Studying Black Holes on Horizon Scales with VLBI Ground Arrays
Lindy Blackburn, Sheperd Doeleman, Jason Dexter, José L. Gómez, Michael D. Johnson, Daniel C. Palumbo, Jonathan Weintroub, Katherine L. Bouman, Andrew A. Chael, Joseph R. Farah, Vincent Fish, Laurent Loinard, Colin Lonsdale, Gopal Narayanan, Nimesh A. Patel, Dominic W. Pesce, Alexander Raymond, Remo Tilanus, Maciek Wielgus, Kazunori Akiyama, Geoffrey Bower, Avery Broderick, Roger Deane, Christian M. Fromm, Charles Gammie, Roman Gold, Michael Janssen, Tomohisa Kawashima, Thomas Krichbaum, Daniel P. Marrone, Lynn D. Matthews, Yosuke Mizuno, Luciano Rezzolla, Freek Roelofs, Eduardo Ros, Tuomas K. Savolainen, Feng Yuan, Guangyao Zhao
AAstro2020 APC White Paper
Studying Black Holes on Horizon Scales with VLBI Ground Arrays
Lindy Blackburn , , ∗ Sheperd Doeleman , , † , Jason Dexter , José L. Gómez , Michael D. Johnson , ,Daniel C. Palumbo , , Jonathan Weintroub , , Katherine L. Bouman , , , Andrew A. Chael , , , ,Joseph R. Farah , , , Vincent Fish , Laurent Loinard , , Colin Lonsdale , Gopal Narayanan ,Nimesh A. Patel , Dominic W. Pesce , , Alexander Raymond , , Remo Tilanus , , , Maciek Wielgus , ,Kazunori Akiyama , , , , Geoffrey Bower , Avery Broderick , , , Roger Deane , , Christian M. Fromm ,Charles Gammie , , Roman Gold , Michael Janssen , Tomohisa Kawashima , Thomas Krichbaum ,Daniel P. Marrone , Lynn D. Matthews , Yosuke Mizuno , Luciano Rezzolla , Freek Roelofs ,Eduardo Ros , Tuomas K. Savolainen , , , Feng Yuan , , , Guangyao Zhao Black Hole Initiative at Harvard University, 20 Garden Street,Cambridge, MA 02138, USA Center for Astrophysics | Harvard & Smithsonian, 60 GardenStreet, Cambridge, MA 02138, USA National Radio Astronomy Observatory, 520 Edgemont Road,Charlottesville, VA 22903, USA Massachusetts Institute of Technology, Haystack Observatory,99 Millstone Road, Westford, MA 01886, USA National Astronomical Observatory of Japan, 2-21-1 Osawa,Mitaka, Tokyo 181-8588, Japan Institute of Astronomy and Astrophysics, Academia Sinica,645 N. A’ohoku Place, Hilo, HI 96720, USA Perimeter Institute for Theoretical Physics, 31 Caroline StreetNorth, Waterloo, ON, N2L 2Y5, Canada Department of Physics and Astronomy, Univ. of Waterloo,200 University Avenue West, Waterloo, ON, N2L 3G1, Canada Waterloo Centre for Astrophysics, University of Waterloo,Waterloo, ON N2L 3G1 Canada Department of Physics, University of Pretoria, LynnwoodRoad, Hatfield, Pretoria 0083, South Africa Centre for Radio Astronomy Techniques and Technologies,Department of Physics and Electronics, Rhodes University, Gra-hamstown 6140, South Africa Max-Planck-Institut für Extraterrestrische Physik, Giessen-bachstr. 1, D-85748 Garching, Germany Inst. für Theoretische Physik, Goethe-Universität Frankfurt,Max-von-Laue-Straße 1, D-60438 Frankfurt am Main, Germany Department of Physics, University of Illinois, 1110 WestGreen St, Urbana, IL 61801, USA Department of Astronomy, University of Illinois at Urbana-Champaign, 1002 West Green Street, Urbana, IL 61801, USA Instituto de Astrofísica de Andalucía-CSIC, Glorieta de laAstronomía s/n, E-18008 Granada, Spain Department of Astrophysics, Institute for Mathematics, As-trophysics and Particle Physics (IMAPP), Radboud University,P.O. Box 9010, 6500 GL Nijmegen, The Netherlands Instituto de Radioastronomía y Astrofísica, Universidad Na-cional Autónoma de México, Morelia 58089, México Instituto de Astronomía, Universidad Nacional Autónoma deMéxico, CdMx 04510, México Steward Observatory and Department of Astronomy, Univer-sity of Arizona, 933 N. Cherry Ave., Tucson, AZ 85721, USA University of Massachusetts Boston, 100 William T, Morris-sey Blvd, Boston, MA 02125, USA Leiden Observatory—Allegro, Leiden University, P.O. Box9513, 2300 RA Leiden, The Netherlands Netherlands Organisation for Scientific Research (NWO),Postbus 93138, 2509 AC Den Haag , The Netherlands Shanghai Astronomical Observatory, Chinese Academy ofSciences, 80 Nandan Road, Shanghai 200030, PRC Key Laboratory for Research in Galaxies and Cosmology,Chinese Academy of Sciences, Shanghai 200030, PRC School of Astronomy and Space Sciences, Univ. of ChineseAcademy of Sci., No. 19A Yuquan Road, Beijing 100049, PRC Korea Astronomy and Space Science Institute, Daedeok-daero776, Yuseong-gu, Daejeon 34055, Republic of Korea Department of Astronomy, University of Massachusetts,01003, Amherst, MA, USA Max-Planck-Institut für Radioastronomie, Auf dem Hügel 69,D-53121 Bonn, Germany Aalto University Department of Electronics and Nanoengi-neering, PL 15500, FI-00076 Aalto, Finland Aalto University Metsähovi Radio Observatory, Metsähovin-tie 114, FI-02540 Kylmälä, Finland California Institute of Technology, 1200 East CaliforniaBoulevard, Pasadena, CA 91125, USA Princeton Center for Theoretical Science, Jadwin Hall,Princeton University, Princeton, NJ 08544, USA NASA Hubble Fellowship Program, Einstein Fellow ∗ [email protected], † [email protected] a r X i v : . [ a s t r o - ph . I M ] N ov Introduction
This white paper outlines a process to design, architect, and implement a global array ofradio dishes that will comprise a virtual Earth-sized telescope capable of making the firstreal-time movies of supermassive black holes (SMBH) and their emanating jets. Thesemovies will resolve the complex structure and dynamics at the event horizon, bringing intofocus not just the persistent strong-field gravity features predicted by general relativity,but also details of active accretion and relativistic jet launching that drive galaxy evolutionand may even affect large scale structures in the Universe. SMBHs are the most massiveand most compact objects predicted by Einstein’s theory of gravity. They are believedto energize the luminous centers of active galaxies, where they convert the gravitationalpotential energy of infalling matter to radiant power and jetted outflows of charged particlesthat can stretch to hundreds of thousands or even millions of light years. We propose toturn the extreme environment of their event horizons into laboratories where astronomers,physicists, and mathematicians can actively study the black hole boundary in real-time, andwith a sensitivity and angular resolution that allow them to attack long-standing fundamentalquestions from new directions.
Figure 15 as a conservative representation of our fi nal M87imaging results.The fi ducial images from each pipeline ( Figure 11 ) , as wellas the conservative, blurred, pipeline-averaged images ( Figure 15 ) provide some evidence for evolution in the ringstructure between April 5, 6 and 10, 11. We discuss thisevolution in more detail in Appendix E ( Figure 33 ) . Somechange in the image structure between April 5, 6 and 10, 11 isnecessitated by the variations seen in the underlying closurephases ( Paper III ) . We fi nd more variation in the image pairsthat are separated by larger intervals, suggesting that thesevariations are intrinsic. However, we cannot unambiguouslyassociate the observed variability with coherent evolution of aspeci fi c image feature.Figure 16 shows the visibility amplitude and phase for eachof the three April fi ducial images as a function of vectorbaseline. Note that no restoring beam is required for CLEAN inthis visibility-domain analysis. Each image produces nulls inthe visibility amplitude near the SMA – SMT baseline, con-sistent with the observed amplitudes ( see Figure 2 ) . Thevisibility phase shows rapid swings at these nulls. Thevisibilities of the images from the different pipelines are mostsimilar near the EHT measurements, as expected. On longerbaselines than those sampled by the EHT, the DIFMAP imageproduces notably higher visibility amplitudes than those of the eht-imaging and
SMILI images, as expected from the factthat the
DIFMAP image is fundamentally a collection of pointsources.
Measuring the variation in images produced in a parametersurvey Top Set allows us to evaluate image uncertainties due tothe explored imaging choices. Figure 17 shows uncertaintiesrelated to imaging assumptions from the largest Top Set ( that ofthe eht-imaging parameter survey ) on April
11 data.Reconstructed image uncertainties are concentrated in theregions with enhanced brightness, notably in the three “ knots ” inthe lower half of the ring ( Figure 17; top panel ) . These are alsothe regions that show the most variation among different imagingmethods ( Appendix I compares their azimuthal pro fi les ) .Visibility-domain modeling provides another method to assessimage structure. In Paper VI, we explore fi tting simple crescentmodels to the data. For instance, a crescent with a brightnessgradient and blurring reproduces the north – south asymmetry inimages without additional azimuthal structure ( the “ blurred andslashed with LSG ” crescent of Paper VI ) . However, this modelgives CP2 c between 3.2 and 11.5 and log CA2 c between 2.2 and 6.6for different days and bands when assuming 0% systematic error ( compare with Table 5 ) . Adding additional degrees of freedom inthe form of three elliptical Gaussian components to the crescent Figure 14.
Fiducial images of M87 on April
11 from our three separate imaging pipelines after restoring each to an equivalent resolution. The eht-imaging and
SMILI images have been restored with 17.1 and 18.6 μ as FWHM Gaussian beams, respectively, to match the resolution of the DIFMAP reconstruction restored with a20 μ as beam. Figure 15.
Averages of the three fi ducial images of M87 for each of the four observed days after restoring each to an equivalent resolution, as in Figure 14. Theindicated beam is 20 μ as ( i.e., that of DIFMAP , which is always the largest of the three individual beams ) . The Astrophysical Journal Letters, ( ) , 2019 April 10 The EHT Collaboration et al. Figure 1: µ as. These represent the highest angular resolution images ever made from the surface of theEarth, and show clearly the predicted photon orbit caused by extreme light bending in the presence of a 6.5billion solar mass black hole. The central dark region occurs because light rays interior to the photon ringspiral into the event horizon. There is clear variation in the structure over the span of 5 days. While ambitious, this vision is grounded in recent remarkable results: on April 10th, 2019,after a decade of developmental efforts from a global collaboration of scientists [18], the EventHorizon Telescope (EHT) announced the first successful imaging of a black hole [17], Figure 1.To accomplish this, the EHT used the technique of very long baseline interferometry (VLBI),in which radio dishes across the globe are synchronized by GPS timing and referenced toatomic clocks for stability, thereby synthesizing a single Earth-sized telescope. Throughdevelopment of cutting edge instrumentation, the EHT extended the VLBI technique to the1.3 mm observing band, and by deploying these systems, created a virtual telescope with thehighest angular resolution currently possible from the surface of the Earth. First experiments[15, 16] confirmed event horizon-scale structures in both Sgr A ∗ and M87. Build-out of thefull EHT enabled detailed imaging of the black hole “shadow” of M87, which is formed bythe lensed photon orbit of the 6.5 billion solar mass black hole at the galaxy’s core. CurrentEHT imaging capability is limited by the sparsity of VLBI baseline coverage, and targeted1 H T µ as0:00 0:05 0:10 0:15 0:20 E H T - II T r u t h K) T r u t h ng E H T E H T Figure 2:
Reconstructing moviesof flares from Sgr A ∗ with the EHT.Bottom row: Simulated images ofa “hot spot” orbiting Sgr A ∗ witha period of ∼ minutes (ModelB from [9, 14]). Upper rows:Corresponding reconstructions ofthe model with the EHT2017 andngEHT arrays merging 230 and345 GHz, demonstrating the po-tential to study the evolution offlares in Sgr A ∗ on timescales ofminutes [26, 8]. Reconstructionsare performed with visibility am-plitudes and closure phases, reflect-ing calibration similar to that of theEHT2017 data. expansion of the EHT array by augmenting existing stations, as well as developing new sites,can greatly increase the scope of EHT core science over the next decade. We refer to theexpanded array as the next-generation EHT, or ngEHT. A separate white paper is dedicatedto a complementary expansion of the EHT array by deployment of a radio telescope in orbitaround the Earth. The detection of the black hole shadow in M87 [17] has opened up the opportunity forrepeated experimental studies of strong gravity and horizon scale accretion and jet launchingwith ngEHT. Future observations will measure the detailed shape and size of the black holeshadow and surrounding photon ring, allowing direct tests of the Kerr metric describingblack holes in general relativity. An ngEHT will also address fundamental questions aboutthe role of magnetic fields in the accretion and jet launching process as traced by the observedtime-variable, polarized synchrotron radiation.
Measuring the shape and size of the shadow and surrounding lensed photon ring provides anull hypothesis test of General Relativity [24]. The mass and distance of Sgr A ∗ are knownto ∼
1% [21], so the precision of GR tests for this source will be limited primarily by thefidelity of EHT data and the ability to extract the emission corresponding to the black holecircular photon orbit and interior shadow. Positional measurements of luminous matter (“hotspots”) orbiting near the event horizon, as shown in Fig. 2, can be used to map the spacetimemetric near the black hole and constrain the black hole spin. Combining one or more suchEHT measurement of Sgr A ∗ with other observations [e.g., 20] allows a test of the “no hair”theorem [22, 10, 31]. 2igher angular resolution allows more precise measurements of the shadow size and shape,while increased dynamic range improves image fidelity and allows us to extract the thin,bright photon ring feature from the more diffuse surrounding emission. Snapshot imagingof Sgr A ∗ on timescales of minutes is required to track relativistic motions around the blackhole. Magnetic fields play an outsized role in accretion and jet formation. The magnetorotationalinstability [MRI, 2] is thought to transport angular momentum and drive accretion ontothe central black hole. Dynamically important magnetic fields can cause instabilities andflaring on horizon scales [34]. The polarized synchrotron radiation observed by the ngEHTtraces magnetic field geometry (Fig. 3), while its time variability encodes the dynamics ofspiral waves driven by the MRI and magnetic flux eruption events associated with strongmagnetic fields. Triggered multi-wavelength campaigns are needed to fully take advantage ofngEHT’s capability to spatially resolve structures associated with the energetic, high energyflares from Sgr A ∗ . The X-ray radiation in Sgr A ∗ flares suggests that particles can beaccelerated to high energy even around a quiescent black hole [13]. Spatially resolving theirradio counterparts will provide new constraints on the acceleration mechanism. Blurred Sim ngEHT Image
Figure 3:
Comparison of po-larization map of a simulationof M87 [11] blurred to half thenominal resolution of ngEHT(left), and a polarimetric re-construction of synthetic ngEHTdata generated by the simula-tion (right). ngEHT enableshigh fidelity polarimetric re-constructions, revealing the or-dered, horizon-scale fields in thissimulation of a “magnetically-arrested” disk.
Snapshot polarimetric imaging with the future EHT can reveal the structure and dynam-ics of magnetic fields. Simultaneous polarimetric observations at 230 and 345 GHz will allowprobing the magnetic field degree of ordering, orientation, and strength through Faradayrotation studies. Spectral index analyses will probe other plasma properties, such as theelectron density and temperature.
The processes that govern the formation, acceleration and collimation of powerful relativisticjets in active galactic nuclei (AGN) and X-ray binaries are a half-century-long mystery inblack hole physics. The leading scenarios rely on magnetic fields to extract rotational energy,either from orbiting material [4] or from the black hole itself [5]. Magnetic fields downstreamfurther collimate and accelerate the jet to relativistic speeds. EHT observations of M873rovide a unique opportunity to study jet launching, collimation, and acceleration at thebase in the immediate vicinity of the black hole.Figure 4 shows reconstructed 3D GRMHD simulations of the jet launching region in M87with current and sample ngEHT arrays. The addition of short baselines anchored to existinglarge apertures, combined with observations at progressively higher frequencies will improveboth the imaging dynamic range and angular resolution to study formation, collimation andacceleration of relativistic jets, not only in M87 but also in other nearby AGN [7]. Thisopens new possibilities for linking jet power to black hole spin, accretion rate, and diskmagnetization through direct comparison of observation and simulations on scales down tothe event horizon. Triggered observations and centralized data processing will increase thecadence of the observations, allowing the study of time variable jet ejections first near theblack hole, and subsequently as they emerge from the AGN core. the EHT-II reconstructions, which include empirically verified error budgets and estimated performance of future sites, demonstrates that connecting the horizon-scale structure and dynamics near the black hole to the emergence and launch of the M87 jet is achievable. Figure 3: EHT baseline coverage for M87 (left) and SgrA* (right). Each point shows the April 2017 1.3mm EHT coverage (black), the 1.3mm coverage anticipated for EHT-II (black and red), and the added coverage at 0.87mm (blue).
EHT-II images of M87 will open new avenues to understand how black holes launch and power relativistic jets. For instance, if the jet is powered by the spinning black hole, then magnetic fields threading the horizon are predicted to rotate at approximately half the angular frequency of the black hole (e.g., Blandford & Znajek 1977; Macdonald & Thorne 1982). With EHT-II observations of M87 over several weeks, this effect will be directly observable via polarimetric movie reconstructions. In addition, measurements of the magnetic field strength via Faraday rotation from joint 230+345 GHz observations will test reveal whether the jet power corresponds to predictions from the Blandford-Znajek mechanism.
Figure 4:
Left:
GRMHD snapshot from a simulation of M87 (Chael et al. 2019). Main panel is log scale; inset is linear scale.
Middle:
Reconstruction using EHT2017, revealing the circular ~ 40 µas ring surrounding the black hole shadow but no jet.
Right:
Reconstruction using EHT-II, including both 230 and 345 GHz, revealing both the black hole and its jet. ngEHT
Figure 4: GRMHD snapshot from a simulation of M87 (Chael et al. 2019). Main panel islog scale; inset is linear scale. Middle: Reconstruction using EHT2017, revealing the circular ∼ µ as ring surrounding the black hole shadow but no jet. Right: Reconstruction usingngEHT, including both 230 and 345 GHz, revealing both the black hole and its jet. The quality of ngEHT images/movies and their corresponding traction on key science ques-tions depend on the baseline coverage of the array as well as overall sensitivity, observingfrequency, bandwidth, and observing/scheduling constraints. Additional improvements inimaging and analysis algorithms will further drive design requirements and trade-offs indefining the ngEHT instrument systems and array architecture. We advocate a formal sys-tem engineering approach, in which key science questions are used to define and select acrosstechnical elements for the array. In Table 1 we outline an abbreviated science traceabilitymatrix for the ngEHT. The full array design must be explored using system engineeringdriven simulation and science optimization process resulting in an expanded STM to definea phase of ngEHT design, followed by a phase of implementation, both timed to deliver afunctional ngEHT array by the end of the coming decade.4
TM Shortform for Ground White PaperRev 9 July 2019Science Goals Measurement Requirements Developments: Array, Instrument, and Algorithms
Are SMBHs described by the Kerr metric? 1. Angular resolution of about 10 μas2. Accelerated baseline sampling to enable static imaging of SgrA* 1. Enable 345 GHz observations2. Identify & characterize candidate sites3. Optimize baseline coverage for Sgr A*4. Sufficient sensitivity for a fully-connected array5. Methods to study intra-day variation of Sgr A*What drives accretion onto a SMBH and triggers flaring events? 1. Movies of the Sgr A* accretion flow on sub-ISCO timescales 2. Polarimetric movies of Sgr A* and M87 to study magnetic field turbulence & multi-wavelength flares3. Coordinated multi-wavelength & triggered observations 1. Optimize array for rapid baseline sampling 2. Simultaneous 230 & 345 GHz with dual-polarization3. Methods for polarimetric movie reconstructions 4. Triggered turn-key VLBI scheduling5. Strategic array redundancy to reduce sensitity to weather and site loss What is the role of the SMBH in forming, collimating & powering a relativistic jet? 1. Horizon-scale polarimetric imaging to measure magnetic field structure2. Faraday rotation measurements to measure magnetic field strength3. Increased image dynamic range from ~10 to ~100 to connect the black hole, jet & counter-jet4. Movies of the M87 jet-launching region over multi-month timescales 1. Co-temporal 230 & 345 GHz with dual-polarization2. Increased sensitivity through wider bandwidths 3. Optimize baseline coverage for M87 horizon scale and jet launching region4. Improved calibration & algorithms for multi-scale and high dynamic range imaging5. Enhance array operations to optimize duration, cadence & quality of observations
Table 1: A short form science traceability matrix (STM). The STM links the key sciencequestions in the first column with the top level requirements for astronomical measurementsin the second, and these drive the specifications for detailed design, of the array configura-tion, instrument developments, and software post processing algorithms in the third column.System engineering will expand this STM in the early phases of an upgrade.
Current EHT images are already exceptionally rich scientifically. Following system engi-neering practices, and referencing the STM in figure 1, we propose to extend the scientificpotential of ground-based mm VLBI observations by quadrupling the current recorded in-stantaneous bandwidth of the EHT, adding a 345 GHz capability, and incorporating newsites to the existing array. This last possibility stems from the important realization thatsingle large apertures in the array (phased ALMA in Chile, the Large Millimeter Telescopein Mexico, and future phased NOEMA in France) provide such high sensitivity that addingsmall-diameter dishes in ideal geographic locations can dramatically improve imaging fidelity– even for sites where the atmospheric conditions are more variable than is typical for currentmm/submm facilities (Figure 7). By roughly doubling the number of antennas in the EHTthrough addition of several new small diameter dishes as well as new stations ngEHT canreconstruct not just images of extraordinary detail, but movies of the dynamics near theblack hole event horizon.
At the bandwidth projected for the ngEHT, antenna diameters between 6 and 12 m will besuitably sensitive for new nodes in the array. The Greenland Telescope is an example of asuccessful relocation of a 12 m ALMA dish, and a similar dish is being relocated to Kitt Peakin Arizona. In ngEHT Phase I, designs for new dishes will be explored using approaches that5igure 5: Distribution of stations around the globe. Stations that participated in theEHT2017 observing campaign are labeled in yellow, while the additional stations that willbe present in the EHT2020 array are labeled in orange. Several possible new site locationsfor the ngEHT are labeled in cyan. Current EHT2017 baselines are shown in magenta.have been successfully used for sub-mm class antennas for the SMA and ALMA [29, 28].Candidate locations for newly designed dishes will be selected based on weather for sub-mm observing, existing infrastructure, and improvement to the spatial frequency coverageof the array [30, 27]. Small dishes are particularly effective close to major ngEHT anchorsites. An example of an expanded ngEHT array, feasible by 2027, is shown in Figure 5.Corresponding improvements in the ( u, v ) -coverage for the EHT science targets are shownin Figure 6. The EHT presently samples 4 GHz bandwidth in dual polarization and two sidebands for atotal of 16 GHz. This corresponds to 64 Gbps for two-bit recording and Nyquist sampling.This matches ALMA’s current bandwidth, though efforts are underway to double the ALMAbandwidth in each sideband over the next decade. The majority of the other EHT sitesalready employ receivers with 8 GHz sidebands, and those that do not are typically in theprocess of upgrading. A doubling of bandwidth per sideband for the EHT would result in arecord rate of 128 Gbps.The ability to simultaneously observe the 1.3 mm (230 GHz) and 0.87 mm (345 GHz)EHT observing bands dramatically improves imaging and movie rendering capability ofthe EHT (Figures 2, 4) as well as polarization observations of, for example, Faraday rota-tion. The ngEHT with 8 GHz per sideband, dual polarization, and simultaneous dual band230/345 GHz capability requires a recording rate of 256 Gbps. A dual-band EHT receiver6 − − λ ) − − − v ( G λ ) µ a s Sgr A ∗ EHT 2017 (230 GHz)ngEHT(230 GHz) ngEHT(230+345 GHz) − − − λ ) − − − v ( G λ ) µ a s M87
EHT 2017 (230 GHz)ngEHT(230 GHz) ngEHT(230+345 GHz)
Figure 6: Fourier space coverage of the EHT primary science sources for the EHT2017 array[18] and for the proposed expanded ngEHT array as shown in Figure 5. OVRO, HAY, KPand GAM sites are excluded from 345 GHz operations. S e n s i t i v i t y o f b a s e li n e ( r m s / s ) Baseline sensitivity to small dish ( = 0.6)
ALMA 2017 actualLMT 2017 actualALMA-II projectedLMT-II projectedNOEMA projectedtarget performance
Figure 7:
Key anchor stations in the ngEHT withsufficient sensitivity to connect small dishes to theentire array on ∼ few second atmospheric timescales.A star marks a low level of correlated flux expectedover long ngEHT baselines. Performance for 2017is taken over 2 GHz of bandwidth and the observedmedian sensitivity of ALMA and LMT during EHTApril 2017 observations, and this is extended to thefull projected bandwidth at 230 GHz for ALMA-IIand LMT-II. NOEMA is calculated for a 12-elementarray under nominal weather conditions, and thesmall ngEHT remote site is evaluated for moder-ately poor line-of-sight opacity of 0.6. that will serve as the prototype for other suitable telescopes in the EHT array will be firstinstalled and commissioned on the LMT, to immediately enhances the high resolution ca-pability of the EHT. 1.3 mm and 0.87 mm wavelength sideband-separating dual-polarizationmixers have been built for facilities such as ALMA and the Institut de RadioastronomieMillimetrique (IRAM). A dual-frequency receiver will be deployed to additional telescopesin the array.New back end development takes advantage of advances in FPGA and ADC capabilities.We are developing a back end capable of processing four 8-GHz bands using 16 Gsps samplers.This is a total of 128 Gbps at 2-bit quantization, available in a compact rack mount box.The European DBBC3 is another platform capable of supporting 128 Gbps using 8 Gspssamplers. A key development is to match the back end rate in data capture and transport.The current Mark 6 VLBI recorder [35] has in the lab operated at close to twice its design7arget of 16 Gbps, due largely to the steady increase in hard disk density and throughput.One path is to further develop the Mark 6 for reliable operation at higher speeds in the field.COTS recording solutions that use Field Programmable Gate Arrays (FPGA) for high-speedparallel data flow are emerging and will also be explored. The current EHT and VLBI practice of recording data and physically transporting them toa central location for correlation has severe disadvantages. This includes inability to verifyin real-time that an experiment has been setup correctly and is working. Data storage disksused by the EHT are large, expensive and take a long time to back up, which means there is avery real risk of data loss in shipment. Observations from the South Pole Station, absolutelykey for Sgr A ∗ , are hobbled because shipment of disks recorded in April takes effectively sixmonths to return to the correlation center. This causes time delays in the data analysis andrequires that the data from all stations be saved for cross-correlation at one time.With the potential eight-fold increase in collected data volume over the current systems,the EHT will develop an innovative way to store and consolidate data for processing. Freespace laser communication presents an alternative. Data rates of many Terabits/sec arepossible. It is an extremely attractive technology to consider for an entirely new paradigm ofEHT operations, supporting real time data transport and correlation. The TBIRD system,developed at MIT Lincoln Laboratory [33], is one example that will be investigated andevaluated. Development of new analysis methods, specially designed to address challenges in EHT data,were critical for producing the first images of a black hole. Due to the extreme sparsity andcalibration uncertainties for high frequency VLBI, recovering an image of M87 required thedevelopment of new methods that simultaneously performed imaging and calibration, aswell as the design of techniques to validate the result [19]. For the ngEHT to meet its Sci-ence Goals successfully, new techniques for the reconstruction of movies of Sgr A ∗ [26, 8],polarimetric images of M87 [12, 1], and fringe detection and calibration strategies tailoredfor a hybrid array with globally distributed anchor stations and several small dishes [3] arerequired. Ultra-wide bandwidths and simultaneous observing at 230 and 345 GHz will alsomotivate the development of new techniques for multi-frequency imaging, scattering mitiga-tion toward Sgr A ∗ [25], and atmospheric phase transfer [23, 32]. The analysis methods willbe coupled with realistic array data simulation [6] incorporating dynamical source models,instrument characteristics, as well as weather and scheduling models in order to optimizengEHT array design through the use of performance metrics. For example at 230 GHz,using the ( u, v ) filling fraction metric [30] at 500 µ as field-of-view with 100 µ as resolution,the EHT2017 samples 20% of Sgr A ∗ and 25% of M87, while the ngEHT samples 71% ofSgr A ∗ and 99% of M87 (see Figure 6). 8 Organization - Array & Partners
The EHT Collaboration (EHTC), coordinates and conducts EHT observation campaigns andsets the agenda for EHT science and development. A Memorandum of Collaboration bindsa group of thirteen stakeholders that currently include (Academia Sinica Institute of Astron-omy and Astrophysics, University of Arizona, University of Chicago, East Asian Observatory,Goethe-Universitaet Frankfurt, Institut de Radioastronomie Millimetrique, Large MillimeterTelescope Alfonso Serrano, Max Planck Institute for Radioastronomy, MIT Haystack Ob-servatory, National Astronomical Observatory of Japan, Perimeter Institute for TheoreticalPhysics, Radboud University, and the Smithsonian Astrophysical Observatory). Develop-ment of future directions and design is a collaborative and ongoing process within the EHTC,and this APC white paper describes activities that require support of US involvement in thistimely effort. The current EHTC international agreement has also enabled, through partner-ship with ALMA, an open-skies policy for the general astronomy community to use the EHTarray for high resolution, high sensitivity applications. It is anticipated that the currentEHT agreement will be continued, possibly evolving to accommodate a larger operationalcomponent with more observing epochs and a doubling of antennas in the array. Retainingopen-skies access to the EHT array is similarly anticipated with commensurate impact onthe global astronomy infrastructure.Over the course of EHT build-out, two previously used sites (Caltech SubmillimeterObservatory and the Combined Array for Research in Millimeter-wave Astronomy) weredecommissioned. This was balanced by the inclusion of new facilities, and the proposeddeployment of a number of new modest-diameter dishes will minimize impact on the ngEHTto the loss of existing facilities over the coming decade should that occur.
We envisage the design and eventual implementation of this proposed ngEHT as two sepa-rate phases, both of which exist within the overall organization of the EHT project. Phase Iwill be a design process that optimizes a systematic approach to defining science goals anddevelopment specifications, leading to Preliminary Design Reviews (PDR) and Critical De-sign Reviews (CDR) for main elements by ∼ ∼ ∼ $10M for upgrades to existing sites and ∼ $40–$100M for the addition of 8 new sites for the completed construction of the ngEHTarray, placing it at the intersection of medium ( > $20M) and large ( > $70M) ground projectcategories. Operational cost is anticipated at the level of ∼ $5M annually. As with the initialEHT, the next generation EHT would be comprised of a global network of radio telescopeswith observations and scientific utilization managed through worldwide collaboration, andis expected to be funded through multiple international sources and in kind contributionsfrom partners and ngEHT stakeholder institutions. Based on current EHT support, weestimate the net cost to US funding agencies will be ∼ To build upon the success of the EHT in imaging the SMBH at the center of M87 on horizonscales, we propose the design and implementation of a ground-based Next Generation EHT(ngEHT). This instrument will double the number of antennas in the array, incorporate adual-frequency capability, more than double the sensitivity, and increase the dynamic rangeby more than one order of magnitude over the existing EHT. This will enable fundamentalquestions to be tackled, both in physics (e.g., the space-time metric around a rotating blackhole and deviations from the predictions of GR) and in astrophysics (e.g. the launchingmechanism of jets in AGNs and the role of magnetic fields in black hole accretion).The EHT effort has delivered the first black hole image. The ngEHT will fulfill thepromise of a newly emerging field of research in astronomy and physics: precision imagingand time resolution of black holes on horizon scales.10 eferences [1] K. Akiyama et al. “Superresolution Full-polarimetric Imaging for Radio Interferometrywith Sparse Modeling”. In: AJ 153, 159 (Apr. 2017), p. 159. doi :
10 . 3847 / 1538 -3881/aa6302 . arXiv: .[2] S. A. Balbus and J. F. Hawley. “Instability, turbulence, and enhanced transport inaccretion disks”. In:
Reviews of Modern Physics
70 (Jan. 1998), pp. 1–53. doi :
10 .1103/RevModPhys.70.1 .[3] L. Blackburn et al. In: ApJ 882.1, 23 (Sept. 2019), p. 23. doi : . arXiv: .[4] R. D. Blandford and D. G. Payne. “Hydromagnetic flows from accretion discs and theproduction of radio jets”. In: MNRAS 199 (June 1982), pp. 883–903. doi : .[5] R. D. Blandford and R. L. Znajek. “Electromagnetic extraction of energy from Kerrblack holes”. In: MNRAS 179 (May 1977), pp. 433–456. doi : .[6] T. Blecher et al. “MEQSILHOUETTE: a mm-VLBI observation and signal corruptionsimulator”. In: MNRAS 464.1 (Jan. 2017), pp. 143–151. doi : .arXiv: .[7] B. Boccardi et al. “Radio observations of active galactic nuclei with mm-VLBI”. In: The Astronomy and Astrophysics Review issn : 1432-0754. doi : .[8] K. Bouman et al. “Reconstructing Video of Time-Varying Sources from Radio Interfer-ometric Measurements”. In: IEEE Transactions on Computational Imaging issn : 2333-9403. doi : .[9] A. E. Broderick and A. Loeb. “Imaging optically-thin hotspots near the black holehorizon of Sgr A* at radio and near-infrared wavelengths”. In: MNRAS 367 (Apr. 2006),pp. 905–916. doi : . eprint: astro-ph/0509237 .[10] A. E. Broderick et al. “Testing the No-hair Theorem with Event Horizon TelescopeObservations of Sagittarius A*”. In: ApJ 784, 7 (Mar. 2014), p. 7. doi : . arXiv: .[11] A. A. Chael, R. Narayan, and M. D. Johnson. “Two-temperature, Magnetically Ar-rested Disc simulations of the jet from the supermassive black hole in M87”. In: MNRAS486.2 (June 2019), pp. 2873–2895. doi : . arXiv: .[12] A. A. Chael et al. “High-resolution Linear Polarimetric Imaging for the Event HorizonTelescope”. In: ApJ 829, 11 (Sept. 2016), p. 11. doi : .arXiv: .[13] K. Dodds-Eden et al. “Evidence for X-Ray Synchrotron Emission from SimultaneousMid-Infrared to X-Ray Observations of a Strong Sgr A* Flare”. In: ApJ 698 (June2009), pp. 676–692. doi : . arXiv: .1114] S. S. Doeleman et al. “Detecting Flaring Structures in Sagittarius A* with High-Frequency VLBI”. In: ApJ 695 (Apr. 2009), pp. 59–74. doi :
10 . 1088 / 0004 - 637X /695/1/59 . arXiv: .[15] S. S. Doeleman et al. “Event-horizon-scale structure in the supermassive black holecandidate at the Galactic Centre”. In: Nature 455 (Sept. 2008), pp. 78–80. doi : . arXiv: .[16] S. S. Doeleman et al. “Jet-Launching Structure Resolved Near the Supermassive BlackHole in M87”. In: Science
338 (Oct. 2012), p. 355. doi : .arXiv: .[17] Event Horizon Telescope Collaboration et al. “First M87 Event Horizon TelescopeResults. I. The Shadow of the Supermassive Black Hole”. In: ApJ 875, L1 (Apr. 2019),p. L1. doi : .[18] Event Horizon Telescope Collaboration et al. “First M87 Event Horizon TelescopeResults. II. Array and Instrumentation”. In: ApJ 875, L2 (Apr. 2019), p. L2. doi : .[19] Event Horizon Telescope Collaboration et al. “First M87 Event Horizon TelescopeResults. IV. Imaging the Central Supermassive Black Hole”. In: ApJ 875, L4 (Apr.2019), p. L4. doi : .[20] Gravity Collaboration et al. “Detection of orbital motions near the last stable circularorbit of the massive black hole SgrA*”. In: A&A 618, L10 (Oct. 2018), p. L10. doi : . arXiv: .[21] Gravity Collaboration et al. “Detection of the gravitational redshift in the orbit of thestar S2 near the Galactic centre massive black hole”. In: A&A 615, L15 (July 2018),p. L15. doi : . arXiv: .[22] S. W. Hawking. “Black holes in general relativity”. In: Communications in Mathemat-ical Physics
25 (June 1972), pp. 152–166. doi : .[23] M. Honma et al. “First Fringe Detection with VERA’s Dual-Beam System and ItsPhase-Referencing Capability”. In: PASJ 55 (June 2003), pp. L57–L60. doi : . arXiv: astro-ph/0306060 .[24] T. Johannsen and D. Psaltis. “Testing the No-hair Theorem with Observations in theElectromagnetic Spectrum. II. Black Hole Images”. In: ApJ 718 (July 2010), pp. 446–454. doi : . arXiv: .[25] M. D. Johnson. “Stochastic Optics: A Scattering Mitigation Framework for RadioInterferometric Imaging”. In: ApJ 833, 74 (Dec. 2016), p. 74. doi :
10 . 3847 / 1538 -4357/833/1/74 . arXiv: .[26] M. D. Johnson et al. “Dynamical Imaging with Interferometry”. In: ApJ 850, 172 (Dec.2017), p. 172. doi : . arXiv: .[27] D. V. Lal, A. P. Lobanov, and S. Jiménez-Monferrer. “Array configuration studiesfor the Square Kilometre Array - Implementation of figures of merit based on spatialdynamic range”. In: arXiv e-prints (Jan. 2010). arXiv: .1228] B. Lazareff. Alignment Tolerances for ALMA Optics . Tech. rep. 2001, pp. 1–14.[29] J.M. Moran et al.
A Submillimeter-Wavelength Telescope Array: Scientific, Technical,and Strategic Issues . Tech. rep. 1984, pp. 1–145.[30] D. C. M. Palumbo et al. “Metrics and Motivations for Earth-Space VLBI: Time-resolving Sgr A* with the Event Horizon Telescope”. In: ApJ 881.1, 62 (Aug. 2019),p. 62. doi : . arXiv: .[31] D. Psaltis et al. “A General Relativistic Null Hypothesis Test with Event HorizonTelescope Observations of the Black Hole Shadow in Sgr A*”. In: ApJ 814, 115 (Dec.2015), p. 115. doi : . arXiv: .[32] M. J. Reid and M. Honma. “Microarcsecond Radio Astrometry”. In: ARA&A 52 (Aug.2014), pp. 339–372. doi :
10 . 1146 / annurev - astro - 081913 - 040006 . arXiv: .[33] B. S Robinson et al. “TeraByte InfraRed Delivery (TBIRD): a demonstration of large-volume direct-to-Earth data transfer from low-Earth orbit”. eng. In: vol. 10524. SPIE,2018, pp. 105240V–105240V-6. isbn : 9781510615335.[34] A. Tchekhovskoy, R. Narayan, and J. C. McKinney. “Efficient generation of jets frommagnetically arrested accretion on a rapidly spinning black hole”. In: MNRAS 418(Nov. 2011), pp. L79–L83. doi :
10 . 1111 / j . 1745 - 3933 . 2011 . 01147 . x . arXiv: .[35] A. R. Whitney et al. “Demonstration of a 16 Gbps Station − Broadband-RF VLBISystem”. In: PASP 125 (Feb. 2013), p. 196. doi :10.1086/669718