Can More Researchers Use ALMA's Black Hole Polarization Data? A New Archive Says Yes

During global VLBI campaigns, the Atacama Large Millimeter/submillimeter Array in Chile does something unusual. A subset of its 12-meter dishes is electronically tied together, or phased, so those antennas behave like one large station. In that mode, ALMA joins very long baseline interferometry (VLBI) networks such as the Event Horizon Telescope and the Global Millimeter VLBI Array, improving their sensitivity and resolution enough to help image the faint emission around supermassive black holes.

Those campaigns leave behind two kinds of data. One is the global VLBI recording used for the famous black hole images. The other is ALMA's own interferometric visibilities: the interference measurements from the phased array. The second kind is scientifically rich, especially when it carries polarization information, but it is difficult to calibrate and image. A new project called VAPOLA aims to change that.

A Pipeline for ALMA's Overlooked VLBI Visibilities

ALMA's phased mode, called the ALMA Phasing System, turns 37 to 42 of its 12-meter antennas into a single station; the 7-meter Atacama Compact Array antennas are excluded. The data structures differ from normal ALMA observations, and polarization calibration adds another layer of complexity. Exploiting these data has required specialized expertise and substantial manual effort. The paper is by Alejandro Mus, Ciriaco Goddi, and colleagues, and it presents VAPOLA, described as the first online, multi-epoch, multi-band repository of high-level ALMA products from these global VLBI campaigns. VAPOLA is built on an automated pipeline that starts from fully calibrated ALMA (QA2) data and produces science-ready products with minimal user intervention. The paper describing the repository has been posted as a preprint on arXiv.

VAPOLA stands for VLBI AGN POLarization data with ALMA. It covers observations from 2017 onward, across ALMA Bands 3, 6, and 7, which correspond to wavelengths near 3 millimeters, 1.3 millimeters, and 0.87 millimeters. It focuses on active galactic nuclei—the bright, compact cores of galaxies powered by supermassive black holes—and on Sgr A*, the Milky Way's own central black hole. The repository is hosted at the Italian Centre for Astronomical Archives (IA2) and is designed to grow as new observations become publicly available. All non-proprietary VAPOLA data products are public; proprietary datasets require authentication.

What VAPOLA Hands Researchers

The archive is not a collection of raw telescope files. It provides calibrated visibilities, both averaged into a single channel and retaining all 240 spectral channels. It also provides full-Stokes images: maps of total intensity (Stokes I), two linear polarization states (Q and U), and circular polarization (V). These are made for each individual spectral window, for combined sidebands (pairs of spectral windows), and for the full band. From those images, the pipeline derives linear and circular polarization fractions, polarization angle, spectral index, depolarization, and rotation measure, a quantity that traces magnetized plasma along the line of sight.

The repository also includes tabulated polarization parameters for compact source cores, extracted from visibility-domain fitting. Pre-rendered maps, diagnostic plots, and error maps accompany the main products. Users can download individual files or assemble custom bundles through a web portal. The goal is to let researchers who are not ALMA calibration experts ask questions about magnetic fields, jets, and accretion flows without first becoming experts in the instrument's phased-array quirks.

Sgr A* is a special case because it varies on timescales of minutes. Standard calibration can mistake that variability for a calibration problem. VAPOLA includes a dedicated procedure called minispiral calibration. It uses the surrounding, quasi-stationary minispiral of gas as a reference, subtracts the extended emission, and then fits the compact, time-variable core at each integration. The result is time-resolved light curves in total intensity, linear polarization, and polarization angle for each spectral window. These products complement recent Event Horizon Telescope polarization studies of Sgr A*'s ring and its variability. They also make it possible to search for orbiting hotspots, magnetic reconnection events, and other rapid changes near the event horizon. For Sgr A*, the pipeline deliberately avoids aggressive automated flagging so that it does not throw away time coverage; only visibilities with antenna elevations below 25 degrees are flagged.

Flux Limits, Polarization Fidelity, and Versioned Releases

The paper is explicit about limits. VAPOLA processes ALMA's own interferometric data from VLBI campaigns, not the VLBI baseband recordings or the globally correlated VLBI visibilities themselves. Absolute flux calibration carries uncertainties of about 5 percent in Band 3 and about 10 percent in Bands 6 and 7. ALMA's linear polarization fidelity is guaranteed at the 0.1 percent level of total intensity (Stokes I) only within the inner third of the primary beam, the telescope's main field of view; farther out, instrumental effects can push systematic errors to roughly 0.3 to 0.5 percent near the primary-beam FWHM. Circular polarization is especially challenging to detect.

The archive is also versioned. As calibration scripts improve, the team plans to reprocess data and issue new releases that supersede older ones. For some quantities derived by fitting across frequencies, such as rotation measure, revised values may differ from earlier papers by more than the 1σ fit uncertainties. The repository documents those changes, but users should check versions carefully. And because this is a data-resource paper, it does not claim a new measurement of a black hole or a new test of gravity.

What VAPOLA offers is infrastructure: a way to turn hard-won observations into reusable science. The same data can probe magnetic fields in accretion flows and relativistic jets, map dusty molecular tori around active galaxies, study interstellar molecules toward the Galactic Center, and search for high-frequency transients. The paper notes, for example, that absorption lines from molecules such as HCN, HCO+, CS, and CN appear in these spectra, including CN near 226.91 GHz toward Sgr A*. As black hole astronomy moves toward time-domain and long-term monitoring, the bottleneck is no longer only collecting light. It is making that light usable by a wider community. VAPOLA is an attempt to do exactly that: to take data gathered at the edge of technical possibility and turn it into something a graduate student, a theorist, or an observer in another field can actually open, understand, and use.