M

Report of the Panel on Radio, Millimeter, and Submillimeter Observations from the Ground

M.1 EXECUTIVE SUMMARY

Observations at radio, millimeter, and submillimeter (RMS) wavelengths have played a critical role over the past decade in advancing our understanding of fundamental physics, cosmology, and the formation and evolution of cosmic structures on all scales (planets, stars, galaxies, and galaxy clusters). This record encompasses discoveries made by two types of RMS facilities: experiments, which are designed, built, and used by dedicated teams to address focused sets of science questions, and observatories, which offer diverse and flexible sets of observational capabilities to broad communities of astronomers and can therefore address wide ranges of science questions. In the next decade, existing and new RMS facilities of both types are poised to make exciting discoveries related to nearly all of the high-priority science questions and discovery areas identified by the Astro2020 science panels. In this report, the Panel on Radio, Millimeter, and Submillimeter Observations from the Ground (“RMS panel”) outlines the investments by U.S. federal agencies that would be most valuable for ensuring that this potential is realized.¹ This vision is balanced in terms of operational mode (experiment versus observatory) and scale (large versus medium versus small), and includes the following elements, listed in order of decreasing cost of construction per project:

• Design, construction, and early operation of a large new observatory, the Next Generation Very Large Array (ngVLA). This facility is conceived as an array of antennas distributed across North America, operating at frequencies from 1.2 to 116 GHz, and would replace two existing federally funded facilities—the Karl G. Jansky Very Large Array (VLA) and the Very Long Baseline Array (VLBA). The ngVLA would provide dramatic improvements in the ability to detect and image faint astronomical signals at high angular resolution, enabling routine observations of cold gas flows inside distant galaxies, annular gaps produced by newly formed planets in the inner parts of protoplanetary disks, and features on the surfaces of nearby stars. With broad, flexible capabilities and science-ready data

___________________

¹ See Appendix A for the overall Astro2020 statement of task, the set of panel descriptions that define the panels’ tasks, and additional instructions given to the panels by the steering committee.

• products accessible to a diverse community of users, the ngVLA would epitomize the strengths of observatory-mode science and enable discoveries in new areas that cannot currently be imagined.
• Design, construction, and early operation of a large new “stage 4” experiment to study the cosmic microwave background (CMB), CMB-S4. This facility would build on the achievements of previous (second- and third-generation) CMB experiments in Antarctica and Chile, deploying a suite of small- and large-aperture telescopes equipped with unprecedented numbers of detectors spanning many bands across a decade in frequency. Working together, the CMB-S4 telescopes would conduct two complementary surveys probing the afterglow of the Big Bang, placing unprecedentedly tight constraints on the strength of primordial gravitational waves and the contribution of light particles to the density of matter in the universe. In addition to its unique ability to address long-standing questions of cosmology and fundamental physics, CMB-S4 would be poised to shed light on the growth of cosmic structures (in particular, formation of the first galaxy clusters) and the properties of explosive transients in a new frequency regime.
• Significant funding to support mid-scale projects. In the past decade, the National Science Foundation (NSF) mid-scale programs offering up to $70 million of funding per project have provided essential support to a number of cutting-edge RMS initiatives. While future funding decisions would be determined by competitive proposal calls, the RMS panel has identified four key areas in which outstanding opportunities exist for new mid-scale facilities to address compelling science questions. Listed in order of nearest to most distant observational target(s), these are:
  • Broadband, high-cadence, spectropolarimetric imaging of the Sun to trace flares, shocks, and coronal mass ejections and to understand the drivers of space weather;
  • High-resolution imaging of jets driven by supermassive black holes in the centers of galaxies to determine how such jets are launched and powered;
  • Surveying the static and time-variable radio sky with an innovative new “radio camera” to address a wealth of science questions using statistical samples of star-forming galaxies and fast radio bursts; and
  • Mapping the evolution of neutral atomic hydrogen (HI) gas in the very early universe, at epochs before galaxies and black holes were sufficiently numerous to ionize it.

Calls for mid-scale funding issued on a regular basis over the next decade would accommodate a range of projects reaching readiness on different time scales, and would enable agile and cost-effective approaches to addressing new science opportunities.

• Ongoing support for three key capabilities—long-term timing of pulsars, development of new instrumentation (including software), and mitigation of radio frequency interference (RFI)—that are not tied to single facilities. The precision timing of pulsars encompasses both individual systems and large networks of objects, with the latter aimed at enabling the detection of low-frequency gravitational waves, and requiring ongoing searches to expand existing networks. These efforts require continued access to substantial observing time on the Arecibo telescope (see Box M.1) and the Green Bank Telescope (GBT), with the ngVLA and a future mid-scale facility potentially contributing as well near the end of the decade. Given the critical importance of Arecibo and the GBT for pulsar timing, and their value in addressing other high-priority science questions, continued federal (and, if available, state) funding to support healthy fractions of “open time” scientific observations at these observatories would be very important. More broadly, to maintain the capacity to build and exploit innovative new RMS instrumentation (including software) and train the next generation of instrument builders, dedicated federal funding (e.g., via the NSF Advanced Technologies and Instrumentation program) remains critical,

• as does the need for platforms where new instruments can be deployed. The rapid increase in RFI from terrestrial sources and satellite constellations poses a severe threat to radio astronomy. The alleviation of this threat will require increased support for RFI protection and mitigation for all new and existing ground-based facilities operating at RMS wavelengths.

The RMS panel recognizes its role as helping to populate a menu of options from which the Astro2020 steering committee will choose in recommending an ambitious program for the next decade. As an input to this process, the RMS panel has also identified three top-level principles governing its overall vision. First, it would be important for facility operations budgets over the next decade to include full support for the U.S. share of the Atacama Large Millimeter/submillimeter Array (ALMA), a hugely productive and scientifically vibrant observatory that has set a new standard for how an RMS facility can serve the entire astronomical community. Second, RMS science will flourish best with a balanced program of investments at large, medium, and small cost scales, where “small” investments include individual investigator grants that can support the training of graduate students and postdoctoral researchers—and with the level of investment at any one cost scale not becoming so large that balanced investment at a smaller cost scale is precluded. Third, in the construction and operation of RMS facilities, the astronomical community needs to engage constructively, respectfully, and substantively with stakeholders from outside that community in managing cultural and environmental concerns.

M.2 THE RMS LANDSCAPE IN 2020

M.2.1 Looking Back

In August 1931 in Holmdel, New Jersey, using a 100-foot-long antenna mounted on four Ford Model T tires, American physicist and engineer Karl Jansky detected radio waves originating in the center of the Milky Way. Following developments in radar in World War II, the discipline of RMS astronomy grew and ramified to produce a stunning range of technological advances and scientific discoveries. RMS telescopes can observe in isolation, as single dishes sensitive to the diffuse emission from interstellar gas clouds and the faint pulses of spinning neutron stars, or in concert, as arrays (sometimes continental or intercontinental in scale) producing exquisitely sharp images of galaxies, black holes, and protoplanetary disks. RMS detectors on these telescopes can distinguish the signatures of specific atoms and molecules from those of thermal plasmas that glow because they are warm, and in turn from the nonthermal emission produced by charged particles accelerating in strong magnetic fields. Modern astronomers can leverage these RMS technologies and techniques to study phenomena ranging from explosive events on the surface of the Sun to tiny ripples in the cosmic microwave background that represents the afterglow of the Big Bang.

An important dimension of RMS astronomy is the fact that groundbreaking discoveries are regularly made by facilities that operate in two different modes. Experiments are conceived, constructed, and exploited by dedicated teams to address focused sets of science questions, with design parameters optimized to deliver the best possible performance in addressing those questions. Observatories are designed and built to be broadly capable, and thus able to address wide ranges of science questions—often much wider than the original designers and builders imagined. An observatory achieves its full potential by eliciting the most creative and ambitious ideas from the broadest possible community of astronomers; this consideration explains why the National Radio Astronomy Observatory (NRAO), the Green Bank Observatory (GBO), and the Arecibo Observatory have made it part of their core mission to expand their user communities through support, outreach, and training activities. Experiments and observatories are essential complements to each other in the progress of science. Figure M.1 shows examples of discoveries made since 2010 by observatories, experiments, and a combination of the two.

Scientifically, highlights from the past decade of RMS observations have extended across all areas of astronomy. Multiple experiments have detected lensing B-mode signatures in the CMB, and the CMB-galaxy lensing cross power spectrum. Fast radio bursts (FRBs) have gone from unconfirmed curiosities to daily events that show promise as cosmological tools for probing the distribution of ionized gas across the universe, as new facilities yield increasing numbers of detections and localizations. Centimeter wavelength measurements of neutron star masses more than twice that of the Sun have provided the strongest constraint yet on the super-nuclear equation of state, while radio observations of a binary neutron star merger have offered robust constraints on the geometry of the explosion and the expansion of the universe. At (sub)millimeter wavelengths, ALMA’s unparalleled sensitivity and resolution have allowed unprecedentedly detailed mapping of the molecular gas and dust in nearby galaxies (from which stars form) and around young stars (from which planets form). The striking ALMA image of the disk surrounding the young star HL Tau revealed exquisitely detailed structure—nested rings and gaps thought to be created by embedded planets—which, in conjunction with other studies, strongly suggests that planet formation is more extensive, more diverse, and earlier-starting than anticipated. ALMA spectral line observations have revealed the kinematics of disks around forming stars and black holes, thereby enabling measurements of their masses, and have probed the gas mass reservoirs in high-redshift galaxies, revealing the factors that drive the cosmic star formation history. Detailed images of gravitationally lensed galaxies have demonstrated the potential to detect and measure the masses of dark matter subhalos, while closer to home, spectacular ALMA images of the solar chromosphere have provided essential data for studying the outer layers of the Sun.

Technologically, emerging trends from the past decade have enabled faster, low-cost prototyping and development of new facilities. At centimeter wavelengths, new developments have largely followed the trajectory of commercial products and are drawn from four categories: low-cost, low-noise amplifiers at increasingly high frequencies, and low-cost, high-bandwidth digital samplers (both driven by telecommunications needs); continued expansion of computing and especially highly parallelizable graphics processing units (driven by data science/gaming needs); and high-bandwidth network switches (driven by telecommunications and high-performance computing needs). Taken together, these devices make possible inexpensive, sensitive, wide-bandwidth radio receivers, correlators, and complex real-time data processing, such that instruments can be rapidly prototyped using commercial off-the-shelf components and software-defined radio tools. At (sub)millimeter wavelengths, the commercialization of cryogenics and the development of large arrays of superconducting detectors have dramatically increased the numbers of detectors that can be deployed on wide-field telescopes, allowing sky-statistics-limited (rather than detector-limited) analyses to be conducted for the first time.

Programmatically, the most dramatic development of the past decade has been the emergence of ALMA as a facility that engages the full (in terms of both wavelength and geography) astronomical community.

High demand for observing time by a user base much broader than the traditional RMS community, a large and growing stream of impressive, published results, and a well-defined pathway to future upgrades using dedicated development funding have set ALMA apart from previous RMS observatories. Two key factors in ALMA’s success are its superb imaging performance and its investment in pipeline development and provision of high-quality data products (through its archive) to users who are non-experts and/or have limited computing capacity at their home institutions. Both factors will also be relevant to the success of future RMS observatories.

M.2.2 Looking Forward to the Next Decade

In looking ahead to a decade (2022–2032) that will include the centenary of Karl Jansky’s pioneering discovery, the RMS panel relied on inputs from a variety of sources. Foremost among these was an impressive suite of white papers—innovative, ambitious, and wide-ranging—that laid out projects and priorities for the next 10 years and beyond. In reviewing these white papers and engaging with the teams who submitted them, the panel was guided by the high-priority science questions and discovery areas identified by the six Astro2020 science panels. In most cases, the RMS panel concurred with the science panels’ own assessments of where and how RMS observational capabilities could help address these questions; in a few cases, opportunities for impact beyond those identified by the science panels were factored into the RMS panel’s analysis. In line with the additional guidance provided to the Astro2020 survey, the RMS panel also considered the findings and recommendations of the National Academies reports Solar and Space Physics: A Science for a Technological Society (2012),²Exoplanet Science Strategy (2018),³ and An Astrobiology Strategy for the Search for Life in the Universe (2019).⁴

To characterize the ability of existing and proposed RMS facilities to address high-priority science questions and discovery areas (in concert with multi-wavelength and multi-messenger observations, theoretical work, and laboratory investigations), the RMS panel developed a scoring rubric with three categories. These distinguish areas where (1) a facility would make a contribution in addressing a science question (or any of its sub-questions) that would be irreplaceable and unique relative to other facilities with U.S. community access; (2) a facility would make a very significant contribution in addressing a science question but would not be sufficient to address that question by itself (e.g., in the absence of observations at other wavelengths); and (3) a facility would have an impact in addressing a science question, but would be one of several facilities playing supporting roles. To assess the risks and costs of large projects (i.e., those requiring >$70 million of federal funding), the RMS panel considered presentations from and extensive documentation provided by the respective project teams, while also making use of independent analyses by the Aerospace Corporation in the context of a Technical, Risk, and Cost Evaluation (TRACE) process (see Appendix O). The panel adopted a “hybrid” approach for these assessments, producing syntheses of risk registers and schedule and cost estimates from the project teams and from the TRACE analyses that are informed by panel expertise and represent the panel’s best judgments.

As a result of these deliberations, the RMS panel arrived at a set of four priorities for new or enhanced federal investment over the next decade. In order from highest to lowest construction cost per project, these investments are:

___________________

² National Research Council, 2013, Solar and Space Physics: A Science for a Technological Society, The National Academies Press, Washington, DC.

³ National Academies of Sciences, Engineering, and Medicine, 2018, Exoplanet Science Strategy, The National Academies Press, Washington, DC.

⁴ National Academies of Sciences, Engineering, and Medicine, 2019, An Astrobiology Strategy for the Search for Life in the Universe, The National Academies Press, Washington, DC.

• Design, construction, and early operation of a large new observatory, the Next Generation Very Large Array (ngVLA).
• Design, construction, and early operation of a large new “stage 4” experiment to study the CMB, CMB-S4.
• Significant funding to support mid-scale projects with costs of up to (at least) $70 million—where exciting opportunities exist in areas that include (but are not limited to) broadband, high-cadence, spectropolarimetric imaging of the Sun; high-resolution imaging of jets driven by supermassive black holes; surveying the static and time-variable radio sky; and mapping the evolution of neutral atomic hydrogen (HI) gas in the very early universe.
• Ongoing support for three key capabilities: the long-term timing of pulsars (which will require continued support for operation of the Arecibo telescope and the GBT), the development of new instrumentation (including software), and the mitigation of RFI.

The scientific motivations for these investments are discussed in detail in the relevant sections below, as are risk, cost, and programmatic issues for the ngVLA and CMB-S4. Table M.1 provides a concise visual representation of the RMS panel’s scoring of all relevant existing and proposed facilities against the Astro2020 high-priority science questions, with darker shades indicating areas where facilities can make more substantial contributions. The report concludes with suggestions of guiding principles for achieving balance (existing versus future facilities, large versus medium versus small cost scales, astronomers versus other stakeholders) within the RMS portfolio.

TABLE M.1 High-Priority Science Questions Versus RMS Facilities

NOTES: Top row indicates future large facilities (dark yellow), future mid-scale opportunities (MSOs, light yellow), existing facilities that would continue to operate (blue), and existing facilities that would be replaced by the ngVLA (pink). Columns MSO1 through MSO4 refer to opportunities in the areas of broadband solar imaging, high-resolution imaging of jets, surveys of the static and time-variable radio sky, and HI in the early universe, respectively. Rows list science questions and discovery areas, which are identified by panel acronym (ISM, EAS, SSSP, COEP, GAL, COS) and number (1, 2, 3, 4, D) elsewhere in the report. Each cell is shaded to indicate the role of the facility in addressing the topic. Dark green indicates that the facility is irreplaceable (at any wavelength) for addressing one or more sub-questions within a topic, and unique relative to other facilities with U.S. community access; medium green indicates that the facility is essential, but not sufficient to address a topic by itself; light green indicates that the facility is one of many with supporting roles in addressing a topic.

M.3 A LARGE NEW OBSERVATORY: THE ngVLA

M.3.1 Introduction

Since its completion in 1980, the VLA has been an outstandingly versatile and productive facility for advancing knowledge about the universe at centimeter wavelengths. The VLA identified the first Milky Way “microquasar” driving jets of energetic particles at close to the speed of light, determined the location of the black hole at the center of the Milky Way, discovered the first complete “Einstein ring” produced by gravitational lensing of a distant galaxy, and detected the first radio-wavelength counterpart to an explosive gamma-ray burst (GRB) event. In 2012, completion of the Expanded VLA (EVLA) upgrade yielded a more capable facility, with more sensitive receivers covering a wider range of frequencies and a correlator able to process wider bandwidths. Even after these improvements, however, the VLA has been hampered by the surface accuracies of its antennas, which limit its performance at the higher frequencies where emission from thermal processes is strongest, and by the number and allowable configurations of those antennas, which limit the quality of the images it can produce. Similarly, since coming online in 1993, the VLBA has blazed new trails in measuring the Hubble constant (via observations of water megamasers) and revealing the structure of the Milky Way. It too has been upgraded in bandwidth and frequency coverage, but its limited sensitivity (particularly at higher frequencies) constrains users’ ability to take full advantage of its superb angular resolution. As reflected in the Astro2020 science panel reports (see also Table M.1), a wealth of discovery opportunities would be within the grasp of a centimeter-wavelength successor to the VLA and VLBA that offered an order of magnitude improvement in sensitivity, and the ability to image sources on scales of arcminutes to fractions of a milliarcsecond (as appropriate for their surface brightness on those scales) across two decades in frequency. The RMS panel supports the funding of such an ngVLA in order to realize this scientific potential.

M.3.2 Science Case

The ngVLA design concept (see below) has been optimized through extensive community engagement to deliver on five key science goals (McKinnon et al., 2019⁵) that are broadly aligned with the Astro2020 high-priority science questions. This section highlights areas in which the ngVLA’s capabilities would allow it to make extraordinary contributions to addressing those questions, grouped by the RMS panel according to three broad science themes.

M.3.2.1 Stars and Planetary Systems

The births, lives, and deaths of stars, and the formation and evolution of planets in orbit around them, represent an important focus for RMS observations. Understanding the origin and prevalence of high-density structures within larger molecular clouds, and the role these structures play in the formation of stellar nurseries and the birth of stars, is critical for elucidating the first stages of the stellar life cycle. Young stellar systems can be explored for evidence of young planets to provide insights on the environments in which planets form and the speed with which they do so. Astronomers are now on the verge of being able to understand the architectures of planetary systems, the means by which planets migrate to different locations within these systems, and the origin, evolution, and prevalence of habitable environments. Understanding stellar activity at all phases of a star’s life is important for predicting potential impacts on planets in its

___________________

⁵ M. McKinnon, A. Beasley, E. Murphy, R. Selina, R. Farnsworth, and A. Walter, 2019, “ngVLA: The Next Generation Very Large Array,” white paper submitted to Astro2020: Decadal Survey on Astronomy and Astrophysics, https://baas.aas.org/pub/2020n7i081/release/1.

habitable zone, while near the end of its life, the process of mass loss and its connection to stellar death become increasingly relevant. The following paragraphs highlight ngVLA capabilities that would address high-priority questions identified by the ISM, EAS, and SSSP science panels.

The ngVLA’s combination of high spatial resolution and centimeter-wavelength continuum sensitivity would resolve protoplanetary disks on scales more than 20 times finer than ALMA, potentially capturing images of planet formation in action as young planets clear gaps in the disk. The inner regions of these disks, within a few astronomical units of the central star, are often too opaque to be studied at the shorter ALMA wavelengths. The ngVLA would image circumstellar disks in hundreds of protoplanetary systems with sufficient resolution (~5 milliarcseconds, corresponding to, for example, 0.6–0.9 astronomical units (AU) for the Taurus molecular cloud) to measure inner-disk surface density perturbations caused by young, forming super-Earths. These disk substructures would offer important insights on the process of planet formation (ISM-4). The short orbital periods of planets close to host stars could potentially be tracked through very sensitive ngVLA imaging of disk structures on time scales as short as a few weeks. High-resolution centimeter and millimeter studies of large samples of disks would help develop a census of planetary system architectures (e.g., distribution, mass, orbital radii) in mature systems to compare to those in newly forming protoplanetary disks (EAS-1), thereby helping researchers to understand the diversity of planetary systems, how our own solar system formed, and how unique our solar system is (EAS-2).

The sensitivity, resolution, and imaging fidelity of the ngVLA would be used to map the physical conditions and gas motions within star-forming cores. Centimeter wavelengths are particularly critical for tracing high-mass star formation in deeply embedded environments, as well as in the densest portions of infalling low-mass stellar cores, where sub-arcsecond resolution spectral studies would help constrain protostellar masses and trace collapse motions (ISM-3). Sensitive spectral line observations across centimeter (e.g., NH₃, deuterated molecules) and millimeter (e.g., N₂H⁺, CO isotopologues) wavelengths at high spatial (<0.1 pc) and velocity (<0.1 km/s) resolution would provide a census of dense gas as a function of environment within our own Milky Way galaxy. The sensitivity of the ngVLA across the centimeter and millimeter bands would further enable sub-arcsecond spectroscopic studies of large samples of galaxies in multiple species (HCN, HCO⁺, CO isotopologues, and different excitation lines), to trace the efficiency of cloud collapse into forming stars (ISM-2). ngVLA mapping of the neutral gas in nearby galaxies on the scale of individual star-forming clouds would make it possible to trace gas flows through the crucial atomic to molecular phase transition within the interstellar medium, as gas moves from being potential fuel for star formation toward the brink of actually forming stars (ISM-1).

Over half of the dust and heavy elements in the interstellar medium originate in the winds and outflows of dying stars known as red giants. One of the greatest challenges in understanding the physics, geometries, and time scales of these winds is that the atmospheres of evolved stars are governed by the interplay of complex physical processes, including pulsations, shocks, convection, magnetic fields, and the formation of dust and molecules. Studying these phenomena demands exquisite spatial resolution (corresponding to a small fraction of a stellar radius), coupled with the ability to monitor time-variable behaviors. The wavelength range covered by the ngVLA would probe the regions of red giant atmospheres beyond ~2 optical radii (in both spectral lines and continuum) where stellar winds are launched and accelerated. The ngVLA’s collecting area on long (up to continental) baselines would provide the combined sensitivity and ultra-high resolution (as fine as ~70 μas) needed to probe the spatially resolved atmospheric dynamics, magnetic fields, brightness temperatures, and surface features (e.g., spots and convective cells) of hundreds of red giants and follow their evolution over time (SSSP-3).

The ngVLA would have the power to resolve over 10,000 stars within the Milky Way, including the radio surfaces and extended atmospheres of hundreds of main sequence stars. Observables include both the thermal emission from their photospheres and chromospheres and coherent nonthermal and incoher-

ent gyrosynchroton emission resulting from magnetic activity and stellar coronae. The ngVLA would be able to resolve the crucial zone where the physical processes driving stellar activity manifest themselves, and its ultra-wide bandwidths would enable the dynamic spectro-imaging and spectropolarimetry necessary to study the evolution of flares, coronal mass ejections, and other active and magnetically driven phenomena (SSSP-1), including those in active binaries (SSSP-2). Such studies of stellar activity have taken on a heightened importance because of their relevance to space weather in extrasolar planetary systems, which may impact the development of life (SSSP-4). Measurements with the ngVLA’s continental baselines would also be able to trace binary orbits (including those of ultracool dwarfs), thereby enabling direct mass measurements (SSSP-1).

M.3.2.2 Black Holes and Galaxies

Supermassive black holes (SMBHs) are both physically and phenomenologically central to their host galaxies. In their vicinity, gravitational potential energy of infalling material and spin energy are converted to copious electromagnetic radiation and powerful jets that extend hundreds to thousands of light-years from their bases. The resulting radiative and mechanical feedback can have long-lasting impacts, including regulation of star formation in the host galaxies, although the transfer of energy from the base of a jet to the surrounding medium is still poorly understood. From a population perspective, detailed understanding of the role that black holes play in star formation and the evolution of their host galaxies will require a census of black hole growth through merging and accretion over cosmic time. At lower masses, spectacular gravitational wave detections have revealed the mergers of individual stellar-mass black holes forming more massive black holes, yet the extent of a population of intermediate-mass black holes, between the extremes of SMBHs and stellar-mass black holes, remains uncertain, and questions remain concerning the masses and spins of binary black holes prior to merger and how those properties map to stellar progenitors. This subsection highlights how the ngVLA would address key open questions on black holes and galaxies as identified by the COEP and GAL science panels.

The ground-breaking Event Horizon Telescope (EHT) observation of the base of the relativistic jet in M87 (see Figure M.1c) would be complemented by the ngVLA’s exceptional ability to trace details of the structure and acceleration of relativistic particles along the full lengths of that and many other jets, from scales of a few parsecs to hundreds of kiloparsecs (well beyond the 0.01 pc region probed by EHT). The continental (~9,000 km) baselines of the ngVLA, together with its full polarization capability, would provide the sub-milliarcsecond imaging needed from centimeter to millimeter wavelengths to trace details of the jet inclination, lateral structure, magnetic field strength, and variation of the Lorentz factor away from the launch point region in many black hole systems. These studies would reveal the composition of the jets, how particles are accelerated, and how the jet parameters vary with distance from where the jets are launched (COEP-3).

Beyond the study of relativistic jet properties, the ngVLA’s resolution, sensitivity, and imaging fidelity across the centimeter to few-millimeter band would be critical in searches for the elusive accretion signatures of intermediate-mass black holes and in efforts to develop a census of binary black holes to determine the role of mergers in the formation of supermassive black holes (COEP-4). Sensitive, high-resolution ngVLA spectral line studies of molecular (low-energy [J] CO) emission would reveal hidden details of the interaction of relativistic jets with their surrounding interstellar media (GAL-3).

Galaxies themselves often contain vast reservoirs of fuel for star formation, and these reservoirs are strongly influenced by feedback from stellar winds and supermassive black holes. Studies of high-redshift systems reveal that molecular gas broadly traces the cosmic star formation rate history of galaxies. The ngVLA would enable detailed measurements of the masses and kinematics of molecular gas clumps on sub-kiloparsec scales in typical (Milky Way–like) galaxies out to beyond z ~2, when the universe was

one-quarter of its current age. These observations would be coupled with ngVLA studies of small-scale feedback, which is known to regulate accretion and reduce star formation efficiency (GAL-2). On larger scales, individual galaxies are embedded within a circumgalactic and intergalactic medium that must be accounted for in our understanding of the star formation fuel reservoirs of individual systems. The ngVLA would enable resolved spectral study of this surrounding material on scales of kiloparsecs down to hundreds of parsecs (GAL-D). ngVLA observations of the dense environment near the center of our own Milky Way, including the Central Molecular Zone and the Circumnuclear Ring, could reveal details of the energetics, motions, and physical characteristics of gas that serves as a template for understanding distant galaxies that cannot be observed at such high spatial resolution (GAL-4).

M.3.2.3 Transient Sources and the Explosive Universe

The study of transients cuts across both astrophysics and cosmology. These brief, energetic events can trace stellar deaths, which drive the chemical enrichment of their surroundings and lead to the formation of neutron stars and black holes. Signatures of the binary neutron star merger GW170817 were detected across the electromagnetic spectrum after the initial gravitational wave signature of coalescence, beginning with the burst of gamma rays 2 seconds later, followed by an optical counterpart within 11 hours—and a VLA radio detection of an emerging relativistic jet 16 days after the detection of gravitational waves. Such observations have opened a new era of multi-messenger astronomy, where studies of individual events are combined across the electromagnetic spectrum to build a detailed picture of the engine that powers the diversity of transients. Transients can be used as “standard candles” to trace the acceleration of the universe, or as probes of the “missing baryon” content of the universe as recently undertaken for a sample of well-localized FRBs. The COEP and COS science panels have identified a number of questions, discussed below, that require the sensitivity, resolution, and imaging fidelity of a new centimeter/millimeter observatory such as the ngVLA.

With its microjansky flux sensitivity (owing to improved receivers and increased collecting area) and the sub-milliarcsecond resolution provided by its continental baselines, the ngVLA would be ideally suited to characterize the energy sources driving explosive transient events. It would map the energy distribution of the explosive ejecta driven outward by the ignition of the transients, search for newly launched relativistic jets, and constrain the development of pulsar wind nebula-like emission. These observations would transform our understanding of explosive phenomena in the universe (COEP-2).

Monitoring of compact binary mergers using the high resolution and excellent imaging fidelity of the ngVLA would move well beyond the excellent initial studies of GW170817 with the VLA and the High Sensitivity Array (VLBA supplemented by the VLA and GBT), which discovered and tracked the radio afterglow of an off-axis relativistic jet driven by the merger. ngVLA monitoring of these systems would allow early detection of newly formed relativistic jets and study of jet evolution when present. As part of a larger multi-messenger study, these new observations would permit detailed mapping of the initial merger conditions to the energetic impact on the local environment (COEP-D). Increased sensitivity of the gravitational wave network over the next decade will yield large samples of events that would require sensitive, high-resolution radio follow-up with an instrument such as the ngVLA, which would search for and image emission from electromagnetic counterparts. These observations would represent a critical step in building a large sample of standard sirens to independently probe the cosmic distance scale (COS-4).

M.3.3 Design Concept

At the broadest level, the vision for the ngVLA entails an order of magnitude increase in capabilities over those of existing facilities. Both the ngVLA project’s own key science goals and the high-priority questions

identified by the Astro2020 science panels demand coverage of broad and continuous frequency ranges between 1.2 and 116 GHz, velocity resolution as fine as 100 m/s, sub-milliarcsecond angular resolution, and high-fidelity imaging capabilities on scales from milliarcseconds to arcminutes. The technical realization that satisfies these specifications consists of 244 reflector antennas of 18 m diameter and 19 reflector antennas of 6 m diameter, all at fixed locations. The Main Array (MA) and Short Baseline Array (SBA) would have 214 × 18 m and 19 × 6 m antennas, respectively, centered on the current VLA site in New Mexico but distributed on baselines ranging from meters to ~1,000 km across the southwestern United States and into Mexico. This configuration would allow for the sampling of a broad range of spatial scales (from arcminutes to milliarcseconds). A Long Baseline Array (LBA) of 30 × 18 m antennas would be located in 10 clusters (mostly at existing VLBA sites), providing continental-scale baselines and sub-milliarcsecond resolution. All antennas would be connected by optical fiber to a single flexible signal-processing center, allowing for real-time correlation of all antennas simultaneously and operation in subarrays. Each antenna would feed a suite of cryogenically cooled receivers allowing operation from 1.2 to 116 GHz (except for the 50–70 GHz range, where the atmosphere is opaque), which would provide access to the HI and CO(1–0) emission lines at z = 0, and with spectral resolution better than ~0.1 km/s. An ambitious software and archive effort is projected to allow science-ready data products to be generated promptly and shared with users via server-side visualization and analysis platforms, thereby enhancing prospects for archival research, and lowering barriers to astronomers who may not be interferometry experts or may not have substantial computing resources at their home institutions.

Compared to other large existing and planned arrays—specifically, the Square Kilometre Array (SKA), in which the United States is not a partner, and ALMA—the ngVLA would provide unique capabilities. These include continuous frequency coverage from 1.2 to 50 GHz and from 70 to 115 GHz, unique coverage of a key frequency range (15–35 GHz, important for studying terrestrial planet formation, water megamasers associated with z <0.5 galactic nuclei, and cold molecular gas in z >2.3 galaxies), superior point-source sensitivity at all common frequencies compared to the SKA (1.2–15 GHz) and ALMA (35–50 and 70–116 GHz), and access to northern hemisphere sources. In its science reach, the ngVLA would deliver substantial quantitative improvements in the ability to study nonthermal phenomena, a qualitatively new ability to explore the thermal universe, and a powerful complement to ALMA’s higher frequency capabilities. The ngVLA would realize this science potential by allocating the bulk of its observing time (like ALMA) in response to principal investigator (PI) proposals, rather than (like the SKA) to large, predefined surveys.

M.3.4 Cost, Schedule, and Risks

The ngVLA project team has prepared a detailed project design, plan, schedule, and cost model. Further design and development work is proposed to occupy the next few years, followed by a decade-long construction phase starting in 2025 and a full (steady-state) operations phase running from 2035 through 2054. Costs (including contingency) are estimated by the project team as ≈$0.1 billion for design and development and ≈$2.4 billion for construction in 2020 U.S. dollars, translating to ≈$0.1 billion and ≈$3.2 billion in then-year dollars. The TRACE analysis estimates design and construction costs summing to ≈$3.2 billion in 2020 dollars, translating to $4.2 billion in then-year dollars. The RMS panel has arrived at a “hybrid” estimate for the total construction cost that is roughly $130 million (in 2020 dollars) lower than the TRACE value (rounding to ≈$3.1 billion in 2020 dollars and ≈$4.2 billion in then-year dollars) but concurs with the TRACE adjustments for (1) a project management and systems engineering “overhead” higher than that for the EVLA upgrade, as appropriate for the ngVLA’s larger geographical extent and greater need for international coordination; (2) schedule threats related to an uncertain time scale for antenna prototyping and a high antenna delivery rate (assumed to be three per month in the

steady state) to be sustained over many months; and (3) a higher level of overall contingency (reserve). With regard to (2), the panel notes that the vast majority of the ngVLA’s science potential would still be realized if longer integration times were needed to compensate for a modest reduction in the number of antennas—although preserving the concept’s excellent ultraviolet (UV) coverage and therefore imaging performance would remain paramount goals. Importantly, all of the above numbers refer to total design and construction costs, of which NSF would contribute only 75 percent; given information available as of mid-2020, the panel views prospects for international partner contributions at the desired 25 percent level as excellent. Annual operations costs for the ngVLA are projected to rise from ≈$147 million in 2035 to ≈$244 million in 2054 in then-year dollars; again, current expectation is that only 75 percent of these costs would need to be borne by NSF. Adopting the RMS panel’s construction cost estimate and a mean annual operations cost of ≈$100 million in 2020 dollars, the operation-to-construction cost ratio would be ~3 percent, which is at the low end of the envelope filled by previous large projects. The panel appreciates the project team’s explicit inclusion of ngVLA decommissioning (at the level of ≈$0.2 billion in 2020 dollars) in its calculation of total life cycle costs.

Owing to its planned use of mature technology in most areas, the ngVLA would be a project with low technical risk. Technical specifications for the antennas are not overly stringent compared to the current state of the art, or indeed to other observatories that are already operational, although the cost implications of these specifications will become clearer once a satisfactory 18 m antenna prototype exists. Risks related to correlator and receiver development are also low and well understood by the ngVLA project team; risks related to RFI are recognized, and the team is working to develop appropriate mitigation strategies. Last, the RMS panel has considered the risk to the ngVLA’s scientific productivity that could arise if its very large image sizes—driven by its large field of view and long baselines—overtax the typical home-institution computing resources of its users. The ngVLA project team understands the scale of this challenge and has plans to address it—for example, via a user-friendly archive coupled with server-side visualization and analysis platforms, although this type of functionality remains an active area of development.

M.3.5 Additional Programmatic Guidance

The RMS panel views the ngVLA as an exciting concept for a flexible, powerful, PI-driven observatory that would address a wide range of Astro2020 high-priority science questions. In support of the project’s long-term success, the RMS panel offers three suggestions for its implementation. First, the panel views international participation as essential to the success of the ngVLA project, given the value of sharing technical expertise as well as costs. It would therefore be important for NSF to be proactive in enabling full participation by international partners through both the construction and operations phases of the project. Second, the panel views the growth of a community of future ngVLA users as vital to the ultimate success of the project. In previous decades, growth of the community of future ALMA users was supported by the tight integration of research and training at (sub)millimeter wavelength facilities funded by NSF’s University Radio Observatories (URO) program. With the demise of the URO program, the existence of a broad community of ngVLA users tomorrow would require the growth of a broad community of VLA users today. The RMS panel therefore suggests that concrete progress in making the VLA accessible to non-expert users (e.g., via observatory-specified calibration strategies, automatically defined schedules, standard correlator modes, pipeline-reduced data products, and server-side visualization and analysis tools),⁶ which can inform detailed design and costing of user interfaces for the ngVLA, factor positively in agency decisions on the

___________________

⁶ J. Kern, B. Glendenning, and J. Robnett, 2019, “The Science Ready Data Products Revolution at the NRAO,” white paper submitted to Astro2020: Decadal Survey on Astronomy and Astrophysics, https://ui.adsabs.harvard.edu/abs/2019BAAS...51g..60K.

start of ngVLA funding. Third, the panel endorses the view that the ngVLA would be a replacement for both the VLA and the VLBA. Full details of when the VLA and VLBA would be decommissioned relative to the progress of ngVLA construction remain to be determined, but as a dedicated VLBI array with continent-scale (~10⁴ km) baselines, the VLBA is globally unique and will remain crucial for astrometric and other science that demands the monitoring of time-varying phenomena with ultra-high angular resolution. The panel therefore suggests that the VLBA remain operational unless and until its capabilities (ideally, upgraded in the near term by bandwidth increases) are supplied by ngVLA/LBA stations. Continuity here would ensure the existence of a community of future LBA users, and would maintain the VLBI capability that is needed for a number of ongoing long-term (e.g., astrometric) observing programs.

M.4 A LARGE NEW EXPERIMENT: CMB-S4

M.4.1 Introduction

The past decade of measurements of the CMB have yielded precision tests of the Lambda cold dark matter (ΛCDM) paradigm and increasing precision on the parameters that describe the universe. In the coming decade, sensitive observations of the CMB have the potential to resolve central questions in cosmology, fundamental physics, and particle physics, while also providing new astrophysical insights. The RMS panel supports funding of the CMB-S4 experiment, which is designed to push CMB measurements across critical sensitivity and measurement thresholds to understand the origins of inflation, search for hidden fundamental particles, map out the distribution of mass and hot gas throughout the universe, and explore time-variable and static millimeter-wave sources. CMB-S4 would apply existing technologies on an unprecedented scale, combining the major ground-based CMB experimental groups and U.S. national laboratories to deliver an instrument matched to scientific need. CMB-S4 is envisioned as a joint NSF and Department of Energy (DOE) project, has been endorsed by the high-energy physics community in the 2014 Particle Physics Project Prioritization Panel (P5) report and by the 2015 National Academies report A Strategic Vision for NSF Investments in Antarctic and Southern Ocean Research,⁷ and has achieved DOE “Critical Decision 0,” which confirms the need for investment in this scientific area.

M.4.2 Science Case

The CMB-S4 experiment has been designed by the U.S. and international cosmology communities to address four science themes. The following subsections connect its capabilities to the science questions identified by the Astro2020 science panels.

M.4.2.1 Cosmology

According to the predominant modern theory of cosmology, in the very first moments after the Big Bang, the universe underwent an exponential expansion known as inflation. Inflation explains key cosmological mysteries, such as the origin of the incredible uniformity of the universe on very large scales and the measured flatness of space, which would otherwise require a precise tuning of the cosmic energy density. However, the physics that drove inflation has not yet been identified, and understanding what set the Big Bang in motion is now one of the primary cosmology science questions for the coming decade (COS-1). There should be signatures of the inflationary epoch encoded in the CMB that will reveal the origins of this

___________________

⁷ National Academies of Sciences, Engineering, and Medicine, 2015, A Strategic Vision for NSF Investments in Antarctic and Southern Ocean Research, The National Academies Press, Washington, DC.

expansion and provide novel information about Grand Unified Theories and quantum gravity, and CMB-S4 is designed to find these signatures.

A clear imprint of the inflation era is a background of gravitational waves echoing through the universe. These inflationary gravitational waves (IGW) introduce vortical patterns (known as “B-mode,” tensor, or curl components) in the observed CMB vector polarization field, which is intrinsically curl-free (and referred to as “E-mode,” scalar, or divergence polarization patterns). The ratio of the tensor to scalar polarization modes in the CMB, known as r, encodes the energy scale at which the inflationary expansion occurred. (For some classes of models that generate additional B-modes during inflation, the picture is not so simple, but these have particular signatures that allow them to be distinguished.) Determining the value of r, therefore, provides a unique window into the earliest moments of the universe—temperature/energy scales that are forever beyond collider experiments. Current data indicate that r <0.06, but also prefer families of models that predict r >0.001. Achieving a measurement of r will clarify the underlying physics driving inflation and provide evidence for the quantization of gravity, while limiting r to less than 0.001 at 95 percent confidence will rule out the leading models of inflation. CMB-S4 is designed around achieving this challenging experimental target, matching the target set by the COS science panel (COS-1). It requires a dramatic increase in the number of CMB bolometers in operation, a wide range of independent frequency bands to separate out contaminating foregrounds, and a combination of large and small angular scales to detect the large-scale IGW B-modes and remove contaminating B-modes from lower-redshift lensing of the CMB. The CMB-S4 design is flexible enough to reoptimize its experimental approach during its 7-year lifetime to refine and improve its constraints if a detection of r is made. No CMB experiment less ambitious than CMB-S4 can achieve the needed sensitivity.

The standard cosmological model is a highly successful description of the evolution of the universe, from the instants after the Big Bang to the present. Within this theory, there remain many key details to understand, many of which are tied to fundamental particle physics. The COS science panel has identified the properties of dark matter and the dark sector (COS-2) as a potential breakthrough area for the decade. In particular, the model of a single dark matter particle has given way to a diverse field of potential “dark sector” contributions to the energy density of the universe, expanded sets of particles and fields that are only weakly coupled to known components of the Standard Model of particle physics, predicted as part of extensions to the Standard Model. The CMB provides a unique opportunity to search for the existence of relativistic particles (“dark radiation”) that contribute to the cosmic energy density but cannot be sensed in laboratory experiments. For example, light relics imprint measurable perturbations in the acoustic oscillations of the primary CMB temperature and polarization power spectra, while ultralight axions affect the formation of structure on small angular scales, which are detectable in gravitational-lensing induced secondary perturbations to the polarization power spectrum at small angular scales. CMB measurements already demonstrate the reality of the cosmic neutrino background predicted by Big Bang cosmology despite the absence of laboratory detections. To accomplish these goals, CMB-S4 is designed with the angular resolution, sensitivity, and sky coverage needed to precisely measure the perturbations caused by relativistic particles that decouple from the hot early universe within the first nanosecond (COS-2b), before the quantum chromodynamics phase transition when quarks bind to form hadrons. Such a measurement is not within reach of current experiments, or planned upgrades before CMB-S4.

Understanding the growth of cosmic structure (COS-3) is a third key science question for the decade and is another interface between fundamental particle physics and cosmology. The universe is suffused with neutrinos, and measurements of neutrino oscillations have demonstrated that these particles have non-zero mass, although these oscillations only determine the differences between the squared masses of the three primary neutrino generations (electron, mu, and tau). The total neutrino mass, and thus the contribution of neutrinos to the energy density of the universe and their influence on structure formation, remains unknown.

CMB-S4 is designed to achieve a critical threshold (COS-3b) in the measurement of the total neutrino mass—measuring the minimum possible value to 5σ precision. Answering this cosmology science question requires pushing CMB measurements to the limits imposed by cosmic variance, and can only be achieved with an experiment on CMB-S4’s scale. CMB-S4 would also make unique measurements of cosmic structure (COS-3a) by mapping the large-scale mass distribution through reconstruction of the CMB lensing potential—enabling fruitful comparisons with tracers of structure at other wavelengths—and by characterizing the motions of clusters within that structure via the kinetic Sunyaev-Zel’dovich (kSZ) effect.

M.4.2.2 Galaxies, Transients, and the Explosive Universe

While CMB-S4 is designed to deliver precise cosmological measurements, its capabilities open up other science areas that have the potential to engage broader swaths of the astronomical community. CMB-S4 would produce unprecedented maps of ~70 percent of the sky at wavelengths between 1 cm and 1 mm, sampling the full area at least every other day. The sensitivity of these maps would enable a variety of science, particularly the study of hot circumgalactic, intergalactic, and intracluster gas. The temporal sampling would open up this wavelength regime to systematic time-domain studies for the first time.

The scattering of CMB photons by hot electrons results in a characteristic CMB spectral distortion, known as the thermal SZ effect, with an amplitude proportional to the integrated pressure of the hot gas. While this technique has seen its greatest use for the detection and characterization of galaxy clusters, the sensitivity and sky coverage of CMB-S4 maps would make it possible to explore the ionized gas of the circumgalactic medium (GAL-D). Stacking analyses would measure the circumgalactic medium pressure profile to megaparsec radii and constrain the contributions of active galactic nucleus (AGN) and supernova feedback.

As each new wavelength regime (gamma ray, X ray, optical, infrared, centimeter) has been opened up to systematic time-domain surveys over the past two decades, the number and nature of transient sources have continued to surprise. The millimeter-wave regime probed by CMB experiments is largely unexplored, with only a single, limited experiment over the past decade. Yet the potential sources span a range of exciting possibilities (COEP-2d), ranging from high-redshift and/or orphaned GRB afterglows (peaking in the millimeter regime), to the mysterious fast/blue optical transients like AT2018cow, to tidal disruption events and AGN variability that may be linked to neutrino emission (COEP-D) and black hole accretion physics (COEP-4). Many of these sources peak quickly in the millimeter regime, especially those enshrouded in dust that may be invisible at other wavelengths. Although event rates are uncertain, CMB-S4 would probe all of these with a new combination of cadence and depth, opening up new avenues for follow-up at higher angular resolution with facilities like ALMA and (potentially) the ngVLA, and perhaps also identifying new solar system objects and galactic transients like stellar flares.

M.4.3 Design Concept

CMB-S4 is designed to take advantage of two well-established millimeter-wave observing sites to conduct two simultaneous surveys. A 7-year ultra-deep survey of 3 percent of sky would take advantage of continuous visibility and outstanding weather conditions at the South Pole. This effort would use 18 Small Aperture Telescopes (SATs), each of diameter 0.5 m, to observe over six bands from 30–270 GHz. In addition, a single Large Aperture Telescope (LAT) of diameter 6 m would be used for delensing purposes, operating at 20–270 GHz. In parallel, a 7-year deep/wide survey of 70 percent of the sky would take advantage of the superior sky coverage accessible from the Atacama Desert in Chile. This effort would use two of the same LATs observing at 30–270 GHz. All ~500,000 detectors required would be of the transition edge sensor type, with cryogenic multiplexing readouts.

M.4.4 Cost, Schedule, and Risks

The CMB-S4 project team has prepared a detailed project design, plan, schedule, and cost model, which are compatible with NSF and DOE protocols for management of large projects and have already been refined through several rounds of internal review. First light is proposed for 2026, with the end of construction in 2028 leading to a full (steady-state) operations phase that concludes in 2035. Rigorous costing by the team in accordance with DOE methodology implies costs (including contingency) of ≈$30 million for design and development and ≈$500 million for construction in 2020 U.S. dollars, summing to ≈$600 million in then-year dollars. The TRACE analysis estimates design, development, and construction costs summing to ≈$660 million in 2020 dollars, translating to $700 million in then-year dollars. The RMS panel has arrived at a “hybrid” estimate for design, development, and construction costs that is slightly higher than the project team’s (≈$560 million in 2020 dollars), concurring with the TRACE adjustments for (1) a higher assumed “overhead” for information technology, computing, and software during construction, and (2) schedule threats related to the timely fabrication of an unprecedented number of cryogenic detectors. With regard to (2), the panel notes that impacts on schedule and cost are substantially reduced by the 1 year of schedule contingency that is already built into the CMB-S4 project plan. In the construction phase, DOE:NSF cost sharing is expected to be in the ratio 7:5, implying that the panel’s estimated total cost for design, development, and construction would translate to costs of ≈$330 million in 2020 dollars (≈$370 million in then-year dollars) to DOE, and ≈$230 million in 2020 dollars (≈$260 million in then-year dollars) to NSF. These estimates make no assumptions about cost savings that might be possible if CMB experiments aligned with CMB-S4 were to make in-kind contributions of infrastructure.

A preliminary bottom-up estimate by the CMB-S4 team implies an annual operations cost of ≈$33 million in 2020 dollars (≈$55 million in then-year dollars) after averaging over the experiment’s nominal 7-year lifetime. In this phase, DOE:NSF cost sharing is tentatively expected to be in the ratio 1:1, with further sharing of the NSF portion among multiple divisions under consideration. Adopting the RMS panel’s construction cost estimate and a mean annual operations cost of $100 million in 2020 dollars, the operation-to-construction cost ratio would be ~6 percent. The CMB-S4 project plan notes that normal end-of-life decommissioning costs for South Pole and Chile infrastructure are anticipated.

Owing to significant heritage from previous generations of CMB experiments, including the ongoing third-generation Simons Observatory (SO) and South Pole Observatory (SPO), CMB-S4 would be a project with medium/low technical risk. The primary source of programmatic risk is the challenge of scaling up to a high production rate across multiple fabrication sites in order to deliver a large number of cryogenic detectors on a tight timeline. As noted above, a year of contingency in the project schedule already provides substantial mitigation on this front. Through a dedicated working group guided by external reviews, the CMB-S4 project team is exploring other risk reduction strategies—for example, enlisting more facilities beyond the planned three DOE labs in the detector fabrication effort. The RMS panel has concluded that the scaling challenge here is not insignificant, but that prospects for mitigation are good.

M.4.5 Additional Programmatic Guidance

The RMS panel views CMB-S4 as a powerful, cosmology-focused experiment that would address Astro2020 priority science questions at a level that no other concepts can. In support of the project’s long-term success, the RMS panel offers the following two suggestions for its implementation. First, the panel suggests that third-generation CMB experiments aligned with CMB-S4—specifically, the SPO and the “nomi-

nal” version of the SO—be high priorities for federal support.⁸ Besides training students and postdoctoral researchers, thereby empowering them to play vital future roles in CMB-S4, these experiments are poised to help retire technical risk for CMB-S4 and usefully inform its strategies for surveying the sky and removing foreground signals. Second, the panel views it as appropriate for an experiment at the cost scale of CMBS4 to be more “observatory-like” in seeking broad engagement with astronomers beyond the traditional CMB community, and ensuring that (for example) plans for data management and event alerts maximize opportunities for transient science to the extent possible without sacrificing the primary cosmology goals. The panel therefore suggests that an articulated plan for engaging the broader astronomical community be a precondition for the start of CMB-S4 funding.

M.5 SIGNIFICANT FUNDING TO SUPPORT MID-SCALE PROJECTS

Over the past decade, funding of projects at the mid-scale level (for NSF, currently defined as costing $2 million to $70 million, with awards made via competitive proposal calls) has become an important, agile, and cost-effective mechanism for enabling the construction and operation of world-class RMS facilities in a variety of science areas, while training the next generation of scientists and instrument builders. Funded mid-scale projects have included the EHT, which can image the environs of supermassive black holes in and beyond the Milky Way, and multiple experiments that explore the early universe by probing the CMB or the epoch of reionization (EoR), when the intergalactic medium transitioned from being mostly neutral to mostly ionized. Funded projects have also included new instruments or improvements for large single-dish telescopes operating in observatory mode, including a more accurate surface for the GBT, a new multi-pixel camera for Arecibo, and a new multi-band millimeter camera for the Large Millimeter Telescope (LMT). Most of these awards have been made through the Mid-Scale Instrumentation Program (MSIP), which was established by the Division of Astronomical Sciences in response to a recommendation of the New Worlds, New Horizons in Astronomy and Astrophysics⁹ (Astro2010) decadal survey. Very recently, two awards in support of RMS projects have been made through the broader Mid-scale Research Infrastructure-1 (MSRI-1) program, which was established in support of one of NSF’s “10 Big Ideas.”¹⁰ The substantial oversubscription of both MSIP and MSRI-1 funding is an indicator of high demand in this cost range and high quality of funded projects. That RMS projects in particular have competed so successfully in these programs reflects the wealth of scientific opportunities in this wavelength regime.

Based on the ambition and creativity of the Astro2020 white papers, the RMS panel is confident that funding for mid-scale projects will be as valuable and as impactful in the next decade as it has been in the past. In particular, based on the Astro2020 high-priority science questions, the RMS panel has identified four areas in which outstanding scientific opportunities exist for new mid-scale RMS facilities. These areas are discussed in detail below, with reference to the specific white papers that inspired them, although because each still requires navigating a complex path to a successfully competed MSRI-2 (up to $70 million) proposal, the panel is highlighting them as exciting opportunities rather than endorsing concepts exactly as presented. In three of the four areas, future investment would build on previous MSIP and/or MSRI-1 funding. The panel views the limited previous investment in the fourth area (solar broadband imaging) as a missed opportunity, given that earlier versions of the Frequency Agile Solar Radiotelescope (FASR) concept

___________________

⁸ The RMS panel considered whether intermediate sensitivities (superior to the third-generation CMB experiments’ and inferior to CMB-S4’s) would be worth pursuing as a separate decadal goal. However, facilities delivering such sensitivities would not by themselves address the Astro2020 COS panel questions and would find it difficult (in terms of schedule) both to build on lessons from third-generation experiments and to inform CMB-S4 strategy.

⁹ National Research Council, 2010, New Worlds, New Horizons in Astronomy and Astrophysics, The National Academies Press, Washington, DC.

¹⁰ National Science Foundation, “NSF’s 10 Big Ideas,” https://www.nsf.gov/news/special_reports/big_ideas/index.jsp.

were strongly endorsed by the Astronomy and Astrophysics decadal surveys in 2000 and 2010 and the Solar and Space Physics decadal surveys in 2002 and 2012, but only the subset of its capabilities represented by the Expanded Owens Valley Solar Array (EOVSA) have been implemented. It would be important to structure future mid-scale funding competitions so that research in areas like ground-based solar physics is not inherently disadvantaged by the fact that it is pursued across more than one NSF division.

In order of nearest to most distant observational target(s), a first key area of mid-scale opportunity is broadband, spectropolarimetric imaging of the Sun. To date, solar radio observations from EOVSA and the Murchison Widefield Array (MWA) have made significant advances using spectro-imaging observations of the Sun, revealing the evolving spatial and energy distributions of high-energy electrons in flares, mapping spatial and temporal changes of the coronal magnetic field, tracing the origin of coronal heating, and performing 3D mapping of the magnetic field in sunspots. However, each facility has limitations: because of its small number of antennas, EOVSA does not have the dynamic range needed for high-fidelity imaging of rapidly time-varying phenomena or the low-frequency coverage needed to study coronal plasma emission, while the MWA is not a dedicated solar array and cannot provide the observing time or infrastructure needed for a comprehensive view of the Sun. Bastian et al. (2019)¹¹ present a concept for FASR, a facility optimized (in terms of bandwidth, sampling rate, angular resolution, and imaging dynamic range) to study the extreme ranges of flux density and temporal variations of the Sun over a frequency range from 0.2 to 20 GHz. Such a dedicated solar facility would allow daily imaging of the dynamic solar atmosphere from the middle chromosphere through the solar corona with a cadence of several times per second, and would have the ability to image narrower frequency bands with time resolution as fine as 20 ms. To achieve this performance, two separate arrays of antennas would be spread over ~3 km footprints (not necessarily at the same site): a ~64 element array of 2 m antennas operating from 2 to 20 GHz, and a ~48 element array of 6 m antennas operating at 0.2–2 GHz. Such a facility would be spectacularly powerful for understanding the dynamic atmosphere of the Sun, solar activity (SSSP-1), and all key components of space weather (SSSP-4), sampling both thermal plasma and nonthermal particles and uniquely sensitive to solar magnetic fields (SSSP-3). It would also be an invaluable partner to solar space-based missions in the coming decade. The arrays would launch a new era of “4D” studies of the Sun through dynamic imaging spectroscopy with unprecedented spatial, temporal, and frequency resolution (SSSP-D) to probe the evolution of complex solar phenomena and the couplings between them. Such spectro-imaging of the Sun at radio wavelengths would reveal the extraordinarily complex range of phenomena that occur over various spatial and temporal scales within stellar atmospheres and directly probe how these processes affect the physics and dynamics of the solar atmosphere, including the temperature structure, the circulation of material, the driving of winds, and the ejection of plasma through coronal mass ejections (CMEs). These insights in turn would inform our understanding of the behavior of other stars (SSSP-3). While the basic concept remains well-aligned with Astro2020 priorities, significant changes in technology have occurred since FASR was first proposed. A logical first step toward a full MSRI-2 scale proposal would be a redesign of the original 2010 concept (via MSIP/MSRI-1 funding) that takes into account technological advances and cost-savings opportunities driven by commercial developments over the past decade.

A second key area of mid-scale opportunity is high-resolution imaging of jets driven by supermassive black holes in the centers of galaxies. The current state of the art in terms of resolution is provided by the EHT, an experiment that has regularly combined a number of telescopes around the world (including ALMA) for VLBI observations at a relatively short (1.3 mm) wavelength. Because angular resolution is proportional

___________________

¹¹ T. Bastian, H. Bain, R. Bradley, B. Chen, J. Dahlin, E. DeLuca, J. Drake et al., 2019, “Frequency Agile Solar Radiotelescope: A Next Generation Radio Telescope for Solar Astrophysics and Space Weather,” white paper submitted to Astro2020: Decadal Survey on Astronomy and Astrophysics, https://science.nrao.edu/science/astro2020/apc-white-papers/201-dea27a2fda-59160361b76946a46b3f82_FASR_APC_White_Paper.pdf.

to observing wavelength and inversely proportional to separations between telescopes, the EHT has been able to deliver an unprecedentedly sharp view of the center of the galaxy M87 (see Figure M.1c). Doeleman et al. (2019)¹² present a concept for an expansion of the EHT that would entail (1) building 10 new 10 m diameter telescopes at additional sites around the world, in order to improve image fidelity, and (2) quadrupling the recording bandwidth, in order to improve sensitivity and enable simultaneous polarimetric imaging at 1.3 mm and 0.87 mm. By virtue of its improved imaging performance and higher angular resolution, such a facility would provide essential insights on the question of how jets are formed and powered (COEP-3). M87, whose SMBH drives a powerful jet at 99 percent of the speed of light and has already been imaged by the EHT at an angular resolution comparable to its projected Schwarzschild radius, would be a uniquely promising target for a more capable facility. By making sensitive, multi-frequency, spatially and temporally resolved observations with higher dynamic range, it would be possible to shed light on the details of how magnetic fields extract rotational energy from the black hole and/or its surrounding accretion disk to drive M87’s jet. Larger samples (of tens to hundreds of systems) would enable studies of jet physics on scales farther from the Schwarzschild radius, and through the EHT’s ability to resolve binary black holes at high angular resolution would also help address the question of how supermassive black holes grow (COEP-4, GAL-3). A logical first step toward a full MSRI-2 scale proposal here would be completion of the design and prototyping work that has recently been funded through an MSRI-1 award to the EHT team.

A third key area of mid-scale opportunity is the surveys of the static and time-variable radio sky that would be enabled by an innovative new “radio camera” instrument. Technological advances in low-noise, room-temperature amplifiers and commercial computer power and networking have made it possible to conceive of large (of order ~1,000 element) arrays of radio dishes whose signals are combined to produce science-ready images in real time without deconvolution—that is, a true radio camera. An array operating in the GHz range with baselines extending out to 15 km would have angular resolution of a few arcseconds. Hallinan et al. (2019)¹³ present the Deep Synoptic Array 2000 (DSA-2000) concept, consisting of 2000 × 5 m steerable dishes covering the entire 0.7–2 GHz frequency range and designed from the ground up for survey science. Such a radio camera could efficiently survey the entire observable sky with multiple pointings over multiple epochs, produce full-Stokes maps with noise well below a μJy/beam, and reveal of order a billion radio sources. The resulting data sets would address a large and diverse subset of the Astro2020 high-priority science questions. An enormous catalog of FRBs, triggered and communicated in real-time, would enable exploration of the diversity of explosive phenomena across the electromagnetic spectrum (COEP-2). Time-domain searches would also contribute to searches for radio afterglows of compact object mergers detected by the Laser Interferometer Gravitational-Wave Observatory (LIGO) and Virgo (COEP-D), for CMEs from other stars (SSSP-4), and potentially for technosignatures as tracers of life on exoplanets (EAS-D). A radio survey camera with the large collecting area of the DSA-2000 concept could make a very significant contribution to pulsar timing efforts in support of gravitational wave detection (see below). Beyond time-domain science, a broadband radio survey of the entire sky to unprecedented depth would enable studies of how supermassive black holes form and grow in concert with their host galaxies (GAL-3), and searches for the signatures of dark matter annihilation (COS-2). A logical first step toward a full MSRI-2 scale proposal here would be the further development of a design (building on the DSA-2000 team’s existing prototype array, recent MSIP funding for a 110-element precursor, and ongoing characteriza-

___________________

¹² S. Doeleman, L. Blackburn, J. Dexter, J.L. Gomez, M.D. Johnson, D.C. Palumbo, J. Weintroub et al., 2019, “Studying Black Holes on Horizon Scales with VLBI Ground Arrays,” white paper submitted to Astro2020: Decadal Survey on Astronomy and Astrophysics, https://arxiv.org/abs/1909.01411.

¹³ G. Hallinan, V. Ravi, S. Weinraub, J. Kocz, Y. Huang, D.P. Woody, J. Lang et al., 2019, “The DSA-2000: A Radio Survey Camera,” white paper submitted to Astro2020: Decadal Survey on Astronomy and Astrophysics, https://arxiv.org/abs/1907.07648.

tion of possible sites in the western United States) and supporting partnerships that would be compatible with the upper limit on mid-scale funding by NSF.

A final key area of mid-scale opportunity, probing the largest cosmic scales, is mapping the evolution of HI in the early universe. Parsons et al. (2019)¹⁴ discuss the current status of this exciting field, in which a detection of the earliest phase of the transition from a mostly neutral to mostly ionized intergalactic medium has recently been reported by the Experiment to Detect the Global EoR Signature (EDGES). Several other current experiments, including the Hydrogen Epoch of Reionization Array (HERA), Phase II of the MWA (MWA-II), and the Large-Aperture Experiment to Detect the Dark Ages (LEDA) are in the process of obtaining data that could deliver the first HI power spectrum of the EoR and potentially confirm the EDGES result interferometrically. While this field is in its infancy, the potential impact of opening an entirely new window on cosmic evolution is enormous, and there is considerable value in having independent experiments with different designs (and therefore different observational systematics). Technical lessons and scientific results from these projects, supplemented by further prototyping of hardware, software, and analytical techniques (e.g., refined methods for in situ mitigation of potential systematic errors, new array elements, direct imaging Fast Fourier transform [FFT] correlators, and real-time calibration), would inform the design of one or more next-generation experiments by the end of the decade. Possible architectures could include an “EoR imager” focused on cross-correlations with multi-wavelength probes of structure at redshifts z <12, and a “Cosmic Dawn Array” focused on z >12 power spectrum measurements where non-HI probes are unavailable. Connections to high-priority science questions are clear: characterization of HI in the early universe would directly constrain the thermal history of the intergalactic medium and the topology of reionization (GAL-1), and open up the use of the pre-reionization “Dark Ages” as a cosmological probe (COS-D). Constraints on the 21 cm power spectrum would also aid understanding of the distribution of dark matter on small scales (COS-2). A logical first step toward a full MSRI-2 scale proposal here would be assimilation of the results and lessons from HERA and MWA-II en route to the design of one or more next-generation experiments. The RMS panel notes that insights on observational systematics gained from the current generation of EoR experiments—for example, related to beam characterization and coupling between neighboring antennas—would also inform the design of possible future HI intensity mapping experiments targeting lower redshifts, such as the Packed Ultra-wideband Mapping Array (PUMA) that was presented to Astro2020 in conceptual form.

The above set of mid-scale opportunities reflects not only the diversity of RMS science that can be supported at this level of investment but also the diversity of phases in which different projects may find themselves relative to the decadal survey cycle. By issuing calls for mid-scale proposals on a regular basis, NSF would accommodate projects that become funding-ready at different points in the decade.

M.6 CROSSCUTTING CAPABILITIES: PULSAR TIMING, INSTRUMENTATION DEVELOPMENT, AND RFI MITIGATION

M.6.1 Pulsar Timing, and Continuing Support for Arecibo and the Green Bank Telescope

Pulsars are highly magnetized, rapidly rotating neutron stars whose beams of emission represent some of the most stable clocks in the universe. By making exact measurements of the arrival times of pulses from individual pulsars and large networks of pulsars, astronomers can draw conclusions about the properties of the neutron stars and the spacetime through which the pulses travel on the way to Earth. To be suc-

___________________

¹⁴ A. Parsons, J.E. Aguirre, A.P. Beardsley, G. Bernardi, J.D. Bowman, P. Bull, C.L. Carilli et al., 2019, “A Roadmap for Astrophysics and Cosmology with High-Redshift 21 cm Intensity Mapping,” white paper submitted to Astro2020: Decadal Survey on Astronomy and Astrophysics, https://arxiv.org/abs/1907.06440.

cessful, pulsar timing programs must satisfy several important criteria: (1) they must involve observations at multiple frequencies, so that the effects of intervening interstellar gas on pulse arrival times can be corrected; (2) they must be long-term, to enable more precise measurements of secular changes in those arrival times (owing to the effects of precession, general relativity, or low-frequency gravitational waves); (3) they must be observed with an uninterrupted cadence, to prevent the loss of information on phasing and/or long-wavelength gravitational wave sensitivity; and (4) in the case of pulsar timing networks, they must be accompanied by pulsar search programs that can identify new objects for inclusion in those networks in order to improve sensitivity to gravitational waves. With the exception of (4), where large single-dish telescopes or very closely packed arrays are needed for efficient searches, these criteria can in principle be satisfied by many possible combinations of current (Arecibo, GBT, VLA) and potential future (ngVLA, mid-scale radio survey camera) RMS facilities.

From the Astro2020 science panel reports, it is clear that pulsar timing capabilities are critical for tackling a number of high-priority science questions: the mass and spin distributions for neutron stars and black holes (COEP-1); the growth of supermassive black holes (COEP-4, GAL-3); the synthesis of information from electromagnetic, particle, and gravitational wave signals (COEP-D); the properties of dark matter (COS-2; pulsar timing arrays can potentially detect anomalies owing to lensing by dark matter lumps); and the cosmological implications of gravitational waves (COS-4). From the RMS panel’s interaction with the North American Nanohertz Observatory for Gravitational Waves (NANOGrav) team, which is a global leader in the use of pulsars for gravitational wave detection, it is equally clear that large amounts of observing time with sensitive facilities will be critical for increasing the size of the timing sample from its current ~80 to 200 millisecond pulsars by the end of the decade. (A “minimum observing program” for NANOGrav would require ~1,200 hours a year of timing observations with broadband receivers on facilities that include at least one with Arecibo-scale collecting area, plus ~1,000 hours a year for pulsar searches; an “ideal observing program” would use up to 2.5× as many hours for timing at a higher cadence.) Given the uncertainties in whether and when the ngVLA or a mid-scale radio survey camera would become available to contribute to pulsar timing efforts, the desire to support continuing U.S. leadership in this area, the need to avoid any gaps in timing coverage, and the fact that Arecibo and the GBT are the only existing or proposed U.S. facilities that can effectively contribute to pulsar search efforts, the RMS panel views continued operational support for Arecibo and the GBT over the next decade as essential.

While pulsar timing represents their clearest science driver in the context of Astro2020, both Arecibo and the GBT are also poised to deliver abundantly in addressing other high-priority science questions. Used as auxiliary very long baseline interferometry (VLBI) stations, both facilities greatly improve the sensitivity of high-resolution observations of jets launched by neutron star mergers, tidal disruptions, and other explosive events (COEP-2, COEP-3). Arecibo’s planetary radar capability plays a vital role in characterizing small solar system bodies, whose properties inform the understanding of debris disks around other stars (EAS-1, EAS-3). Sensitive single-dish observations complement the capabilities of the VLA (and potentially the ngVLA), allowing detections of faint emission lines from sources too extended or too low in surface brightness to be studied by interferometric arrays (ISM-1, ISM-2). To deliver on their broad scientific potential, it is important for both observatories to have stable bases of NSF (and, if available, state) funding that can support healthy fractions of peer-reviewed “open time” scientific observations.¹⁵ Time purchases by outside partners will continue to be important going forward, but it is clear from recent experience that large commitments to such partners can deprive observatories of scheduling flexibility and jeopardize their

___________________

¹⁵ For the GBT, whose wide range of high-priority observing modes creates an unusual level of scheduling complexity, a “healthy fraction” would be at least 50 to 60 percent.

ability to execute high-priority fixed-time observations of the sort described above (e.g., for VLBI, radar, and the timing of pulsars outside the NANOGrav network).

M.6.2 Instrumentation Development

As discussed above, the past decade’s scientific advances at RMS wavelengths have been strongly enabled by technological advances and the work of talented instrument builders (including software developers) to leverage them. Maintaining the capacity for instrumentation development across the U.S. astronomical community is essential for the future health and progress of the field. The challenge is to align funding with projects that (1) address technology needs, (2) are on a scale where students and postdoctoral researchers can engage, and (3) have platforms for deployment. While MSIP and MSRI funding of midscale projects plays a valuable role here, smaller-scale projects and cutting-edge technology development efforts are also important and are well matched to dedicated funding via smaller grants (e.g., as provided by the NSF Advanced Technologies and Instrumentation program, which has supported EDGES and new instruments for the GBT and LMT, among other projects). Even if funded, however, new instrumentation can be exploited only if there are facilities where it can be deployed. Historically, this role has been filled mainly by single-dish telescopes that can accommodate a range of guest and/or facility instruments. Recent examples include Arecibo, the GBT, and the LMT, although the closure of the Caltech Submillimeter Observatory (CSO) means there is now a U.S. “deployment capability gap” at the shortest RMS wavelengths. User-contributed instrumentation is also increasingly compatible with large-aperture synthesis arrays, with digitized voltage streams from individual telescopes packaged into standard Ethernet packets and broadcast to multiple backends for different commensal uses (radio transients, pulsars, technosignatures, etc.). For both single-dish and array observatories, the RMS panel suggests that ability and willingness to accommodate the deployment of user-contributed instrumentation (encompassing both hardware and software) factor positively in discussions of federal funding levels. The panel also notes that the scientific potential of a new instrument can only be fully exploited if all of its software needs—including, when relevant, the processing of extremely large data sets—receive robust long-term support. Overall, continued support for the development and deployment of new instruments, the development and continued support of requisite data processing software, and the training of the next generation of instrument builders is a key RMS investment for the next decade and beyond.

M.6.3 Mitigation of Radio Frequency Interference

A significant challenge for all ground-based facilities operating at radio wavelengths in the coming decade will be the need to contend with the growing problem of RFI from human-made sources. The rapid increases in the quantity, bandwidth, and power of RFI from terrestrial sources and satellite constellations pose an existential threat to radio astronomy.¹⁶ Without suitable mitigation efforts, RFI will increasingly impact the ability to detect spectral line emission from atoms and molecules at all redshifts, as well as faint sources of thermal and nonthermal continuum emission that require averaging over large frequency bandwidths.

Protection of the radio sky requires a multi-faceted approach, and there are a number of ways that funding agencies can support this effort. One is to advocate for protection of radio observatories from sources of RFI geographically, spatially, and temporally. Examples of geographical and spectral protection

___________________

¹⁶ L. van Zee, D. DeBoer, D. Emerson, T.E. Gergely, N. Kassim, A.J. Lovell, J.M. Moran, T.J. Pearson, S. Ransom, and G.B. Taylor, 2019, “Spectrum Management,” white paper submitted to Astro2020: Decadal Survey on Astronomy and Astrophysics, https://baas.aas.org/pub/2020n7i136/release/1.

include the preservation of existing radio quiet zones and protected (passive) frequency bands, respectively. Temporal separation includes the exploration of coordinated dynamical sharing of the spectrum between various users. In addition, the RMS panel encourages agencies to provide adequate funding to all current and future RMS facilities for the development of RFI protection and mitigation strategies, including specially designed hardware and sophisticated software tools to excise RFI without indiscriminately deleting signals from real astronomical sources (e.g., temporally varying fast radio bursts and other transients). The panel also encourages agencies to increase the levels of funding they already provide to astronomers and scientific institutions for training, advocacy, and public communication on RFI threats and mitigation approaches. Among these efforts are continued advocacy for scientific use of the spectrum as part of the overall management of the radio spectrum as a shared resource through the work of the National Academies Committee on Radio Frequencies (CORF) and other U.S. and international organizations.

M.7 GUIDING PRINCIPLES

From the earliest stages of the Astro2020 process, multiple stakeholders within the astronomical community have encouraged the development of an ambitious, exciting, and scientifically motivated program for the next decade. The RMS panel has taken this encouragement to heart in arriving at the set of investments described above, which it is offering for the consideration of the Astro2020 steering committee. To inform the steering committee’s deliberations on which selections to order from the lengthy “menu” defined by the program panels’ reports, the RMS panel also offers three top-level principles governing its overall vision. First, the panel views it as important that facility operations budgets for the next decade include full support for the U.S. share of ALMA. As discussed above (and reflected in Table M.1), ALMA is a productive and scientifically vibrant observatory, which has already engaged an impressively broad swath of the global astronomical community and is poised to make further progress on many of the next decade’s high-priority science questions. Continuing operations funding over the next decade would enable further facility improvements within the envelope of the current development budget, with more ambitious and costly improvements possible in future decades. Second, the panel expects that science will flourish best with a program of investments extending over large, medium, and small cost scales, properly balanced so that bigger investments do not crowd out their smaller cousins. This principle is informed by the history of discovery in astronomy, which shows that major disruptive discoveries are made in diverse ways—by individuals and small groups, or by large teams; using modest observing resources, or operating on the cutting edge; driven by serendipity and human inspiration, or achieved through dogged persistence. More generally, literature citation patterns demonstrate that discovery and development in science are strong functions of team size: large teams tend to excel in developing existing ideas, whereas small, agile, risk-tolerant teams are more likely to make disruptive discoveries.¹⁷ The panel’s support of large new facilities is therefore intertwined with its support for traditional individual investigator grants, opportunities for small teams to pursue ambitious observing programs, funding to support the development of new technologies and instrumentation, and facilities that offer opportunities for student training and/or substantial amounts of open time for diverse, risk-tolerant investigations. These more modest investments offer the potential for outsized science return.

The panel’s third governing principle relates to the constructive, respectful, and substantive engagement of the astronomy community with stakeholders from outside that community in addressing environmental and cultural concerns, including at intersections with indigenous rights. The design and construction of

___________________

¹⁷ L. Wu, D. Wang, and J.A. Evans, 2019, “Large Teams Develop and Small Teams Disrupt Science and Technology,” Nature 566:378–382, https://www.nature.com/articles/s41586-019-0941-9.

new ground-based facilities operating at RMS wavelengths offer opportunities to set high standards for professional astronomers’ interactions with indigenous communities, allowing consent to emerge from a sustained and genuinely collaborative process. A rigorous environmental impact assessment can provide an initial sense of the full spectrum of concerns about a new facility, but for all RMS facilities, ongoing consultation with community stakeholders is necessary to minimize negative impacts of operations on surrounding areas and associated cultural activities. To make sure that telescope sites are ultimately returned to their original conditions, projects need to understand and budget for all decommissioning activities before receiving construction funding. To help ensure that RMS facilities’ impacts are as positive for their immediate communities as for society at large, the panel suggests that agencies provide funding for meaningful stakeholder engagement at all phases of the project life cycle.