Studying black holes on horizon scales with space-VLBI
Kari Haworth, Michael D. Johnson, Dominic W. Pesce, Daniel C. M. Palumbo, Lindy Blackburn, Kazunori Akiyama, Don Boroson, Katherine L. Bouman, Joseph R. Farah, Vincent L. Fish, Mareki Honma, Tomohisa Kawashima, Motoki Kino, Alexander Raymond, Mark Silver, Jonathan Weintroub, Maciek Wielgus, Sheperd S. Doeleman, Jose L. Gomez, Jens Kauffmann, Garrett K. Keating, Thomas P. Krichbaum, Laurent Loinard, Gopal Narayanan, Akihiro Doi David J. James, Daniel P. Marrone, Yosuke Mizuno, Hiroshi Nagai
AAstro2020 APC White Paper
Studying black holes on horizon scaleswith space-VLBI
Kari Haworth , ∗ , Michael D. Johnson , , ∗ , Dominic W. Pesce , , Daniel C. M.Palumbo , , Lindy Blackburn , , Kazunori Akiyama , , , , Don Boroson , Kather-ine L. Bouman , Joseph R. Farah , , , Vincent L. Fish , Mareki Honma , ,Tomohisa Kawashima , Motoki Kino , , Alexander Raymond , , Mark Silver ,Jonathan Weintroub , , Maciek Wielgus , , Sheperd S. Doeleman , , José L.Gómez , Jens Kauffmann , Garrett K. Keating , Thomas P. Krichbaum ,Laurent Loinard , , Gopal Narayanan , Akihiro Doi , David J. James , ,Daniel P. Marrone , Yosuke Mizuno , Hiroshi Nagai Center for Astrophysics | Harvard & Smithsonian, 60 Gar-den Street, Cambridge, MA 02138, USA Black Hole Initiative at Harvard University, 20 GardenStreet, Cambridge, MA 02138, USA Massachusetts Institute of Technology, Haystack Observa-tory, 99 Millstone Road, Westford, MA 01886, USA National Radio Astronomy Observatory, 520 EdgemontRoad, Charlottesville, VA 22903, USA National Astronomical Observatory of Japan, 2-21-1 Osawa,Mitaka, Tokyo 181-8588, Japan Massachusetts Institute of Technology, Lincoln Laboratory,244 Wood St, Lexington, MA 02421 California Institute of Technology, 1200 East CaliforniaBoulevard, Pasadena, CA 91125, USA University of Massachusetts Boston, 100 William T, Mor-rissey Blvd, Boston, MA 02125, USA Kogakuin University of Technology & Engineering, Aca-demic Support Center, 2665-1 Nakano, Hachioji, Tokyo 192-0015, Japan Mizusawa VLBI Observatory, National Astronomical Ob-servatory of Japan, 2-12 Hoshigaoka, Mizusawa, Oshu, Iwate023-0861, Japan Department of Astronomical Science, The Graduate Uni-versity for Advanced Studies (SOKENDAI), 2-21-1 Osawa,Mitaka, Tokyo 181-8588, Japan Department of Astronomy, University of Massachusetts,01003, Amherst, MA, USA Instituto de Astrofísica de Andalucía-CSIC, Glorieta de laAstronomía s/n, E-18008 Granada, Spain Max-Planck-Institut für Radioastronomie, Auf dem Hügel69, D-53121 Bonn, Germany Steward Observatory and Department of Astronomy, Uni-versity of Arizona, 933 N. Cherry Ave., Tucson, AZ 85721,USA The Institute of Space and Astronautical Science, JapanAerospace Exploration Agency, 3-1-1 Yoshinodai, Chuou-ku,Sagamihara, Kanagawa 252-5210, Japan Institut für Theoretische Physik, Goethe Universität, Max-von-Laue Str. 1, D-60438, Frankfurt am Main, Germany Instituto de Radioastronomía y Astrofísica, UniversidadNacional Autónoma de México, Morelia 58089, México Instituto de Astronomía, Universidad Nacional Autónomade México, CdMx 04510, México ∗ [email protected], [email protected] a r X i v : . [ a s t r o - ph . I M ] S e p Introduction
In 2019, after decades of effort by an international team, the Event Horizon Telescope (EHT)Collaboration presented the first image of a black hole [11, 12, 13, 14, 15, 16]. The impactof this release, both scientifically and among the public, was extraordinary and felt aroundthe globe. The capability to image black holes on event horizon scales enables entirelynew tests of General Relativity (GR) near a black hole and opens a direct window intothe astrophysical processes that drive accretion, flaring, and jet genesis. The EHT image,revealing the supermassive black hole (SMBH) in M87, was captured using a global very-long-baseline interferometry (VLBI) network operating at 230 GHz [12]. Taking the next stepstowards precise tests of GR and time-domain studies of accretion flows will require sharperresolution, higher observing frequencies, and faster sampling of interferometric baselines.The angular resolution of ground-based VLBI is approaching fundamental limits. In-terferometer baseline lengths are currently limited to the diameter of the Earth, imposinga corresponding resolution limit for a ground array of ∼ µ as at an observing frequencyof 230 GHz. Observations at higher frequencies can improve the resolution but become in-creasingly challenging because of strong atmospheric absorption and rapid phase variations,severely limiting the number of suitable ground sites and the windows of simultaneous goodweather at many global locations.The extension of the EHT into space with the addition of a single orbiting element wouldcircumvent these limitations and enable a wealth of new scientific possibilities: • The high time resolution afforded by the rapid ( u, v ) -filling will enable reconstructedmovies of black hole accretion flows. • The improved resolution will increase the number of spatially resolvable black holeshadows to dozens, yielding a corresponding number of black hole mass measurements. • Sharper images will reveal turbulent plasma dynamics, allowing further study of thecrucial role magnetic fields play in black hole feeding and in jet launching.The highest operating frequency of space-VLBI to date is 25 GHz (with RadioAstron),and a number of technical challenges must be overcome to access significantly higher frequen-cies. These challenges would be mitigated by anchoring an orbiting element to the highlysensitive elements in the EHT such as ALMA, the LMT, and NOEMA, permitting the useof a modest aperture (of ∼ -meter size) in space and reducing costs. In this white paper,we present a conceptual mission design for a first such submillimeter space mission, whichwe expect to fall within the medium cost category of the Astro2020 survey. A companionwhite paper details the concurrent expansion of the EHT ground array. We review the key science drivers for submillimeter space-VLBI. For additional details, see[27], [17], and [33]. 1 − λ ) − − N o rt h - S o u t h B a s e li n e ( G λ ) Single Transit Coverage
Space-GroundGround-Ground − − λ ) − − N o rt h - S o u t h B a s e li n e ( G λ ) Full Day Coverage
Space-GroundGround-Ground
Figure 1: Left: 230 GHz baseline coverage of Sgr A ∗ of the EHT2020 array with and withouta polar LEO over 45 minutes. Right: Same as left, over a full day. In both cases, the additionof the polar LEO dramatically improves the baseline coverage. ∗ Measuring the shape and diameter of the black hole shadow in Sgr A ∗ would provide a nullhypothesis test of GR [29] and would yield precise constraints on the surrounding spacetimeand the possibility for black hole alternatives. However, the relatively small mass of thisblack hole [ . × M (cid:12) ; 19] results in correspondingly short dynamical timescales (of orderten minutes) for the system, and the current EHT array lacks sufficient baseline coverage toform images on these timescales.The addition of a Low-Earth Orbit (LEO) dish to the EHT array would provide sufficientinstantaneous ( u, v ) -coverage to be sensitive to the (cid:46) hour dynamical timescales of Sgr A ∗ [27]. When observing in tandem with ground-based stations, a LEO station accrues coveragein one 45-minute half-orbit comparable to a full night of observation with the expanded EHTarray expected for 2020 (EHT2020), as shown in Figure 1.Figure 2 shows static images of a GRMHD simulation of Sgr A ∗ formed on half-orbittimescales for the EHT2020 with and without a LEO co-observing [6, 7, 8]. A LEO orbiterenables reconstruction of the black hole shadow and fine wisps of plasma on single half-orbittimescales.However, the evolutionary timescale of Sgr A* may not permit static imaging, even overa period as short as 45 minutes. However, the combined ( u, v ) -plane filling rate provided bya LEO contribution to the EHT array also enables high-fidelity dynamical imaging of Sgr A ∗ [27, 3, 21]. Figure 3 shows dynamical reconstructions of the same GRMHD simulation usingthe EHT2020 and EHT2020+LEO arrays. The rapid baseline sampling of the orbiter isnecessary to recover complex structure in an evolving accretion flow.Reconstructed movies of Sgr A ∗ would elucidate the nature of coherent orbiting features2 µ as G H z Blurred Sim.
NRMSE = 0.31
EHT2020
NRMSE = 0.18
EHT2020+LEO µ as G H z NRMSE = 0.45 NRMSE = 0.29
Figure 2: Left column: a ray-traced snapshot from a 40 degree inclined GRMHD simulationof Sgr A ∗ [8], blurred by the ensemble average scattering kernel at 230 and 345 GHz [22].Center, Right columns: reconstructions with the EHT2020 and EHT2020+LEO arrays using45 minutes of synthetic data. Normalized root-mean-square error (NRMSE) is shown relativeto the blurred true image. A single Low Earth Orbiter enables successful reconstruction ofthe black hole shadow and diffuse plasma features, while the ground arrays have insufficientinstantaneous ( u, v ) coverage in this short interval. As discussed in [27], sparser sampling at345 GHz worsens reconstructions that use very brief observations.such as “hotspots” ∗ [18] and the origin of the flaring events observed to occur approximatelydaily across many wavebands [24, 39]. A large ( u, v ) -plane filling fraction will also enablerobust modeling in the visibility domain [4, 18], allowing measurements of both the radiusand period of very compact orbits. While adding a LEO station to the EHT array would not substantially increase the availablephysical lengths of baselines, it could improve the angular resolution by extending the ob-serving frequency from 230 GHz up to 690 GHz. For Earth-diameter baselines, the angularresolution improves from ∼ µ as at 230 GHz to ∼ µ as at 690 GHz. Even with just twoground stations co-observing with the orbiter, the left panel of Figure 4 shows that a LEOstation observing at 690 GHz could accumulate a comparable ( u, v ) -plane filling fraction tothe current (ground-based) EHT array operating at 230 GHz. Imaging with the 690 GHzcoverage alone would thus provide improved resolution, albeit with a limited dynamic range. ∗ By “hotspot,” we refer generically to any luminous and compact region on a short–period orbit. = 0 M S i m u l a t i o n
60 as0.4 E H T E H T + L E O t=250 M0.470.33 t=500 M0.50.35 t=750 M0.460.33 Figure 3: 345 GHz Starwarps reconstructions of an ensemble-average scattered [22] GRMHDsimulation [8] of Sgr A* with the EHT2020 and EHT2020+LEO arrays (matching the sim-ulation in in Figure 2). The normalized root-mean-square error relative to the blurred trueimage is quoted at the bottom of each panel, showing pixel-wise accuracy for each recon-struction. Frames from the reconstruction are shown every 250 M, where M is the black holemass expressed as unit of time via t M = GM/c . The improved baseline coverage of theexpanded and space-enabled array yield a sharp reconstruction of the black hole shadow, aswell as diffuse extended plasma features, while the current ground array produces imagesdominated by artifacts that do not accurately localize the peak of emission. For additionaldetails, see Palumbo et al. [27].In addition, with a LEO station operating at very high frequencies and able to perform ob-servations across a wide range of baseline lengths, we expect to sample the power spectrumof the turbulent accretion flow on very fine spatial scales, for the first time observation-ally testing our understanding of the magnetorotational instability and angular momentumtransport in the inner part of the accretion disks [1, 20, 38].Finer angular resolution also provides access to additional targets with spatially resolvedblack hole shadows. Given a uniform distribution of SMBHs in flat space, we expect thenumber of sources that can be resolved ( N ) to increase roughly as the cube of the maximum ( u, v ) -distance. At ∼ µ as angular resolution, the number of known SMBHs that are expectedto have resolvable black hole shadows will increase from N ≈ (Sgr A ∗ and M87) to N (cid:38) (see right panel of Figure 4). † Each spatially resolved shadow provides a corresponding black † This number accounts only for those SMBHs with well-measured masses from [36], and so represents a roughlower limit on the number of sources with spatially resolvable black hole shadows. Synchrotron opacity will East-West Baseline (G λ ) − − − N o r t h - S o u t h B a s e li n e ( G λ ) ALMA-SPT-LEO at 690 GHz µ as) − − − − F l u x d e n s i t y ( J y ) G H z G H z G H z G H z Figure 4:
Left : Sgr A ∗ baseline coverage of the ALMA-SPT-LEO subarray at 690 GHz, withpoints shown every minute over the course of 24 hours. Right : 230–690 GHz flux versusblack hole shadow size for SMBHs with known masses; fluxes have been taken from theNASA/IPAC Extragalactic Database, and masses were tabulated by [36]. Minimum fringespacings at different observing frequencies are shown as vertical dashed lines.hole mass-to-distance ratio estimate, a constraint on the black hole spin, and an additionalopportunity to study SMBH accretion flow and jet physics on horizon scales.
The sensitivity of an interferometric baseline depends on the geometric mean of the twotelescope sensitivities, the recorded bandwidth, and the coherent integration time. The firstproperty allows small telescopes (e.g., an orbiter) to form sensitive baselines when pairedwith a large telescope (e.g., ALMA). The second allows digital enhancements (e.g., widerrecorded bandwidths) to offset limitations in telescope sensitivity. The third ties sensitivityto phase stability, which is limited by the atmosphere and reference frequency.The RMS noise on a VLBI baseline is σ RMS = η − Q (cid:112) ( SEFD × SEFD ) / (2 ∆ ν T ) , whereSEFD and SEFD are the system-equivalent flux densities (system noise in units of effectiveflux above the atmosphere) of each antenna, ∆ ν and T are the integration bandwidth andtime respectively, and η Q ≈ . is a loss factor for 2-bit quantization.For Sgr A ∗ , the flux density currently seen on the longest EHT baselines is ∼
100 mJy[23], guiding sensitivity requirements for an orbiter. To maintain a S/N of 4 over 32 GHz(averaging two polarizations) and 5-seconds of integration (sufficient to track rapid atmo-spheric phase at 230 GHz), an orbiter–anchor station baseline would require an orbiter SEFDless than . × (ALMA), . × (NOEMA), . × (IRAM30m), . × (LMT),and . × (SMA) to connect successfully to one of the anchor stations in Table 1. Once preclude horizon-scale observations for some sources, but the opacity decreases with increasing frequency. LBA,GMVA ( ) - m Orbiter: G H z G H z G H z G H z ( σ ≤ m J y ; t c = s ) G H z ( s ) G H z ( . s ) EHT ( ) EHT ( ) - Orbiter SEFD ( Jy ) B a nd w i d t h ( G H z ) Figure 5: Sensitivity for a single LEOsatellite on a baseline to ALMA as afunction of the orbiter SEFD and aggre-gate bandwidth. Colored regions show ex-pected sensitivity requirements for fringedetections for Sgr A ∗ and M87 (see §3 fordetails). With the current bandwidth ofthe EHT, a 3.5-meter orbiter would haveample sensitivity for strong detections at230 and 345 GHz. At 690 GHz, detectionswould require a larger dish or longer in-tegration times, which may be possiblewith simultaneous multi-frequency obser-vations.phase referenced to the anchor station, the orbiter could also connect to other smaller dishesin a ground array through further coherent averaging.Once the required sensitivity for the orbiter is understood, we can conduct trade studiesto determine the instrumentation requirements (see Figure 5). Selecting technology that iseither presently available or on a near-term development path, the satellite is envisioned tohave a 3 - 4 meter dish and will process two polarizations of two 8 GHz bands of data at 230GHz, 345 GHz, or 690 GHz. Next, we will discuss the details of the instrument subsystems. Most technical elements of a VLBI satellite are independent of the specific orbital geometry.We now discuss the current status and development trajectories for these common elements,summarized in Figure 6. These include the antenna (§4.1), receiver (§4.2), digital processingsystem (§4.3), timing reference (§4.4), on-board storage and downlink (§4.5), and bus (§4.6).6 urface SEFD (Jy)Telescope Aperture rms ( µ m)
86 GHz 230 GHz 345 GHz 690 GHzALMA 54 ×
12 m 25 33 51 128 1017NOEMA 12 ×
15 m 35 80 188 586 . . .LMT 50 m 80 143 563 2590 . . .
IRAM30m 30 m 55 365 962 2685 . . .
SMA × m 20 . . . . . . GBT 100 m 300 118 . . . . . . . . . ngVLA (full) × m 160 7.4 . . . . . . . . . ngVLA (compact) × m 160 17 . . . . . . . . . Space 3.5 m 15 21600 36600 113000 260000Table 1: Existing and future planned large ground facilities that could serve as anchor sta-tions for a space antenna. For the sensitivity estimates, we have assumed observations at 45degree elevation, with zenith opacity of 0.05 (ALMA, SMA), 0.09 (NOEMA, IRAM30m, 0.13(LMT) at 230 GHz, and 0.05 (GBT, and the planned ngVLA [34]) at 86 GHz. Additionalbeam efficiency of 0.7 was assumed for all sites, except for 0.8 at SMA and a hypothet-ical 3.5 m Space orbiter. Receiver temperatures are taken from stations specifications orprojections. Receiver temperatures for the Space antenna are based on maturing HEMTtechnology and are listed in subsection 4.2.
Data Processing System 3-meter Dish Receiver with Cooling Block Down Converter Digital Back End Storage/ DownlinkTiming Ref.GPS
Rigid or foldableFoldable dish requires development from TRL-4Rigid similar to Herschel HEMT-based systems maturing quickly Ground-based prototype in developmentSpace-rated OCXO’s commercially available. Space-qualified packaging and frequency-stabilizing feedback requires development Commercially available high-capacity SSDs>200 Gbps downlink to be demonstrated by NASA TBIRD 2020
Figure 6: Space-VLBI instrument for a LEO orbiter with a summary of component devel-opment state.
Observing weak sources requires large antenna apertures; however, because anchoring to anEarth-based dish allows for a relatively small reflector in space, the aperture size requirementof 3 - 4 meters is grounded in what’s technologically realistic in the near-term. Spacecraftand launch vehicle constraints, not yet determined, define the requirement of size at launch.We review the technology for both a fixed dish, which would require a dedicated launchvehicle, and a deployable dish.The Herschel satellite flew a fixed 3.5-m antenna with <6 micron surface accuracy [28],7ufficient for 230, 345, or 690 GHz operation. A similar concept was developed by Northrup-Grumman Innovation Systems, with a 2-m prototype tested successfully in a balloon launch.Extending to 3.5 meters will require retooling, but the path to build is well understood.If the launch volume is limited, the precision solid surface reflector segments and support-ing structure must be collapsed for launch and deployed once in orbit. Large deployable meshreflectors operating <40 GHz have significant flight heritage [26], but deployable reflectorsoperating at >200 GHz present two new challenges. First, operation at these higher fre-quencies will require RMS surface accuracy of <20 microns, a level previously unattained inheritage systems. Second, solid surface deployable reflectors will require different, unproven,methods of collapsing the reflector. High-Strain Composite (HSC) precision deployable struc-tures could be used to achieve the precision and alternate deployment approaches with lowersize, weight and power than existing technology, and can deploy to within 2.5 microns oftheir previous position[10]. This technology, currently at Technology Readiness Level (TRL)4, could be matured for flight in 3-5 years.
In general, sub-millimeter receivers sensitive enough for VLBI observations require cryogeniccooling. In space, cooling to 50 K is routine using a mature single-stage Stirling cooler [9]and is a reasonable target for a near-term mission. That temperature range prohibits usingsuperconducting-insulator-superconductor (SIS) mixers and instead points to the use of highelectron mobility transistor (HEMT) technology. Quickly-maturing research-grade devicesexist for 230, 345, and 690 GHz bands with noise temperatures of 100 [31], 300 [30, 9], and600 K [30, 9], respectively. Additionally, at frequencies above 345 GHz, cooled Schottkydiode receivers might be suitable and have flown previously [25].
The digital processing system (DPS) is required to convert two 8 GHz analog intermediatefrequency (IF) signals into digital packets for downlink. The DPS consists of block downcon-verter (BDC) and a digital backend (DBE). The DBE has two Analog to Digital Converters(ADCs) and a Field Programmable Gate Array (FPGA). Nyquist sampling of 8 GHz requiresthat the ADC samples at 16 GSps. With four bit samples, this is 128 Gbps on the FPGAinput quantized to >64 Gbps (including packet overhead) on the output. While quantizationand packetization require minimal processing, the high data throughput means selecting ahigh-performance FPGA.Current ground systems run at 8 Gbps to process two 2-GHz bands [37]. A new systemdesigned to take advantage of progress made in FPGA and ADC technologies will run at128 Gbps to process four 8-GHz bands, with a prototype expected within a year. This canbe deployed as is to the ground-based system; to use in space will require a redesign ofthe board for space qualification and to reduce power consumption. The ADC chips andother components have corresponding space versions. Advances have been made in radiationtolerant high-performance FPGAs, with the Xilinx Kintex Ultrascale FPGA providing highdata rates in a space-qualified part. 8he pathway from the ground-based DPS to a space-ready design is well-understood andis estimated to take three years.
VLBI requires an extremely stable time reference. The precise specifications required aregoverned by the observing frequency, the integration time, and the tolerance for phase deco-herence. Current ground-based arrays deploy hydrogen masers for this purpose. At 300 GHzobserving frequency, a hydrogen maser has adequate phase stability of a part in or better,corresponding to coherence losses at 345 GHz less than 5%, on timescales up to 100 s [35].Hydrogen masers have flight heritage but are large (approx. 35 kg; [2]) and requiresignificant power (approx. 100 W; [5]). High-performance oven-controlled quartz crystaloscillators (OCXO) are small and power efficient, and in laboratory environments, theirshort-term stability approaches what is needed for VLBI. However, current space-qualifiedOCXO devices don’t meet the requirements, so development of the packaging and frequency-stabilizing feedback is needed if they are to be used as an orbiting VLBI time reference.We estimate that OCXO development for space could be completed within two years. Packets will be generated at 64 Gbps from the DBE. Assuming a LEO orbit and a 1/2 dutycycle, this results in approximately 22 TB of data per orbit to be handled by a combinationof downlink and onboard storage.Downlink technology is currently available via laser communication at >100 Gbps. TheNASA TBIRD (TeraByte InfraRed Delivery) design is based on commercially-available,highly-integrated 100 Gbps modems, multiplexed into the single telescope using commercially-available wavelength-division-multiplexing fibers, so the downlink speed is configurable tomultiple 100’s of Gbps at the cost of power and weight. A TBIRD system that will deliverbursts of 200 Gbps to a single ground terminal from a LEO CubeSat is being prepared fora 2020 flight [32]. The ground terminals will be small and easily deployable so that severalcan be fielded, thus increasing the frequency of both seeing the LEO pass over and having acloud-free line of sight.For high capacity on-board storage, solid state recorders are available today with 1 TBoffered in chewing-gum-sized packages. These are easily multiplexed to much higher datavolumes.
VLBI requires that the source be well-localized within the primary beam of each single radiotelescope dish. For a 3.5-meter dish diameter, 1.3 mm observing wavelength, and a pointingrequirement to within ∼
10% of the beam width, this gives δθ (cid:46) , which is readily providedthrough independent optical star-tracking.Antenna position and velocity must be known in order to coherently average the cor-related signal across finite bandwidth and time. Initial searches are conducted with widesearch windows in the associated delay, delay-rate, and acceleration, with residual values9eing used to refine orbit determination, as is currently done for RadioAstron [40]. Delaywindows of ∼ few µ s ( ∼ km in distance along line-of-sight) are routinely fit in high-bandwidthVLBI post-processing, as well as delay-rate windows of ∼ few ps/s ( ∼ mm/s). For integra-tions along Earth orbit, there are often residual acceleration terms which need to be fit aswell. This is computationally demanding, but adds a level of complexity that is similar tothat required to fit the unknown velocity. Expansion of the EHT in the coming decade will include both extending the existing groundarray, as detailed in a companion white paper, and adding at least one site in space. Thoughtechnological considerations differ, these two goals fall under one mission concept: to buildon the success of the current EHT and expand spatially resolved studies of black holes.A space based antenna in LEO would enable routine observations at 345 and 690 GHz.Operation of the EHT array at higher frequency would permit imaging of dozens of addi-tional black holes. It would also enable higher precision measurements of the photon ring intargets already observed by EHT, as well as finer-resolution studies of the turbulent plasmasurrounding the black hole. Relative to 230 GHz observations, these higher frequencies havelower synchrotron opacity and, for Sgr A ∗ , sharply reduced interstellar scattering.The position of an antenna in LEO changes rapidly compared to the usual changes inVLBI antenna positions driven by the rotation of the Earth. Baselines to an LEO an-tenna thus sweep rapidly through the ( u, v ) -plane, providing sufficient coverage on shorttimescales to reconstruct images of Sgr A ∗ on its dynamical timescale. The combination ofhigher observing frequencies and faster baseline sampling—unique to an orbiting VLBI arrayelement—will thus permit time-resolved studies of the Galactic Center and precision testsof general relativity.While we have not identified any fundamental obstacles in the subsystem technologydevelopment paths, all system components require further work on the scale of 2 to 5 yearsto be ready for a space extension to the EHT. The technical development toward launchingan exploratory high-frequency VLBI orbiter in LEO would lay the path for future expansionsof the array into medium Earth orbit (MEO; b max ≈ –3 D ⊕ ) or geosynchronous orbit (GEO; b max ≈ D ⊕ ), enabling a further factor of ∼ µ as-scale imaging.10 eferences [1] 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 .[2] P. Berthoud et al. “Development of the active space hydrogen maser for the ACESspace experiment of ESA”. In:
Proceedings of the 20th European Frequency and TimeForum . IEEE. 2006, pp. 379–383.[3] 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 : .[4] 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 .[5] L. Cacciapuoti. “The ACES Mission: Scientific Objectives and Present Status”. In: ESASpecial Publication . Vol. 621. ESA Special Publication. June 2006, p. 164.[6] A. A. Chael et al. “High-resolution Linear Polarimetric Imaging for the Event HorizonTelescope”. In: ApJ 829, 11 (Sept. 2016), p. 11. doi : .arXiv: .[7] A. A. Chael et al. “Interferometric Imaging Directly with Closure Phases and ClosureAmplitudes”. In: ApJ 857, 23 (Apr. 2018), p. 23. doi : .arXiv: .[8] A. Chael et al. “The role of electron heating physics in images and variability of theGalactic Centre black hole Sagittarius A*”. In: MNRAS 478 (Aug. 2018), pp. 5209–5229. doi : . arXiv: .[9] Goutam Chattopadhyay et al. “A 340 GHz cryogenic amplifier based spectrometerfor space based atmospheric science applications”. In: International Conference on In-frared, Millimeter, and Terahertz Waves, IRMMW-THz (2017), pp. 1–2. issn : 21622035. doi : .[10] Michael A Echter et al. “Recent Developments in Precision High Strain CompositeHinges for Deployable Space Telescopes”. In: . 2018, p. 0939.[11] Event Horizon Telescope Collaboration et al. “First M87 Event Horizon TelescopeResults. I. The Shadow of the Supermassive Black Hole”. In: ApJ 875.1, L1 (Apr.2019), p. L1. doi : .[12] Event Horizon Telescope Collaboration et al. “First M87 Event Horizon TelescopeResults. II. Array and Instrumentation”. In: ApJ 875.1, L2 (Apr. 2019), p. L2. doi : .[13] Event Horizon Telescope Collaboration et al. “First M87 Event Horizon TelescopeResults. III. Data Processing and Calibration”. In: ApJ 875.1, L3 (Apr. 2019), p. L3. doi : . 1114] Event Horizon Telescope Collaboration et al. “First M87 Event Horizon TelescopeResults. IV. Imaging the Central Supermassive Black Hole”. In: ApJ 875.1, L4 (Apr.2019), p. L4. doi : .[15] Event Horizon Telescope Collaboration et al. “First M87 Event Horizon TelescopeResults. V. Physical Origin of the Asymmetric Ring”. In: ApJ 875.1, L5 (Apr. 2019),p. L5. doi : .[16] Event Horizon Telescope Collaboration et al. “First M87 Event Horizon TelescopeResults. VI. The Shadow and Mass of the Central Black Hole”. In: ApJ 875.1, L6(Apr. 2019), p. L6. doi : .[17] Vincent L. Fish, Maura Shea, and Kazunori Akiyama. “Imaging Black Holes and Jetswith a VLBI Array Including Multiple Space-Based Telescopes”. In: arXiv e-prints ,arXiv:1903.09539 (Mar. 2019), arXiv:1903.09539. arXiv: .[18] GRAVITY Collaboration et al. “Detection of orbital motions near the last stable cir-cular orbit of the massive black hole SgrA*”. In: A&A 618, L10 (Oct. 2018), p. L10. doi : . arXiv: .[19] 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: .[20] Xiaoyue Guan et al. “Locality of MHD Turbulence in Isothermal Disks”. In: ApJ 694.2(Apr. 2009), pp. 1010–1018. doi : . arXiv: .[21] M. D. Johnson et al. “Dynamical Imaging with Interferometry”. In: ApJ 850, 172 (Dec.2017), p. 172. doi : . arXiv: .[22] M. D. Johnson et al. “The Scattering and Intrinsic Structure of Sagittarius A* at RadioWavelengths”. In: ApJ 865, 104 (Oct. 2018), p. 104. doi : .arXiv: .[23] R.-S. Lu et al. “Detection of Intrinsic Source Structure at ∼ Schwarzschild Radii withMillimeter VLBI Observations of Sagittarius A*”. In: ApJ 859, 60 (May 2018), p. 60. doi : . arXiv: .[24] D. P. Marrone et al. “An X-Ray, Infrared, and Submillimeter Flare of Sagittarius A*”.In: ApJ 682 (July 2008), pp. 373–383. doi : . arXiv: .[25] Gary J. Melnick et al. “The [ITAL]Submillimeter Wave Astronomy Satellite[/ITAL]:Science Objectives and Instrument Description”. In: The Astrophysical Journal doi : . url : https://doi.org/10.1086%5C%2F312856 .[26] Thomas Murphey. “Historical perspectives on the development of deployable reflec-tors”. In: . 2009, p. 2605. 1227] D. C. M. Palumbo et al. “Metrics and Motivations for Earth-Space VLBI: Time-Resolving Sgr A* with the Event Horizon Telescope”. In: arXiv e-prints (June 2019).arXiv: .[28] G. L. Pilbratt et al. “Herschel Space Observatory. An ESA facility for far-infrared andsubmillimetre astronomy”. In: A&A 518, L1 (July 2010), p. L1. doi : . arXiv: .[29] D. Psaltis et al. “A General Relativistic Null Hypothesis Test with Event Horizon Tele-scope Observations of the Black Hole Shadow in Sgr A*”. In: ApJ 814, 115 (Dec. 2015),p. 115. doi : . arXiv: .[30] Theodore J. Reck, William Deal, and Goutam Chattopadhyay. “Cryogenic performanceof HEMT amplifiers at 340GHz and 670GHz”. In: IEEE MTT-S International Mi-crowave Symposium Digest (2014), pp. 1–3. issn : 0149645X. doi : .[31] Theodore Reck et al. “A 230 GHz MMIC-Based Sideband Separating Receiver”. In: IEEE Transactions on Terahertz Science and Technology issn :2156342X. doi : .[32] 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.[33] Freek Roelofs et al. “Simulations of imaging the event horizon of Sagittarius A* fromspace”. In: A&A 625, A124 (May 2019), A124. doi : .arXiv: .[34] R. J. Selina et al. “The ngVLA Reference Design”. In: Science with a Next GenerationVery Large Array
517 (Dec. 2018), p. 15.[35] A. R. Thompson, J. M. Moran, and G. W. Swenson Jr.
Interferometry and Synthesisin Radio Astronomy, 3rd Edition . Springer, 2017. doi : .[36] Remco C. E. van den Bosch. “Unification of the fundamental plane and Super MassiveBlack Hole Masses”. In: ApJ 831.2, 134 (Nov. 2016), p. 134. doi :
10 . 3847 / 0004 -637X/831/2/134 . arXiv: .[37] L. Vertatschitsch et al. “R2DBE: A Wideband Digital Backend for the Event HorizonTelescope”. In: PASP 127 (Dec. 2015), p. 1226. doi : .[38] Justin Walker, Geoffroy Lesur, and Stanislav Boldyrev. “On the nature of magneticturbulence in rotating, shearing flows”. In: MNRAS 457.1 (Mar. 2016), pp. L39–L43. doi : . arXiv: .[39] F. Yusef-Zadeh et al. “Simultaneous Multi-Wavelength Observations of Sgr A* During2007 April 1-11”. In: ApJ 706 (Nov. 2009), pp. 348–375. doi : . arXiv: .[40] M. V. Zakhvatkin et al. “RadioAstron orbit determination and evaluation of its resultsusing correlation of space-VLBI observations”. In: arXiv e-prints , arXiv:1812.01623(Dec. 2018), arXiv:1812.01623. arXiv:1812.01623 [astro-ph.IM]