Metadata for the Flux Density Calibration of the April 2021 and March 2022 Event Horizon Telescope Science Data

arXiv:2611.16606 · astro-ph.IM · Submitted 2026-11-12 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Astrophysics. The day in astrophysics: observations, models and what the data forces.

Vera: We're looking at the paper: "Metadata for the Flux Density Calibration of the April 2021 and March 2022 Event Horizon Telescope Science Data".

Jocelyn: Here's the gist of the paper. The paper reports on the metadata associated with observations from April 2021 and March 2022 at 1.3 mm wavelengths for the Event Horizon Telescope (EHT) science data, highlighting differences in sensitivity treatment and array changes since 2018.

Vera: Campaign Overview The EHT has been conducting science observations annually since 2021 after the initial campaigns in 2017 and 2018 <ref:2611.16606#pg4>.

Jocelyn: Significant adjustments were needed to enable a new EHT campaign, such as controlling some telescopes remotely <ref:2611.16606#pg5>.

Vera: For example, the Greenland Telescope (GLT) operations required coordination of efforts between the GLT staff in Taiwan and engineers employed by the Pituffik Space Base in Greenland (Chen et al., 2023) <ref:2611.16606#pg5>.

Paper summary: Vera: So, looking at this paper, "Metadata for the Flux Density Calibration of the April two thousand twenty-one and March two thousand twenty-two Event Horizon Telescope Science Data," what it really boils down to is that they’ve documented all the technical differences in how they calculated sensitivity and gain curves for different stations since two thousand eighteen <ref:2611.16606#pg1,Metadata for the Flux Density Calibration of the April 2021 and March>.

Jocelyn: It means that when we look at the actual science data from those years, we need this detailed metadata to correctly calibrate the flux density measurements.

Subrahmanyan: And the implication for cosmology is that having these precise calibration tables allows us to get a more accurate measurement of the signals we see, which feeds into our understanding of how fast the universe is expanding near us <ref:two thousand six hundred eleven point one six six zero six#pg3.

Vera: It’s about making sure the numbers we use to interpret those observations are as solid as possible, accounting for all those array changes and different instruments.

Jocelyn: The title of this paper highlights that it's specifically about the flux density calibration, which is the first step in getting any meaningful astrophysical result from these EHT observations.

Subrahmanyan: So we’re looking at how robust this calibration framework is when you combine data from so many different single dishes and phased arrays <ref:two thousand six hundred eleven point one six six zero six#pg5.

Vera: It confirms that the process of collecting and processing metadata needs to evolve alongside the growing complexity of the EHT array.

Jocelyn: It’s a necessary step for anyone looking at those raw one point three mm wavelength observations from April two thousand twenty-one and March two thousand twenty-two data to understand how they were measured <ref:2611.16606#pg2>.

Conclusion: Vera: So, this paper is all about the metadata they created for their April twenty twenty one and March twenty twenty two science data from the Event Horizon Telescope observations at one point three millimeter wavelengths.

Jocelyn: It’s essentially a technical manual for how they calibrated those measurements.

Subrahmanyan: The core focus here is showing how sensitivity treatments and array changes have evolved since eighteen.

Vera: Exactly. They detail the differences in sensitivity treatment and how things have changed with the array since eighteen to calibrate their flux density data correctly.

Jocelyn: It’s important because that calibration directly impacts what we actually measure from those black hole images, which is crucial for time domain radio astronomers like myself looking at pulsar signals.

Subrahmanyan: From a theoretical standpoint, these details are what allow us to connect the observed flux density back to the physical structure of spacetime and how matter behaves near extreme gravity.

Vera: So, it’s about making sure that when we look at those images, the numbers we get from them are trustworthy because they accounted for all those telescope and array adjustments.

Jocelyn: And it shows the ongoing effort to keep up with a constantly changing observational setup across so many different stations.

Subrahmanyan: The implication is that as the EHT keeps growing and adding new stations, this kind of detailed metadata becomes absolutely necessary for reliable cosmological constraints on structure formation.

Vera: It’s a bit of an engineering deep dive into the data preparation process, but it underpins everything we try to figure out about black holes using these telescopes.

Jocelyn: It gives us a clearer picture of the observational pipeline's current state and what they need to fix next.

Subrahmanyan: So, if you want to understand how reliable those early EHT measurements are, you have to look at this level of calibration detail.

Vera: And that leads us into some of the specific issues they found with the metadata tables themselves.

C. Romero-Cañizales, J. Y. Koay, L. Blackburn, L. D. Matthews, V. Fish, R. P. J., Tilanus, G., B., , , , , , ,

Department of Astrophysics IMAPP Radboud University Academia Sinica Institute of Astronomy and Astrophysics Niigata University Center for Astrophysics Harvard & Smithsonian Black Hole Initiative at Harvard University Massachusetts Institute of Technology Haystack Observatory Steward Observatory and Department of Astronomy University of Arizona Institut de Radio Astronomie Millimétrique East Asian Observatory Instituto de Astronomia Geofísica e Ciências Atmosféricas Dipartimento di Fisica Universitá degli Studi di Cagliari INAF Osservatorio Astronomico di Cagliari Max Planck Institut für Astronomie Max Planck Institut für Radioastronomie European Southern Observatory Canadian Institute for Theoretical Astrophysics

astro-ph.IM

Submitted: 2026-11-12

Updated: 2026-11-12

Comments: 23 pages, 5 figures, EHT Memo Series 2026-L1-01

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 91/100

The gist: The paper reports on the metadata associated with observations from April 2021 and March 2022 at 1.3 mm wavelengths for the Event Horizon Telescope (EHT) science data…

Key concepts

System Noise Temperature (T*sys)
This represents the overall noise level of a telescope system, accounting for atmospheric effects. It is approximated by multiplying the atmospheric opacity ($ au$) by the system noise temperature ($T_{sys}$). This value is crucial for calculating how sensitive a telescope is to receiving faint signals.
Degrees per Flux Unit (DPFU)
DPFU measures a single dish's characteristic sensitivity, expressed in degrees per flux unit. It combines factors like antenna area and system parameters. This metric helps quantify the raw data quality provided by individual telescopes for calibration purposes.
Field Rotation Angle ($\phi$)
This angle describes how the signal arrives at a telescope's receiver based on its mount type and orientation relative to the sky. It is calculated using several components, including parameters related to the Nasmyth configuration and cabin rotation, which affects how different parts of the array see celestial sources.

Terminology

Summary

The paper reports on the metadata associated with observations from April 2021 and March 2022 at 1. p. 13 mm wavelengths for the Event Horizon Telescope (EHT) science data, highlighting differences in sensitivity treatment and array changes since 2018. p. 2

Campaign Overview

The EHT has been conducting science observations annually since 2021 after the initial campaigns in 2017 and 2018 <ref:2611.16606#pg4>. Significant adjustments were needed to enable a new EHT campaign, such as controlling some telescopes remotely <ref:2611. p. 216606#pg5>. For example, the Greenland Telescope (GLT) operations required coordination of efforts between the GLT staff in Taiwan and engineers employed by the Pituffik Space Base in Greenland (Chen et al. p. 2, 2023) <ref:2611. p. 016606#pg5>. In addition, both the NOrthern Extended Millimeter Array (NOEMA) operated by the Institut de Radio Astronomie Millimétrique (IRAM) in France, and the University of Arizona 12-meter Telescope at Kitt Peak (KT) joined the EHT observations for the first time in that year <ref:2611. p. 216606#pg5>. A total of 10 and 11 stations participated in the 2021 and 2022 EHT campaigns, respectively <ref:2611. p. 216606#pg5>. This makes a total of eight single-dish stations (APEX, IRAM 30 m, SPT, LMT, SMT, JCMT, GLT and KT) and three phased arrays (ALMA, SMA and NOEMA) <ref:2611. p. 216606#pg5>.

Sensitivity Estimates

The antenna sensitivities needed for the ANTAB tables take into account the changes in effective system temperature (T∗sys; Eq. A2) with elevation <ref:2611. p. 016606#pg6>. The maximum gain is assumed to be constant during the campaign for a given station, frequency band, and polarization <ref:2611. p. 316606#pg6>. The gain curve (gE) is normally expressed by a second-order polynomial of elevation (E) <ref:2611. p. 316606#pg7>. Each station provides the information needed to obtain their characteristic sensitivity, or directly provides the DPFUs and gain curve coefficients (see details in Section 3) <ref:2611. p. 316606#pg7>. The DPFU values are given per sideband and polarization; these represent combined values from individual antennas <ref:2611. p. 416606#pg2>.

Station Characteristics

Each station in the EHT array provides raw metadata in a standard format, but there are still some peculiarities in each package that need to be considered <ref:2611.16606#pg3>. Details of the information provided by each station and how this is used to assemble the EHT ANTAB tables are described in subsections <ref:2611. p. 416606#pg3>. For instance, for ALMA, the phased-ALMA calibration workflow is described in detail in Goddi et al. p. 5 (2019) <ref:2611. p. 016606#pg4>. The NOEMA team provides metadata packages containing individual antenna gains and various data per observing track <ref:2611. p. 616606#pg14>.

Pipeline Updates and Issues

The pipeline has been improved to process the metadata from new stations in the EHT array and further automate some tasks <ref:2611.16606#pg19>. The current pipeline consists of a script with functions to read the phased-array metadata from ALMA, NOEMA and SMA <ref:2611. p. 216606#pg19>. A main script reads and validates the metadata from single-dish stations, and calls the two other scripts to produce validation plots, ancillary files, and ANTAB tables <ref:2611.16606#pg19>. Remaining issues include that the ANTAB tables do not contain ALMA T∗sys values for scans on weak targets observed with passive phasing <ref:2611. p. 616606#pg3>.

Field Rotation Dependence

The signal arriving at the main dish (or primary reflector) of each telescope follows a different path to the receiver, depending on the telescope’s mount type <ref:2611.16606#pg20>. Different orientations of the receiver antenna with respect to the celestial sky are captured by the so-called field rotation angle <ref:2611. p. 116606#pg20>. The drive type for all EHT telescopes is Alt–Az, which introduces a rotation of the cabin with respect to the celestial sky as a source is being tracked <ref:2611.16606#pg20>.

Formulae and Nomenclature

Basic formulae describe the relationship among parameters such as Atmospheric opacity at zenith τ = τ0/sin(E) (A1). The system noise temperature T∗sys is approximated by T∗sys ≈ eτ Tsys (A2). p. 3 The Degrees per flux unit for a single dish telescope is defined by DPFUsd = ηA Ageom 2kB (A3) <ref:2611. p. 316606#pg15>. The System equivalent flux density for a single dish telescope is SEFDsd = T∗sys gE DPFUsd (A5) <ref:2611. p. 316606#pg15>.

Metadata Package Contents

The metadata packages in this L1 release are named EHTmetadata 2021April v1.2 and EHTmetadata 2022March v1.0 <ref:2611. p. 016606#pg6>. These tarball packages contain several products, including Raw metadata files provided by telescope representatives <ref:2611. p. 716606#pg18>, ANTAB tables per track and per EHT frequency band <ref:2611. p. 616606#pg18>, Validation plots in the plots/ directory <ref:2611. p. 016606#pg18>, and Supplementary files produced by the pipeline in the ancillary/ directory <ref:2611.16606#pg18>.

Remaining Issues

A common issue is that the ANTAB tables do not contain ALMA T∗sys values for scans on weak targets observed with passive phasing, and end users may need to use a nearby calibrator <ref:2611.16606#pg3>. For the 2021 metadata package, the GLT modeled T∗sys values for track e21e13 were obtained under incorrect assumptions in version v1. p. 71 <ref:2611. p. 016606#pg19>. The ANTAB tables for track e21b09 in the v1. p. 71 package do not indicate that APEX observed in linear polarization <ref:2611. p. 716606#pg3>. In the 2022 metadata package, there are substantial changes to the LMT surface since 2018, leading to an assumption of a flat gain curve without an uncertainty estimate <ref:2611. p. 916606#pg20>. The DPFUs for IRAM 30 m in the ANTAB tables implicitly include the official gain curve instead of the one reported here <ref:2611. p. 816606#pg20>.

Field Rotation Parameters

The field rotation angle ϕ is defined as ϕ = fpar ψpar + fel θel + ϕoff (6) <ref:2611.16606#pg21>. Table 5 provides field rotation parameters for the 230 GHz receivers at the EHT stations participating in the 2021–2022 campaigns <ref:2611. p. 416606#pg20>. The sign of the dependence is related to the Nasmyth configuration and number of reflections prior to the Nasmyth cabin <ref:2611.16606#pg20>.

Naming Conventions

There are different naming conventions for telescopes and for observing tracks which may be preferred by different calibration pipelines and plotting libraries. VEX stands for “VLBI Experiment” and HOPS stands for the “Haystack Observatory Postprocessing System” software. p. 9 Table 6 lists the names and identification codes for all the telescopes that have participated in EHT science observations in 2021 and 2022 <ref:2611. p. 116606#pg20>. Table 7 shows VEX and EHT experiment identification numbers for each track of the 2021 and 2022 EHT observing campaigns, along with participating stations <ref:2611. p. 316606#pg20>. The paper also includes a script custom antab. p. 6py that produces customized ANTAB tables per track and per source <ref:2611. p. 116606#pg18>.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved metadata processing pipeline can automate several tasks including A script with functions to read the phased-array metadata from ALMA, NOEMA and SMA and A main script that reads and validates the metadata from single-dish stations. This automation will reduce manual effort in creating files like ANTAB tables per track and per EHT frequency band.

  2. The system can now perform complex gain curve estimations by using a script that estimates DPFUs, elevation- and time-dependent gain curves, as well as their uncertainties. This allows the AI to directly input gain information from files provided by telescope representatives in non-standard format and automatically create necessary ANTAB headers.

  3. The system can handle missing values by implementing a strategy where it simply interpolate or extrapolate neighboring values to find missing ones, rather than just modeling when there are no values from which to interpolate or extrapolate. This directly addresses issues like GLT modeled Tsys values for track e21e13 by using typical GLT values for Tsys rx as well as real weather conditions on 13 April.

  4. The system can automatically handle polarization labeling inconsistencies by ensuring that it corrects labels, such as the point that Correct labels appear only in the ANTAB tables for APEX observations observed in linear polarization.

Sources

Related papers