal, I. 2010, A&A, 517, A83

    2023

  12. [12]

    P., Waters, B., Schiebel, D., Young, W., & Golap, K

    McMullin, J. P., Waters, B., Schiebel, D., Young, W., & Golap, K. 2007, in As- tronomical Society of the Pacific Conference Series, V ol. 376, Astronomical Data Analysis Software and Systems XVI, ed. R. A. Shaw, F. Hill, & D. J. Bell, 127

    2007

  13. [13]

    T., Jackson, N

    Myers, S. T., Jackson, N. J., Browne, I. W. A., et al. 2003, MNRAS, 341, 1

    2003

  14. [14]

    Offringa, A. R. 2010, AOFlagger: RFI Software, Astrophysics Source Code Li- brary, record ascl:1010.017, ascl:1010.017

    2010

  15. [15]

    1999, in Astronomical Society of the Pacific Conference Series, V ol

    Pence, W. 1999, in Astronomical Society of the Pacific Conference Series, V ol. 172, Astronomical Data Analysis Software and Systems VIII, ed. D. M

    1999

  16. [16]

    Petrov, L. Y . & Kovalev, Y . Y . 2025, ApJS, 276, 38 Pötzl, F. M., Casadio, C., Kalaitzidakis, G., et al. 2025, A&A, 695, A169

    2025

  17. [17]

    2002, arXiv e-prints, astro

    Reynolds, C., Paragi, Z., & Garrett, M. 2002, arXiv e-prints, astro

    2002

  18. [18]

    Schwab, F. R. & Cotton, W. D. 1983, AJ, 88, 688

    1983

  19. [19]

    Shepherd, M. C. 1997, in Astronomical Society of the Pacific Conference Series, V ol. 125, Astronomical Data Analysis Software and Systems VI, ed. G. Hunt & H. Payne, 77

    1997

  20. [20]

    & Moellenbrock, G

    Small, D. & Moellenbrock, G. 2022, CASA Memo 12: The Fringefit Task in

    2022

  21. [21]

    Smirnov, O. M. 2011, A&A, 527, A106 The CASA Team et al. 2022, Publications of the Astronomical Society of the Pacific, 134, 114501

    2011

  22. [22]

    R., Moran, J

    Thompson, A. R., Moran, J. M., & Swenson, Jr., G. W. 2017, Interferometry and Synthesis in Radio Astronomy, 3rd Edition, 3rd edn. (Cham :: Springer International Publishing) van Bemmel, I. M., Kettenis, M., Small, D., et al. 2022, PASP, 134, 114502 van Diepen, G. N. J. 2015, Astronomy and Computing, 12, 174

    2017

  23. [23]

    N., Henstock, D

    Wilkinson, P. N., Henstock, D. R., Browne, I. W., et al. 2001, Phys. Rev. Lett., 86, 584 Article number, page 9

    2001