Event Horizon Telescope Data Overview

• Event Horizon Telescope data is a global VLBI system that produces interferometric measurements to probe event-horizon-scale phenomena in black holes.
• It employs coherent averaging, closure phases, and dual-frequency observations to achieve high-resolution imaging and accurate calibration.
• The data enable precise modeling of accretion processes, photon ring detection, and monitoring of variability in supermassive black holes.

The Event Horizon Telescope (EHT) is a global very long baseline interferometry (VLBI) array operating at submillimeter wavelengths, specifically designed to resolve event-horizon-scale structure around supermassive black holes such as Sagittarius A* (Sgr A*) and M87*. The EHT data encompass an interconnected set of interferometric measurements, calibration strategies, and imaging products that directly probe strong-field gravitational physics, accretion processes, and jet launching at horizon scales. This article provides a comprehensive examination of the structure, calibration, analysis techniques, and scientific utilization of EHT data products.

1. EHT Array Topology, (u, v)-Coverage, and Observational Frequencies

The EHT array comprises geographically distributed stations spanning four continents and the South Pole. For its "Complete" configuration (as of 2014 and realized in subsequent campaigns), principal sites include SMA and JCMT in Hawaii, ALMA, APEX, ASTE in Chile, SMTO in Arizona, CARMA in California, LMT in Mexico, Pico Veleta and Plateau de Bure in Europe, and the South Pole Telescope. Projected baseline lengths span from ∼2,000 km (e.g., CARMA–LMT) up to ∼7,000–9,000 km (Chile–Hawaii, Chile–South Pole), mapping to maximum baseline lengths $B_{\rm max}$ of order $5$–$8\times 10^{6}\,\lambda$ at $230$ GHz and $7$–$12\times10^{6}\,\lambda$ at $345$ GHz.

The interferometric angular resolution is governed by $\theta \simeq \lambda/B_{\rm max}$. At $\lambda\approx1.3$ mm ($230$ GHz) and $B_{\rm max}\sim9,000$ km, the synthesized beam reaches $\theta\sim23~\mu$as, with $0.87$ mm ($345$ GHz) offering up to $15~\mu$as. The $(u,v)$-coverage, a function of Earth rotation, array geometry, and track duration, defines image fidelity and maximal recoverable spatial frequencies (Ricarte et al., 2014).

The EHT nominally observes at two primary bands: $230$ GHz (routine) and $345$ GHz (scattering mitigation). The choice of band affects not only resolution, but also susceptibility to interstellar electron scattering (with $\theta_{\rm scatt} \propto \lambda^2$) and atmospheric calibration requirements.

2. Data Products: Visibilities, Noise, and Calibration

2.1. Complex Visibilities and Closure Quantities

The core EHT data product is the complex visibility $V_{ij}(u,v)$ for each baseline, representing a Fourier component of the sky brightness distribution $I(l,m)$:

$$V_{ij}(u,v) = \iint I(l,m)\,e^{-2\pi i(ul+vm)}\,dl\,dm,$$

where $(u,v)$ are baseline spatial frequencies in units of the observing wavelength $\lambda$.

Amplitude calibration is achieved through a-priori SEFD measurements and system temperature monitoring. Bandpass calibration corrects frequency-dependent complex gains; fringe fitting removes residual station-based delays and phase offsets; gain solution application enforces co-phasing in arrayed dishes (e.g., phased ALMA) (Ricarte et al., 2014).

Closure phases,

$\Phi_{ijk} = \arg\bigl[V_{ij}V_{jk}V_{ki}\bigr],$

are invariant to station-based phase corruption and serve as robust phase observables. Closure amplitudes (quadrangle ratios) remove station-based gain errors.

2.2. Thermal Noise and SNR

The RMS noise on a complex visibility for a given baseline $i$–$j$ is

$$\sigma_{ij} = \frac{\mathrm{SEFD}_i\,\mathrm{SEFD}_j}{2\,(\mathrm{SEFD}_i+\mathrm{SEFD}_j)\sqrt{\Delta\nu\,\tau}},$$

where $\Delta\nu$ is the correlated bandwidth (e.g., $4$ GHz), and $\tau$ the coherence-limited integration time (e.g., $10$ s). Baseline SNR is then $|V_{ij}|/\sigma_{ij}$ (Ricarte et al., 2014).

SEFDs vary widely: at $230$ GHz, representative SEFDs are $\sim110$ Jy (ALMA), $560$ Jy (LMT), $11,900$ Jy (SMTO), $3,500$ Jy (CARMA, phased), and $7,300$ Jy (SPT). At $345$ GHz, SEFDs increase by factors of $2$–$3$ due to higher receiver noise and atmospheric opacity.

3. Precision Techniques: Coherent Averaging, Closure, and Dual-frequency Observations

The precision of EHT measurements is maximized through:

• Coherent averaging: Closure phases can be coherently averaged over timescales $T\gg\tau$, with phase error reduction scaling as $\sim1/\sqrt{T/\tau}$. For $T=10$ min and $\tau=10$ s, this yields a phase uncertainty reduction of $\sqrt{60}\approx7.7$ (Ricarte et al., 2014).
• Robust use of closure quantities: Emphasis on closure phases (and amplitudes) as data products minimizes sensitivity to station-based errors, key for both imaging and model discrimination.
• Dual-frequency and dual-source strategies: $230$ GHz and $345$ GHz observations probe sources at different optical depths and circumvent scattering blur (by a factor of $\sim3$ at $345$ GHz toward Sgr A*). Simultaneous imaging of Sgr A* and M87 helps break geometric and degeneracy constraints common to single-frequency, single-source analyses.

4. Imaging and Model Fitting Methodologies

4.1. Regularized Maximum Likelihood (RML) and Advanced Imaging

Imaging from sparse $(u,v)$ data and rapidly-varying atmospheric phases necessitates regularization. Images $I$ are reconstructed by minimizing

$$\chi^2(I) = \sum_k \frac{|D_k^{\rm (obs)} - D_k^{\rm (mod)}(I)|^2}{\sigma_k^2},$$

where $D_k$ includes visibilities, closure phases, and closure amplitudes. Regularization terms $R(I)$ penalize undesirable image properties:

$S(I) = \chi^2(I) + \lambda R(I).$

RML algorithms support entropy, $\ell_1$ (sparsity), and total variation (TV) regularizers.

Algorithmic advances include:

• Polarimetric MEM: Incorporation of polarimetric entropy, robust to gain systematics.
• Bi-Spectrum Sparse Modeling (BSSpM): Enforces pixel-level sparsity and edge preservation, enabling superresolution at $\sim0.4$ the nominal fringe spacing.
• CHIRP: Data-driven patch-based priors for improved extended-structure recovery and resilience to low-SNR (Fish et al., 2016).

4.2. Model-based Geometric Fitting

Geometric models for the emission structure (e.g., crescents plus photon rings, "m-ring" models, blurred rings) are directly fit to visibilities or closure observables. These enable parameter estimation for diameter, width, brightness asymmetry, and ring thickness, which are then mapped to physical interpretations (e.g., photon ring detection as a Kerr black hole signature) (Ricarte et al., 2014).

5. Science Analysis: Detectability, Variability, and Model Discrimination

5.1. Photon Ring Detection

The EHT is capable of detecting the General Relativity-predicted photon ring, a sharply lensed emission feature with radius $R_{\rm ring} \approx \sqrt{27}\,GM/c^2$ (projected to $\sim27~\mu$as for Sgr A*, $\sim38~\mu$as for M87). GRMHD simulations predict a fractional ring brightness $f_{\rm ring} = F_R/(F_C+F_R)$ ranging from $1\%$ to $10\%$. Achieving a $5\sigma$ ring detection with a $2$ Jy source and $f_{\rm ring}=5\%$ requires an effective baseline sensitivity of $\lesssim20$ mJy, feasible only with phased ALMA and LMT (Ricarte et al., 2014).

5.2. Monitoring Structural Variability

GRMHD and jet models predict event-horizon-scale variability at $5$–$20\%$ amplitude over hour (Sgr A*) to day (M87) timescales. Observationally, multi-epoch closure phase and amplitude tracking are used to discriminate between static and variable models, requiring high-SNR closure phase precision ($<2^\circ$) and dense temporal sampling (Ricarte et al., 2014).

Short baselines (e.g., CARMA–SMTO) constrain overall geometry, while long baselines (e.g. ALMA–SMTO) probe fine structure. Closure phase precisions $\lesssim1^\circ$ on key triangles enable discrimination of $\sim5$–$10\%$ intrinsic structural changes.

6. Practical Recommendations and Future Data Analysis Directions

For robust scientific exploitation, EHT data analysis should prioritize:

• Dual-frequency, dual-source campaigns to break degeneracies in accretion and emission parameters and to mitigate scattering.
• Prioritization of closure phases and coherent integration on short triangles for maximal structural sensitivity.
• Simultaneous, multi-epoch monitoring to distinguish real (astrophysical) from calibration-induced variability and to study dynamical horizon-scale phenomena.
• Advanced regularization and polarimetric imaging: The inclusion of full-Stokes (I, Q, U, V) data, wideband synthesis, and learned or physics-informed priors will enhance recovery of magnetic geometry and plasma physics.

With properly calibrated amplitudes, complex visibilities, and high-precision closure phases, the EHT is configured to detect photon rings at high significance, distinguish among GRMHD flow/jet models, and measure horizon-scale variability—all essential for testing strong-field GR and accretion physics (Ricarte et al., 2014).

---

• (Ricarte et al., 2014) The Event Horizon Telescope: exploring strong gravity and accretion physics
• (Fish et al., 2016) Observing---and Imaging---Active Galactic Nuclei with the Event Horizon Telescope