Next-Generation Event Horizon Telescope

• ngEHT is a next-generation radio interferometer that adds 6–10 high-frequency dishes globally to fill uv-plane gaps and extend baseline coverage.
• It upgrades sensitivity and resolution by quadrupling bandwidth and reducing noise, achieving imaging resolutions as fine as 15 μas and dynamic ranges >500:1.
• ngEHT supports advanced polarimetric and dynamic imaging, enabling robust photon ring detection and detailed studies of SMBH accretion, jets, and strong-field gravity.

The Next-Generation Event Horizon Telescope (ngEHT) is a planned upgrade to the original Event Horizon Telescope (EHT) array, designed to overcome the EHT’s limitations in sensitivity, angular resolution, and dynamic range. The ngEHT will enable horizon-scale imaging and polarimetry of a significantly larger population of supermassive black holes (SMBHs), systematically test strong-field general relativity, and uniquely address the structure and dynamics of SMBH accretion, jet launching, and multi-messenger astrophysics.

1. Array Architecture and Global Coverage

The ngEHT will extend the existing EHT by adding 6–10 new high-frequency dishes at strategically selected geographical locations, including NOEMA (Europe), the Greenland Telescope, a site in southern Africa, southern hemisphere (Chile/Argentina), the US Southwest, and potentially Australia or Antarctica (Zhang et al., 25 Jun 2024). These additional stations will:

• Fill critical north–south and east–west gaps in the $uv$-plane.
• Increase the maximum terrestrial baseline from $\sim10{,}000$ km (EHT) to $\sim12{,}000$ km, corresponding to a maximum spatial frequency of $\sim12$ G$\lambda$ at 345 GHz.
• Densify intermediate $uv$-coverages (2–6 G$\lambda$ at 230 GHz) by $\approx30$–$50\%$, reducing imaging artifacts and boosting dynamic range.

Three-frequency operation is central, with simultaneous or dual observations at 230 GHz ($\lambda\approx1.3$ mm) and 345 GHz ($\lambda\approx0.87$ mm), and in some designs incorporating an 86 GHz/85 GHz band for robust frequency phase transfer (FPT) and source-frequency phase-referencing (SFPR) (Jiang et al., 2022, Rioja et al., 2023).

2. Sensitivity, Resolution, and Bandwidth Upgrades

ngEHT enhancements target the physical limitations of the original EHT in sensitivity and angular resolution by:

• Quadrupling the aggregate observing bandwidth from 4 GHz/pol (EHT) to up to 16 GHz/pol (ngEHT), with multi-band total instantaneous bandwidth often quoted as 16 GHz (Zhang et al., 25 Jun 2024, Johnson et al., 2023).
• Upgrading receivers for lower system noise temperatures, leading to station SEFDs reduced by a factor of $\sim2$ (e.g., SEFD $\sim$1500 Jy at 230 GHz) (Zhang et al., 25 Jun 2024).
• Achieving theoretical imaging resolutions of
  • $$\theta_{230} \simeq \frac{1.3\,\mathrm{mm}}{1.2\times10^7\,\mathrm{m}} \simeq 20\,\mu\mathrm{as}$$ (ngEHT)
  • $$\theta_{345} \simeq \frac{0.87\,\mathrm{mm}}{1.2\times10^7\,\mathrm{m}} \simeq 15\,\mu\mathrm{as}$$
• Reaching baseline thermal noise (per 60 s) of $\sigma_{230} \sim 8$ mJy and $\sigma_{345} \sim 16$ mJy with the nominal SEFDs and $\Delta\nu=16$ GHz (Zhang et al., 25 Jun 2024).
• Imaging dynamic range improvements from EHT’s $\sim100\!:\!1$ to $\gtrsim500\!:\!1$ (static imaging) and $>1000\!:\!1$ in continuous/dynamical campaigns (Zhang et al., 25 Jun 2024, Roelofs et al., 2022).
• Table: Comparative ngEHT Capabilities

Capability	Current EHT (230 GHz)	ngEHT (230 GHz)	ngEHT (345 GHz)	Ground+Space (BHEX)
Max Baseline (km)	$\sim10{,}000$	$\sim12{,}000$	$\sim12{,}000$	$\sim50{,}000$
Resolution $\theta$ ($\mu$as)	20	18	13	5
Bandwidth $\Delta\nu$ (GHz)	4	16	16	16
SEFD (Jy)	3000	1500	3000	1500
Baseline σ (mJy, 60 s)	$\sim40$	$\sim8$	$\sim16$	$\lesssim8$
Dynamic Range	$\sim100\!:\!1$	$\gtrsim500\!:\!1$	$\gtrsim500\!:\!1$	$\gtrsim1000\!:\!1$
Accessible sources	2 (M87*, Sgr A*)	$\sim12$	$\sim12$	$\gtrsim50$

Observations at 345 GHz are crucial for lengthening baselines, surpassing scintillation limits (especially for Sgr A*), and achieving beam sizes necessary for super-resolution and photon-ring analyses (Shavelle et al., 13 Jul 2024, Tiede et al., 2022).

3. Calibration, Phase Transfer, and Astrometry

The ngEHT incorporates advanced calibration methods leveraging simultaneous multi-band observations:

• Frequency Phase Transfer (FPT) and Source-Frequency Phase Referencing (SFPR) exploit the linearity of tropospheric phase fluctuations to transfer solutions from lower (e.g., 85/110 GHz) to higher (230–345 GHz) bands, extending achievable coherence times by $10^2$–$10^3\times$ (from $\sim10$ s at 230 GHz to $10^3$–$10^4$ s) (Jiang et al., 2022, Rioja et al., 2023).
• The detection threshold at high frequencies is set by SNR at the reference (lower) band, allowing routine detection of sources down to $\sim$10 mJy at 255–345 GHz.
• SFPR permits direct measurement of frequency-dependent core shifts with astrometric precision $\sigma_\theta \sim 3\,\mu\mathrm{as}$, enabling registration of structures across frequency bands and micro-arcsecond-level astrometry for Sgr A*, M87*, SMBH binaries, and cluster galaxies.
• Receiver systems must enable simultaneous or near-simultaneous operation at two or more frequencies; tri-band receivers (e.g., 85/230/345 GHz) with integer frequency ratios are strongly preferred to avoid phase ambiguity (Rioja et al., 2023).

4. Science Case Expansion: Population Studies and New Targets

The ngEHT will transition from imaging only Sgr A* and M87* to direct horizon-scale studies of $\mathcal{O}(10^1$–$10^2)$ SMBHs. Quantitative projections (Pesce et al., 2022, Zhang et al., 25 Jun 2024):

• Under $\nu=230$ GHz and April conditions, a Phase 1 ngEHT can conduct:
  • Mass (ring-diameter) measurements for $\sim$50 SMBHs
  • Spin constraints for $\sim$30 SMBHs (polarization proxies)
  • Shadow detections (clear ring interior) for $\sim$7 SMBHs
• For each target, the ability to recover the photon-ring radius and mass is governed by the fundamental formula:

$$\theta_\mathrm{ring} = 2\sqrt{27}\,\frac{G M}{c^2 D} \approx 9\,\mu\mathrm{as}\ \Bigl(\frac{M}{10^9\,M_\odot}\Bigr)\ \Bigl(\frac{D}{10\,\mathrm{Mpc}}\Bigr)^{-1}$$

• Many new targets have $F_{230} \sim 0.03$–$0.2$ Jy and angular diameters $\gtrsim10\,\mu$as, within ngEHT’s sensitivity and resolution.
• Higher Eddington ratios for these new sources may yield larger optical and Faraday depths (complicating polarimetric spin measurements) (Zhang et al., 25 Jun 2024).
• Accessing masses and spins with high accuracy depends on independent inclination constraints and robust modeling that includes possible jet contributions.

5. Polarimetry, Photon Ring Detection, and Dynamic Imaging

ngEHT will enable a new generation of horizon-scale polarimetric science and time-domain imaging:

• Polarimetric detection of the $n=1$ photon ring in Sgr A* at 345 GHz is feasible using a robust circular-polarization ratio observable,

$R(u,v) \equiv \frac{V_R(u,v)}{V_L(u,v)}$

where $V_R$, $V_L$ are the right- and left-circular polarized visibilities. A phase reversal in $R$ at baselines $\gtrsim4$ G$\lambda$ signals the photon ring; detection significance is quantified by the metric

$$\mathrm{SNR}_\mathrm{PR} = \frac{\operatorname{Re}\bigl[ \langle R\rangle_\mathrm{long} - \langle R\rangle_\mathrm{short} \bigr]} {\sigma_\mathrm{noise}}$$

Simulations predict $\mathrm{SNR}_\mathrm{PR} \sim 2$–3 for a week-long campaign at nominal Phase 1 sensitivity (Shavelle et al., 13 Jul 2024).

• For static and dynamic imaging, the quadrupled instantaneous bandwidth and increased station count deliver dynamic range (DR) enhancements from $\sim100$ to $\gtrsim1000$, critical to optically resolving central brightness depressions (shadows) and tracing transient/faint emission (e.g., hot spots, jet knots).
• The ngEHT Analysis Challenges demonstrate that with $\sim$10 new antennas and 4$\times$ bandwidth, event-horizon-scale “movies” of M87* jet dynamics and Sgr A* variability can be robustly reconstructed (Roelofs et al., 2022).

6. Imaging Algorithms, Spectral Reconstruction, and Parameter Estimation

ngEHT advances in both data acquisition and analysis methodology:

• Regularized Maximum Likelihood (RML) and Bayesian imaging pipelines using frequency-joint and polarization-joint constraints outperform independent single-frequency reconstructions. Simultaneous RML multi-frequency imaging naturally recovers both intensity and spectral index ($\alpha$) maps:

$$\ln I_i(x,y) = \ln I_0(x,y) + \alpha(x,y)\,\ln\frac{\nu_i}{\nu_0} + \beta(x,y) \Bigl(\ln\frac{\nu_i}{\nu_0}\Bigr)^2$$

(typically, just the linear term, $\alpha$, is sufficient in ngEHT bands) (Chael et al., 2022).

• These methods improve superresolution, spectral mapping, and propagation of $uv$ coverage between frequencies, yielding NRMSE and cross-correlation metrics superior to band-independent methods.
• For parameter estimation, the Fisher-matrix approach yields error estimates on ring diameter, polarization modes, and shadow width, with statistical confidence thresholds of $<20\%$ fractional error for spin proxies.
• Imaging capabilities are extended to fainter, more extended structures (e.g., detecting jet sheath, recollimation nodes, or outflows) by leveraging high dynamic range and expanded $uv$-coverage.

7. Instrumental Challenges, Limitations, and Future Outlook

Despite significant capability gains, key limitations and challenges persist:

• High-frequency operations (especially 345 GHz) are weather-limited, requiring excellent atmospheric conditions (low PWV), and the small dish apertures at new sites necessitate wide bandwidth and long integration for deep sensitivity (Johnson et al., 2023).
• Robust photon-ring detection in realistic data is complicated by blending with direct image features and possible model degeneracies; hybrid imaging methods can yield false positives if not coupled rigorously to physical models (Tiede et al., 2022).
• Faraday rotation and optical depth effects may compromise polarimetric mass/spin measurements for higher-Eddington sources.
• Full exploitation of dynamic imaging and time-domain science demands highly stable calibration, optimized scan strategies, and improved pipeline automation.
• The envisioned extension to space VLBI (Black Hole Explorer, BHEX) will further increase accessible angular resolution ($\sim 5\,\mu\mathrm{as}$) and sample size (tens more SMBHs).

A plausible implication is that, with sustained hardware, calibration, and algorithmic development, the ngEHT will shift SMBH astrophysics from single-object phenomenology to population-level tests of accretion physics and strong gravity, eventually encompassing both stellar-mass and supermassive regimes over a wide range of cosmic history.