SYMBA : an end-to-end VLBI synthetic data generation pipeline simulating event horizon telescope observations of M 87

Freek Roelofs, Michaël Janssen, Iniyan Natarajan, Roger Deane, Jordy Davelaar, Héctor Olivares, Oliver Porth, Scott Paine, Katherine L. Bouman, R. P. J. Tilanus, Ilse van Bemmel, H. Falcke, Kazunori Akiyama, A. Alberdi, W. Alef, Keiichi Asada, Rebecca Azulay, Anne-Kathrin Baczko, D. R. Ball, Mislav Baloković, John Barrett, Dan Bintley, Lindy Blackburn, W. Boland, Geoffrey C. Bower, Michael Bremer, Christiaan D. Brinkerink, Roger Brissenden, S. Britzen, Avery E. Broderick, Dominique Broguière, Thomas Bronzwaer, Do‐Young Byun, J. E. Carlstrom, Andrew Chael, Chi‐kwan Chan, Shami Chatterjee, Koushik Chatterjee, Ming‐Tang Chen, Y. Chen, Ilje Cho, Pierre Christian, J. E. Conway, J. M. Cordes, G. Crew, Yuzhu Cui, Mariafelicia De Laurentis, Jessica Dempsey, G. Desvignes, Jason Dexter, Sheperd S. Doeleman, Ralph P. Eatough, Vincent L. Fish, E. B. Fomalont, Raquel Fraga-Encinas, Per Friberg, Christian M. Fromm, J. L. Gómez, Peter Galison, Charles F. Gammie, Roberto García, Olivier Gentaz, Boris Georgiev, C. Goddi, Roman Gold, Minfeng Gu, Mark Gurwell, Kazuhiro Hada, M. H. Hecht, Ronald Hesper, Luis C. Ho, Paul T. P. Ho, Mareki Honma, Wu Jiang, Li Huang, D. H. Hughes, Shiro Ikeda, Makoto Inoue, Sara Issaoun, D. J. James, Buell T. Jannuzi, Britton Jeter, Abhishek V. Joshi, Michael D. Johnson, Svetlana G. Jorstad, Taehyun Jung, Mansour Karami, R. Karuppusamy, Tomohisa Kawashima, Garrett K. Keating, Mark Kettenis, J. Kim, Motoki Kino, Jun Yi Koay, Patrick M. Koch, Shoko Koyama, M. Krämer, C. Krämer, T. P. Krichbaum, Cheng‐Yu KuoUvA-DARE 2020 · DOI

Abstract

Realistic synthetic observations of theoretical source models are essential for understanding real observational data and verifying model-based predictions. We present the SYnthetic Measurement creator for long Baseline Arrays (SYMBA), a novel synthetic data generation pipeline for Very Long Baseline Interferometry (VLBI). SYMBA incorporates realistic atmospheric, instrumental, and calibration effects to simulate Event Horizon Telescope (EHT) observations. Using general relativistic magnetohydrodynamics (GRMHD) models of M 87, we demonstrate that while atmospheric and instrumental corruptions are significant, standard calibration techniques can robustly recover source structures. Furthermore, planned array expansions will significantly improve angular resolution and dynamic range.

The Challenge of Realistic VLBI Simulation

Capturing the first image of a black hole shadow (M87*) was a landmark achievement, but the data provided by Earth-sized radio telescope arrays is inherently "dirty." At millimeter wavelengths, the Earth’s atmosphere introduces rapid phase fluctuations, while telescopes suffer from pointing errors and instrumental noise.

Historically, simulations used to test imaging algorithms often relied on overly optimistic models, typically adding only simple thermal noise. This "reality gap" can lead to overconfidence in image reconstruction. To trust that a reconstructed image accurately represents complex physics—such as black hole spin or jet structure—researchers require an end-to-end simulation that mirrors the entire observational and calibration workflow under realistic conditions.

The SYMBA Pipeline Architecture

SYMBA addresses this gap by integrating specialized tools into a unified, dockerized pipeline. The process begins with a theoretical source model and proceeds through three primary stages:

  1. MeqSilhouette (Corruption): This stage applies realistic "corruptions," including tropospheric effects (opacity and turbulent phase noise), receiver noise, and antenna pointing errors (e.g., "wind shake").
  2. rPICARD (Calibration): A CASA-based pipeline that treats synthetic data as real, performing fringe fitting to solve for phase, rate, and delay offsets.
  3. eht-imaging (Reconstruction): The final stage uses Regularized Maximum Likelihood (RML) methods to reconstruct the image from calibrated visibilities.
The SYMBA pipeline architecture, illustrating the flow from initial source models and antenna information through corruption, calibration, and final image reconstruction.

Insights from Synthetic Observations

The application of SYMBA to M 87 models revealed several critical insights into the nature of VLBI data:

  • Atmospheric Dominance: Atmospheric turbulence and pointing errors are far more detrimental to data quality than thermal noise. Without robust fringe fitting, absolute phase information is quickly lost.
  • Calibration Robustness: Despite heavy corruption, standard techniques like network calibration and self-calibration successfully recover the underlying "shadow" structure of the black hole.
  • Sensitivity to Pointing: Large dishes, such as the 32m Large Millimeter Telescope (LMT), are particularly sensitive to pointing errors. Even minor drifts can lead to significant, unrecoverable signal loss due to their narrow beams.

Future Array Configurations and Scientific Potential

A key application of SYMBA is evaluating the impact of future array expansions. While the 2017 EHT array successfully imaged the black hole's "ring," it faced challenges in distinguishing between different physical models, such as thermal versus non-thermal jet models.

The uv-coverage of the EHT array, showing the baseline distribution for the 2017 configuration and the substantial improvements provided by adding new stations like the Greenland Telescope (GLT) and the Africa Millimetre Telescope (AMT).

By simulating the "EHT2020+" configuration—which includes additional stations like the Greenland Telescope (GLT) and the Africa Millimetre Telescope (AMT)—SYMBA demonstrates a marked increase in uv-coverage. These improvements provide the necessary resolution and dynamic range to differentiate between competing GRMHD models based on salient features in the reconstructed images.

Conclusion

SYMBA provides a rigorous framework for testing the limits of VLBI imaging. By bridging the gap between theoretical models and realistic observational constraints, it ensures that the next generation of black hole images is grounded in a deep understanding of instrumental and environmental uncertainties. Future work will focus on modeling time-variable sources and incorporating ionospheric effects for lower-frequency observations.

Summary updated

Abstract

CONTEXT: Realistic synthetic observations of theoretical source models are essential for our understanding of real observational data. In using
\nsynthetic data, one can verify the extent to which source parameters can be recovered and evaluate how various data corruption effects can be
\ncalibrated. These studies are the most important when proposing observations of new sources, in the characterization of the capabilities of new or
\nupgraded instruments, and when verifying model-based theoretical predictions in a direct comparison with observational data.
\nAIMS: We present the SYnthetic Measurement creator for long Baseline Arrays (SYMBA), a novel synthetic data generation pipeline for Very Long
\nBaseline Interferometry (VLBI) observations. SYMBA takes into account several realistic atmospheric, instrumental, and calibration effects.
\nMETHODS: We used SYMBA to create synthetic observations for the Event Horizon Telescope (EHT), a millimetre VLBI array, which has recently
\ncaptured the first image of a black hole shadow. After testing SYMBA with simple source and corruption models, we study the importance of
\nincluding all corruption and calibration effects, compared to the addition of thermal noise only. Using synthetic data based on two example general
\nrelativistic magnetohydrodynamics (GRMHD) model images of M 87, we performed case studies to assess the image quality that can be obtained
\nwith the current and future EHT array for different weather conditions.
\nRESULTS: Our synthetic observations show that the effects of atmospheric and instrumental corruptions on the measured visibilities are significant.
\nDespite these effects, we demonstrate how the overall structure of our GRMHD source models can be recovered robustly with the EHT2017 array
\nafter performing calibration steps, which include fringe fitting, a priori amplitude and network calibration, and self-calibration. With the planned
\naddition of new stations to the EHT array in the coming years, images could be reconstructed with higher angular resolution and dynamic range.
\nIn our case study, these improvements allowed for a distinction between a thermal and a non-thermal GRMHD model based on salient features in
\nreconstructed images.
BibTeX
@inproceedings{roelofs2020symba,
  author = {Freek Roelofs and Michaël Janssen and Iniyan Natarajan and Roger Deane and Jordy Davelaar and Héctor Olivares and Oliver Porth and Scott Paine and Katherine L. Bouman and R. P. J. Tilanus and Ilse van Bemmel and H. Falcke and Kazunori Akiyama and A. Alberdi and W. Alef and Keiichi Asada and Rebecca Azulay and Anne-Kathrin Baczko and D. R. Ball and Mislav Baloković and John Barrett and Dan Bintley and Lindy Blackburn and W. Boland and Geoffrey C. Bower and Michael Bremer and Christiaan D. Brinkerink and Roger Brissenden and S. Britzen and Avery E. Broderick and Dominique Broguière and Thomas Bronzwaer and Do‐Young Byun and J. E. Carlstrom and Andrew Chael and Chi‐kwan Chan and Shami Chatterjee and Koushik Chatterjee and Ming‐Tang Chen and Y. Chen and Ilje Cho and Pierre Christian and J. E. Conway and J. M. Cordes and G. Crew and Yuzhu Cui and Mariafelicia De Laurentis and Jessica Dempsey and G. Desvignes and Jason Dexter and Sheperd S. Doeleman and Ralph P. Eatough and Vincent L. Fish and E. B. Fomalont and Raquel Fraga-Encinas and Per Friberg and Christian M. Fromm and J. L. Gómez and Peter Galison and Charles F. Gammie and Roberto García and Olivier Gentaz and Boris Georgiev and C. Goddi and Roman Gold and Minfeng Gu and Mark Gurwell and Kazuhiro Hada and M. H. Hecht and Ronald Hesper and Luis C. Ho and Paul T. P. Ho and Mareki Honma and Wu Jiang and Li Huang and D. H. Hughes and Shiro Ikeda and Makoto Inoue and Sara Issaoun and D. J. James and Buell T. Jannuzi and Britton Jeter and Abhishek V. Joshi and Michael D. Johnson and Svetlana G. Jorstad and Taehyun Jung and Mansour Karami and R. Karuppusamy and Tomohisa Kawashima and Garrett K. Keating and Mark Kettenis and J. Kim and Motoki Kino and Jun Yi Koay and Patrick M. Koch and Shoko Koyama and M. Krämer and C. Krämer and T. P. Krichbaum and Cheng‐Yu Kuo},
  title = {{SYMBA : an end-to-end VLBI synthetic data generation pipeline simulating event horizon telescope observations of M 87}},
  booktitle = {UvA-DARE},
  year = {2020},
  doi = {10.1051/0004-6361/201936622},
}
Figure