C

Report of the Panel on Cosmology

INTRODUCTION

We live in a remarkable time in human history when—for the first time—we can observe most of our universe and how it has evolved. The study of cosmology has been expanded far beyond Earth and the Milky Way to encompass vast extragalactic distances and the dramatic evolution of our universe. Enabled by the profound technological advances of the past century, the cosmology community has compiled exquisite measurements and made remarkable discoveries about the history and composition of the universe. The results have led us to a simple empirical cosmological model, referred to here as the standard cosmological model, that unifies a wide range of observational phenomena and provides a crisp starting point for astrophysical computations. This model has continued to successfully explain the measured evolution of our universe even as the body of data that might have challenged it has improved by orders of magnitude over the past two decades. Yet the standard cosmological model remains incomplete, lacking an underlying physical explanation of key ingredients. Realistic physical theories predict a wide range of observable signatures, and the opportunity to discover these signatures is the driving motivation for the coming decade of cosmological research.

The foundation of modern cosmology theory is the Hot Big Bang, in which an initially hot, dense, and nearly smooth universe rapidly expands and cools. Out of this early pressure cooker emerges the universe’s present-day composition: the familiar nuclei and electrons of normal matter, the relic heat now encased in the cosmic microwave background (CMB), a cosmic neutrino background, and an unknown dark matter that outweighs normal matter by a factor of 6.

Using the well-understood physics of plasmas, we are able to map the temperature fluctuations seen in the CMB back to the primordial conditions imprinted in the Big Bang. The small primordial fluctuations are inferred to closely follow a specific statistical pattern: Gaussian correlations with no preferred scale and with all components (i.e., dark matter, nuclei, photons, etc.) varying spatially together, maintaining a fixed composition. While simple, this result is profoundly important because it indicates that the density perturbations were established before the Hot Big Bang phase of cosmic evolution. It is remarkable that these inferred properties match exceptionally well to the predictions of the theory of cosmological inflation, in which extraordinarily rapid expansion in the earliest moments of the universe established the large-scale

homogeneity and flatness of the universe while also causing quantum fluctuations to create exactly the kind of density perturbations we observe.

As time passed, these primordial density perturbations grew in amplitude to form the detailed structure of the universe. Observations of this structure, in surveys of both galaxies and the CMB, clearly require something beyond normal matter to explain the experimental results. In the common paradigm, this is cold dark matter (CDM), some unseen gravitating material that moves nonrelativistically in the recent universe. The standard cosmological model posits the CDM as an empirical extreme—noninteracting, nondecaying, and without any thermal motion—but observations to characterize the properties of dark matter on galactic and sub-galactic scales are limited.

In addition, observations of the recent universe have shown that its expansion is currently accelerating. This remarkable discovery is not readily explained by a model containing only matter, but instead indicates a new feature, dubbed dark energy. In the standard cosmological model, this is Einstein’s cosmological constant: an energy and pressure characterizing empty space whose gravitational effect drives the acceleration. While next to nothing is known about the underlying cause of the acceleration, today’s observations are consistent with the energy density being constant in time, as the cosmological constant would be.

There is no doubt that the standard cosmological model is a triumph. By adopting simple versions of inflation, dark matter, and dark energy, the model can match observational results despite orders of magnitude of improvement in cosmological measurements over the past 20 years. But there is also no doubt that the model is incomplete, as these essential components are not found within the standard model of particle physics. The panel stresses that the familiarity of the names of these components must not obscure this crucial problem. While Occam’s razor favors the adoption of the simplest physical theory, the standard model of cosmology is not physically grounded, and particle physics models built to reproduce our cosmological observations almost invariably have observational signatures that deviate from the standard model—deviations we may well be able observe this decade.

State of the Field

In the Astro2010 decadal survey, the Panel on Cosmology and Fundamental Physics presented four questions—(1) How did the universe begin? (2) Why is the universe accelerating? (3) What is dark matter? (4) What are the properties of neutrinos?—as well as a discovery area in gravitational wave astronomy.¹ Progress on observational and experimental data sets to study these topics has been tremendous. Some notable highlights are:

• An explosion of arcminute-scale CMB data has extended the standard cosmological model to unprecedented precision, produced superb catalogues of galaxy clusters, and has opened the frontier of CMB lensing and the kinematic Sunyaev-Zel’dovich (SZ) effect.
• Searches for inflationary gravitational waves have improved in precision by more than a factor of 10, to the point where they disfavor many of the simplest models of inflation.
• Maps of large-scale structure have enabled a range of scientific advances, including (but not limited to) measurements of the cosmic distance scale over a wide range of redshift using the baryon acoustic oscillations (BAO).
• Cosmological weak lensing has leapt forward, with the uncertainty in the lensing-inferred amplitude of late-time density fluctuations decreasing fourfold, to about 3 percent.

___________________

¹ National Research Council, 2011, Panel Reports—New Worlds, New Horizons in Astronomy and Astrophysics, The National Academies Press, Washington, DC, https://doi.org/10.17226/12982.

• Measurements of the Hubble constant from the direct distance scale and from strong gravitational lensing have improved to a precision of about 2 percent, while supernovae measurements at cosmological distances have driven precision on the dark energy equation of state w below 5 percent.
• The first gravitational wave events have been detected, including an initial application of the standard siren method of constraining the cosmic expansion rate. The observed travel time of gravitons resulted in an improvement of more than 10 orders of magnitude in the determination of the speed of propagation of gravitational waves.
• The primordial deuterium-to-hydrogen ratio has been measured to 1 percent precision.
• A wide range of searches for the astrophysical detection of dark matter have occurred, greatly improving the limits on many possible scenarios.

While much of the decade has been marked by a concordance between experimental results and the predictions of the standard cosmological model, not everything agrees. Direct measurements of the Hubble constant, H₀, tend to give higher values than those implied by CMB and large-scale structure data. The Experiment to Detect the Global EoR Signature (EDGES) project reports an 80 MHz spectral distortion consistent with 21 cm absorption from redshift 17, but with an amplitude several times larger than predicted. And there is a haze of gamma-ray emission that peaks at around 1 GeV from the inner Milky Way, consistent with a dark matter annihilation signal but also possibly explained as high-energy emissions from undetected pulsars. Whether resolving these discrepancies will ultimately require a new addition to the standard cosmological model is unknown, but they highlight the importance of a broad experimental program.

The coming decade will provide unprecedented cosmological opportunities. One of the major achievements of the past decade has been the development of a new generation of facilities, now under construction or in early operations, that will push the field dramatically forward. The Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST), Dark Energy Spectroscopic Instrument (DESI), Subaru/Prime Focus Spectrograph (PFS), Euclid, and the Nancy Grace Roman Space Telescope (formerly Wide Field Infrared Survey Telescope) will provide superb optical and near-infrared imaging and spectroscopy surveys. The Spectro-Photometer for the History of the Universe, Epoch of Reionization, and Ices Explorer (SPHEREx) will take observations further into the infrared with low-resolution spectroscopic mapping. The South Pole Telescope (SPT)-3G, Advanced Atacama Cosmology Telescope (ACT), and Simons Array will produce high-sensitivity maps of the CMB polarization at arcminute-scale resolution, while experiments such as the Keck Array, Background Imaging of Cosmic Extragalactic Polarization (BICEP) 3, Spider, and Cosmology Large Angular Scale Surveyor (CLASS) will measure CMB polarization at large angles. The recently launched Extended Roentgen Survey with an Imaging Telescope Array (eROSITA) will produce a sensitive X-ray map of the full sky, with more detailed measurements of individual sources to be made by the Advanced Telescope for High Energy Astrophysics (Athena) mission early in the 2030s. Gaia will continue its mission, increasing its sensitivity to stellar proper motions and astrometric binaries. Gravitational wave (GW) observatories are rapidly extending their reach, including the ongoing development of the Laser Interferometer Space Antenna (LISA) mission. These surveys are complemented by a wide range of narrow-field facilities that allow us to pursue important companion and follow-on studies. The James Webb Space Telescope (JWST) is the coming exemplar of these facilities, but nearly every large telescope plays some role in cosmological science, spanning all wavebands from pulsar timing in the radio, to spectroscopy of faint transients in the optical, to dark matter annihilation searches in the gamma rays.

COSMOLOGY IN THE 2020S AND BEYOND

With both compelling mysteries and extensive observational means by which to explore them, this will be an amazing decade for cosmology. In this report, the panel identifies four major science questions

for the upcoming decade: (1) What set the Hot Big Bang in motion? (2) What are the properties of dark matter and the dark sector? (3) What physics drives the cosmic expansion and large-scale evolution of the universe? (4) How will measurements of gravitational waves reshape our cosmological view? The panel also identified a discovery area: the Dark Ages as a cosmological probe.

These are familiar questions, but our experimental ability to tackle them is increasing rapidly and radically. As the panel explains below, the range of possible discoveries in these foundational areas is very broad. The motivation and capability to explore the unseen constituents and earliest moments of the universe remains one of the central themes of astrophysics.

It is also important to note the deep and increasing connection between cosmology, with its precision observations over enormous volumes, and the rest of astrophysics. Cosmological probes are invariably intertwined with their astrophysical context. These connections are often couched as “systematic uncertainties,” which ignores the synergistic opportunities that come with the co-development of different areas of the field. An obvious example is galaxy formation, which is now highly tied to its cosmological context. The need to embrace and extend these connections will only increase as the data become more sensitive and we seek more subtle cosmological signatures.

A major purpose of this report is to connect these scientific questions to the capabilities needed to pursue them. In the discussion of each question, the panel has identified the key observational and experimental facilities required. Where the panel has identified an important quantitative science goal, it has stated that goal. However, in many cases, discovering the physics behind our cosmology is an exploratory endeavor without a specific threshold. In these cases, the panel has opted to use the term “next generation” to indicate where future yet-unfunded facilities could provide important improvements, typically at the order of magnitude level. The practical selection of precise quantitative requirements will necessarily depend upon a balance of technical opportunities, cost, timing, and risk. The panel uses the term “funded upcoming facilities” to discuss the important role of facilities currently under fully funded construction; in some cases, the gains from these facilities will be so transformative that the panel believes it is critical to assess their early results before identifying opportunities for future generations of experiments.

C-Q1. WHAT SET THE HOT BIG BANG IN MOTION?

A vast number of observations allow us to characterize the state of the universe early in its history when it was hot, dense, and expanding rapidly. One of the fundamental discoveries of modern cosmology is that the primordial density fluctuations, the seeds of the structure of the universe observed throughout cosmic history, were created before the hot phase of the Big Bang. As a result, studying these primordial fluctuations provides a unique window into physics at extremely early times and at energy scales many orders of magnitude above what researchers can access in the laboratory.

The question of what process set the Hot Big Bang in motion and created the seeds of structure has been with us for many decades. Early theoretical developments, together with observations over the past two decades, have established the inflationary paradigm as the dominant picture in the field. In inflation, the universe went through an early period of accelerated expansion that smoothed out prior anisotropies, ending in a dramatic event that filled the universe with high-energy particles. The initial seeds for structure resulting from this period are expected to have simple properties: the statistics of the seeds follow almost perfect Gaussian correlations, invariant in scale, with no spatial variation in the composition among different particles. Deviations from these simple predictions carry most of the information about the inflationary period. In addition, inflationary theories always predict a stochastic background of gravitational waves whose amplitude and scale dependence, if measured, would provide important information about inflation as well as the quantum theory of gravity.

In the past decade, the Planck satellite measured departures from scale invariance of the power spectrum of density fluctuations, in line with the expectations of the simplest inflationary models, and placed exquisite constraints on departures from Gaussianity and fluctuations in the composition of the universe. The BICEP/Keck CMB polarization experiments put stringent upper limits on the amplitude of inflationary gravitational waves. We now stand at a crossroads for the inflationary paradigm because the improved measurements that could be performed in the coming decade will allow us to cross important theoretical thresholds and significantly improve our understanding of the inflationary epoch.

C-Q1a. Primordial Gravitational Waves

Gravitational waves are inevitably produced during an inflationary epoch and would survive to the present day. If inflation occurs at a sufficiently high energy scale, those gravitational waves can be observed through their imprint on the large-scale polarization pattern of CMB maps, the so-called B-modes. The past decade has seen a steady advance in orbital, suborbital, and ground-based experiments to measure CMB polarization at exquisite precision. The amplitude of this gravitational wave background is quantified using the ratio of the amplitude of gravitational waves (tensor) to that of density fluctuations (scalar) produced during inflation, a ratio known as r. At the time of this report, the experimental constraints at r <0.06 (95 percent confidence) already rule out very interesting portions of the parameter space of models.

A concerted effort over the next decade to improve the sensitivity to gravitational waves by a factor of 10–100 would cross important theoretical thresholds. In particular, a measurement of r >0.01 would imply that the inflationary field moved over very large distances, larger than the Planck scale, in field space as inflation proceeded. Such an observation would be highly constraining to quantum gravity theories. Furthermore, models that naturally explain the scale-dependence in the density fluctuations observed by Planck by fixing the spectral index of density fluctuations to be inversely proportional to the number of e-folds of observable inflation predict r >0.001. A next generation of large-angle, high-sensitivity CMB polarization measurements along with arcminute-scale maps to provide the requisite control of foreground gravitational lensing can bring early universe cosmology to these two important discovery thresholds. Even if the gravitational waves are not detected, such limits would lead to a significant improvement in the understanding of the primordial universe.

C-Q1b. Non-Gaussianity of the Large-Scale Structure of the Universe

The statistical properties of the primordial density fluctuations in the universe encode information about the physical processes responsible for their generation. Minimal models of inflation involve a single field that evolves during inflation, serving as a clock that determines when inflation ends and the Hot Big Bang begins. Interactions between fluctuations of this clock field, or between such fluctuations and those of other fields during inflation, generically cause noticeable departures from Gaussian correlations in the distribution of the primordial structural seeds. If such departures can be detected in surveys of cosmological structure, then one can study these inflation-era interactions and constrain the physical origin of perturbations.

A wide range of possible non-Gaussian signals in the statistics of the primordial seeds are of cosmological interest, but one particular form contains an important quantitative threshold for next-generation surveys. In single-field inflation models, fluctuations correspond simply to time delays between different regions of space. This remarkable fact implies that a particular kind of deviation, called local-type non-Gaussianity, must be extremely small in these scenarios (f_NL,local << 1, in the usual parameterization). In contrast, if the observed density fluctuations originate from fields other than the inflationary clock field, or were not created during an inflationary period, interactions between the fields are generically not suppressed and produce

large local-type non-Gaussianities in the primordial seeds, with f_NL,local of order 1 or larger. As a detection of primordial local-type non-Gaussianity with f_NL, local of order unity would falsify single-field inflation, the search for primordial non-Gaussianity, either to detect a signal or to constrain f_NL,local to be below 1 with 5-σ significance, is particularly important.

Advancing to this level of sensitivity will require three-dimensional surveys of very large volume, high sampling density, and exquisite large-scale systematic control to accurately measure a large number of fluctuation modes. The balance of volume, sampling density, and target properties could vary among viable surveys. The National Aeronautics And Space Administration (NASA) SPHEREx mission, as well as redshift surveys such as from DESI and Euclid, will be the next step, but even larger surveys at higher redshift are needed to reach the target value. These will require next-generation high-multiplex spectroscopic facilities, likely in the optical and infrared but possibly at radio wavelengths, with the goal of mapping as many linear-regime modes of the primordial structural seeds as possible, at least a factor of 10 more than funded upcoming facilities. Cross-correlations of such surveys with CMB anisotropy maps can leverage the kinematic SZ effect to improve constraints on the largest-scale density perturbations.

C-Q1c. The Initial Power Spectrum of Density Fluctuations

In most scenarios for the production of primordial structural seeds, different spatial scales are generated at different times. This presents an opportunity to explore the history during the production era by measuring the amplitude of fluctuations over a wide range of scales. The physical processes governing the early universe could involve a number of additional degrees of freedom with a wide range of interactions, and in many cases, these dynamics reveal themselves in features on top of an otherwise smooth primordial power spectrum. Similarly, a detection of large-scale fluctuations between different species of matter would be an important new constraint on inflation and the thermal history of the universe. Detailed measurements of the primordial power spectrum can be advanced with new maps of the CMB, large-scale structure, and clustering of the intergalactic medium; in particular, the surveys required to measure non-Gaussianity will be excellent for measuring the large-scale power spectrum. Deviations at much smaller scales could be detected owing to their impact on CMB spectral distortions or on small-scale structure in early galaxy formation.

Summary of Capabilities Needed for C-Q1

Capabilities needed include a next generation of CMB polarization experiments (both large and small angular scales) to seek the primordial gravitational waves, and a next-generation large-volume redshift survey to seek primordial non-Gaussianity.

C-Q2. WHAT ARE THE PROPERTIES OF DARK MATTER AND THE DARK SECTOR?

Since the Astro2010 decadal survey, the field of dark matter theory and detection has undergone a paradigm shift. In previous decades, the field focused primarily on two dark matter candidates, weakly interacting massive particles (WIMPs) and axions, motivated mainly by their ability to solve long-standing open questions within the Standard Model of particle physics. However, recent work has emphasized that dark matter may arise from a dark sector more analogous to the visible sector of familiar particles, with its own dynamics and forces and with new terrestrial and astrophysical signatures. The non-detection of physics beyond the Standard Model at the Large Hadron Collider (LHC)—notably not finding signatures of supersymmetry—has served to further highlight the possibility that the dark sector need not be closely connected to well-recognized questions in particle physics. The breadth of new dark matter candidates and

dark sector dynamics that have arisen from these recent explorations offers new motivation and opportunities to detect astrophysical signatures of dark matter.

C-Q2a. Dark Sector Signatures in Small-Scale Structure

The only irreducible interaction of the dark matter is through gravity, and it is through the gravitational interaction that all of our knowledge of dark matter in cosmology arises. As such, the way dark matter clusters gravitationally is a unique window into the nature of the dark matter and its attendant forces. Indeed, such measurements have already been very effective on super-galactic scales in the Milky Way for establishing the cold dark matter model. However, the clumpiness of dark matter on small scales is today only loosely constrained, save for the extreme case of objects compact enough to produce microlensing of stars. Many theories of dark matter beyond the WIMP paradigm feature modifications of the scale-invariant power spectrum—for example, resulting in gravitational collapse in the early universe into dark matter mini halos, which can be thousands or even a million times more dense than Lambda cold dark matter (ΛCDM) sub-halos. New forces also generically appear in a broad range of dark matter models, giving rise to dark matter self-interactions and modifications of CDM predictions for the abundance and density profile of halos. Such predictions and others provide compelling motivations to sharply extend our study of small-scale clustering.

The small-scale clustering of dark matter can be explored over a wide range of length and mass scales, ranging from the substructure of clusters, to dwarf galaxies, to sub-planetary-mass relics. This, in turn, relies on a large set of observational opportunities. Higher masses, above a million solar masses, can be sensitively probed by resolved gravitational lenses and by dwarf galaxy counts and mass profiles. These studies will be substantially advanced by the funded upcoming facilities LSST, Euclid, and Roman Space Telescope, and the panel notes the opportunity of the Atacama Large Millimeter/submillimeter Array (ALMA) to characterize gravitational lenses. Next-generation large-aperture optical telescopes will be critical for spectroscopy of both dwarf galaxies and lens systems.

Stellar astrometric measurements, like those of galactic stellar streams and the survival or disruption of wide stellar binaries, probe intermediate masses from the dwarf-galaxy scale down to about a solar mass. This area is being revolutionized by Gaia, and augmented by both LSST and the funded upcoming wide-field high-multiplex optical/infrared spectrographs. At yet lower masses, possibly as low as 10^–14 solar mass, pulsar timing arrays offer a novel opportunity by searching for timing anomalies owing to gravitational lensing by dark matter lumps passing between the observer and the target pulsar, even if the lumps are not compact enough to generate microlensing of the flux. Enhancing the network of radio telescopes capable of precision pulsar timing in order to substantially increase the precision, cadence, and sample size of timing measurementscould considerably extend these low-mass searches.

C-Q2b. Dark Sector Imprints on Big Bang Nucleosynthesis and Recombination

The dark sector can leave other imprints on cosmic evolution. Current measurements still allow significant room for additional dark sector contributions to the early universe’s energy density. One potential contribution—well motivated by numerous extensions of the Standard Model of particle physics—is from the relics of light particles produced thermally in the early universe, here called dark radiation. Dark radiation and its self-interactions can be constrained by arcminute-resolution CMB measurements of the recombination-era damping of waves in the baryon-photon fluid. With a next generation of such experiments, light-particle relics could be detected at an energy density only 1–2 percent of that of the cosmic

neutrino background, allowing detection of relics that thermally decouple from the Standard Model before the quantum chromodynamics phase transition.

Measurements of the baryon density provide another window into the early universe. These come independently from CMB anisotropies and from the light-element abundances predicted from Big Bang Nucleosynthesis (BBN). These agree except for the ⁷Li abundance, which shows a long-standing factor of 2 discrepancy. A measurement of the ⁷Li abundance in low-metallicity diffuse gas could determine whether there is indeed an anomaly in the BBN predictions. If so, dark sector physics is a candidate explanation. For example, dark sector models can produce relativistic by-products that lead to observable signatures in BBN. In addition, further improvements in the measurement of the helium and deuterium abundance can constrain theories of light dark matter. Such abundance measurements require high-resolution ultraviolet and optical spectroscopy with a next generation of larger aperture telescopes.

C-Q2c. Annihilation By-Products

While the LHC has not discovered supersymmetry or any signs of new physics at the weak scale, there are important models of supersymmetric dark matter that the LHC cannot reach, notably those in which the dark matter interacts only with the weak force and all the supersymmetric particles interacting directly with the strong force are too heavy to produce at the LHC. These models typically predict dark matter candidates at the few-TeV mass scale, with cross sections for annihilation to photons that make them observable with next-generation Cherenkov telescopes that have sufficient scope to detect an annihilation cross section to two photons larger than 10^–28 cm³/s for dark matter masses of 1 TeV. This represents a crucial opportunity to search an otherwise-unreachable part of the weak-scale dark matter model space.

At the same time, at a much lower mass scale, the origin of the GeV gamma-ray excess toward the galactic center remains unknown. It may be owing to dark matter or to millisecond pulsars. A goal of the next decade is to solve this puzzle. A next-generation gamma-ray telescope with better angular resolution than the Fermi satellite to resolve point sources in the galactic bulge, or better sensitivity to photons from dark matter annihilation in dwarf galaxies, would be a powerful tool for doing so. Less directly, deeper pulsar searches with next-generation radio telescopes could identify potential sources.

Summary of Capabilities Needed for C-Q2

The search for dark matter signatures is wide-ranging and exploratory, but the next generation of radio telescopes for pulsar timing, large-aperture optical telescopes, high-resolution CMB polarization mapping, GeV telescopes, and TeV-scale Cherenkov telescopes are particularly important to make progress in this field.

C-Q3. WHAT PHYSICS DRIVES THE COSMIC EXPANSION AND LARGE-SCALE EVOLUTION OF THE UNIVERSE?

One of the striking features of dark energy is that it explains not only the accelerating expansion seen in the late universe but also the cutoff of large-scale structure growth. Understanding the physics behind cosmic expansion requires testing both the expansion of space and the growth of structure across cosmic time. Together, these observations will provide an end-to-end test of our standard cosmological model and measure the properties and masses of neutrinos—the last known un-weighed constituent of our universe. The panel stresses the importance of pursuing precise and accurate measurements that span a wide range of redshifts and clustering scales, and of doing so with methods that are complementary in both their

systematic errors and their sensitivity to the physics of the cosmological model. Below, a number of key observations are highlighted to elucidate the physics behind cosmic expansion.

C-Q3a. The Physics of Cosmic Acceleration

One of the most profound discoveries in modern cosmology has been the accelerating expansion rate of the universe, which has now been independently confirmed by multiple probes. The cause of this acceleration, which has been dubbed “dark energy,” remains a mystery. The leading theory is that of a cosmological constant with a constant energy density over time, today comprising approximately 70 percent of the energy density of the universe. Testing for deviations from the cosmological constant model to the practical limits of available methods remains a key goal of the field of cosmology, as such deviations would be a signature of new physics or reveal a breakdown of general relativity (GR) at large scales.

It is critical to probe this acceleration with diverse and independent probes that provide constraints on both distance scales and the growth of structure. Key methods to explore this question in the coming decade include weak gravitational lensing of distant galaxies and the CMB, BAO and redshift-space distortion measurements from redshift surveys, supernovae Ia distance measurements, Hubble constant measurements, and galaxy cluster abundances. The funded upcoming wide-field survey facilities and CMB experiments will produce a major leap forward in both statistical reach and systematic control of these methods. Beyond currently funded capabilities, next-generation CMB surveys offer important new opportunities in both CMB lensing—an emerging field that will provide a longer redshift lever arm for lensing studies of structure growth—and kinematic SZ studies, which combine CMB and redshift survey data to measure large-scale velocity flows. Beyond the major wide-field surveys, other capabilities will be needed to support these methods—for example, deep spectroscopic training data sets for photometric redshifts of weak lensing samples, follow-up telescope resources for Type Ia supernovae, and narrow-field instruments for local distance measurements and strong lensing cosmography. Narrow-field observations would be particularly advanced by the next generation of large-aperture optical telescopes.

The flexibility of the Roman Space Telescope mission will provide an important and powerful capability to investigate opportunities or questions raised by many of the funded upcoming wide-field survey facilities and CMB experiments, listed earlier in this appendix, that will begin earlier in the decade. The middle of the decade would be an excellent opportunity to assess progress and identify further scientific and technical opportunities in the field. The panel notes the particular importance of studying the low-redshift period where dark energy dominates the expansion rate. Opportunities for doing so include (1) standard candle and standard siren methods, which are not limited by the cosmic variance of large-scale structure, and (2) a next generation of densely sampled galaxy redshift and lensing surveys to test the impact of dark energy on structure formation with greater sensitivity.

In order to thoroughly test the predictions of the cosmological model, robust and model-independent tests for deviations from GR’s prediction for the growth of large-scale structure are an essential element of the research program of the field. These tests largely use the same data sets as tests of cosmic acceleration. Existing data have motivated attempts to build modified gravity theories that explain this acceleration while remaining consistent with stringent constraints on gravity from other measurements—a challenging exercise that has deepened the understanding of GR and inspired new observational tests of gravity.

C-Q3b. The Properties of Neutrinos

The relic population of neutrinos forms an important dynamical component of the universe. This fact enables large-scale structure surveys to probe fundamental physics of neutrinos, such as their mass and possibly even their self-interaction cross section. Further, this characterization of neutrino properties will also be necessary to model the cosmological observables that feed into the dark energy measurements described above.

One key goal for next-generation surveys is to determine whether the neutrino mass hierarchy is “normal,” with two similar lower-mass states much below the third, higher-mass state, or “inverted,” with two similar higher-mass states well above the third, lower-mass state. Flavor oscillation experiments imply a minimum sum of the three masses of 0.06 eV in the normal hierarchy and 0.12 eV in the inverted one; distinguishing these cosmologically requires measuring the total mass to a 5-σ precision of 0.06 eV. This is typically done by comparing the amplitude of clustering at z = 1,000 from CMB observations to that in the late-time evolved universe. Because of parameter degeneracies, achieving such tight constraints will also require significant improvements in constraints on the optical depth to reionization, τ. These are obtainable from next-generation large-scale (ℓ<30) CMB E-mode polarization measurements and potentially also from small-scale kinematic SZ measurements and measurements of the Dark Ages. Improving the measurement of τ to the required precision of 0.002 (1 σ), near the CMB cosmic variance limit, in the next decade will be vital for achieving precise constraints on the neutrino mass hierarchy. Similarly, the low-redshift amplitude of clustering needs to be measured to this level; this can be done with the funded upcoming facilities that will measure gravitational lensing of both galaxies and the CMB, cluster abundances, the intergalactic medium, and redshift-space distortions.

C-Q3c. End-to-End Tests of Cosmology

Diverse and precise cosmological measurements that probe multiple epochs of cosmic history will allow for stringent cross-epoch end-to-end tests of cosmology, enabling explorations of the consistency of cosmological models across cosmic time and providing ways to challenge the standard cosmological model. To maximize the power of these tests, measurements at low redshift will need to reach a precision comparable to those from the early universe, notably from observations of the CMB.

As an example, at the precision of today’s (2020) surveys, observations have revealed a growing discrepancy between local and CMB-epoch calibrations of the Hubble constant. While this may be a sign of unaddressed biases in the measurements, it could also be an indication of new physics beyond the standard cosmological paradigm. These tests concern comparison of absolute distance scales in the early and late universe and require improved measurements across cosmic time: local expansion rate measurements at low redshift; galaxies, quasar, and Lyman-alpha forest BAO measurements, strong lensing cosmography, supernovae, and GW standard siren measurements at intermediate redshifts; and small-angle CMB measurements at high redshift. If tensions persist, efforts will be needed to probe scales that can distinguish between changes in the early expansion rate and the speed of sound in the primordial plasma. Such measurements will be accessible to probes of the Dark Ages discussed in the discovery area below.

Another ongoing area of work is to compare the amplitude of structure fluctuations at low and high redshift. Although this was already mentioned in the context of neutrino masses and modified gravity explanations of cosmic acceleration, the evolution of large-scale structure could reveal other extensions in the dark sector, such as late-time decaying particles or time-varying masses.

Summary of Capabilities Needed for C-Q3

To complement the funded upcoming large survey facilities targeting cosmic expansion, next-generation high-angular-resolution CMB experiments to measure lensing and wide-angle CMB polarization maps to improve the measurement of the optical depth to reionization are needed. Support is further needed for facilities for spectroscopic training of photometric redshifts for weak lensing samples, follow-up of Type Ia supernovae, and narrow-field instruments for local distance measurements and strong lensing cosmography, such as could be provided by large-aperture optical telescopes. The panel anticipates that the early returns from the funded upcoming facilities may be transformational, and a mid-decade assessment will be important to shape plans for future investments in this area, including opportunities to tune the observing plan for Roman Space Telescope.

C-Q4. HOW WILL MEASUREMENTS OF GRAVITATIONAL WAVES RESHAPE OUR COSMOLOGICAL VIEW?

Just 10 years ago, the Astro2010 Cosmology and Fundamental Physics panel listed gravitational wave astronomy as its discovery area. In the intervening decade, Laser Interferometer Gravitational-Wave Observatory (LIGO) has observed the merger of tens of binary black holes, and its discovery of a binary neutron star merger has heralded the era of multi-messenger astronomy. In the next decades, gravitational wave measurements will span a wide range of frequencies, from nHz with pulsar timing arrays, to mHz with LISA, to kHz with ground-based instruments. This new astronomical window will open a large dynamic range of time and scale for cosmological inferences, from a potential stochastic gravitational background originating in the early universe, to new particles created in the vicinity of rotating black holes, to the current expansion rate of the universe. Indeed, our understanding of the potential reach of these observations is still maturing, and they may take us in unexpected directions. While some of the topics described below overlap aspects of the previous questions, the panel believes that this new and rapidly expanding view of the universe offers an opportunity for cosmological discovery that needs to be specifically recognized.

C-Q4a. The Stochastic Gravitational Wave Background

The detection in the nHz to kHz bands of a stochastic gravitational wave background beyond that expected from compact-object sources would provide a unique window into the thermal history of the universe during otherwise inaccessible times. As the universe cools, phase transitions and their associated topological defects could produce gravitational waves. Gravitational waves far more intense than those expected from simple inflation models could also arise from the start of the Hot Big Bang, and comparison to the ultra-low-frequency waves being sought in CMB large-angle polarization measurements would probe a large lever arm in the spectrum of gravitational waves. Pulsar timing arrays and gravitational wave detectors are sensitive to this stochastic background, provided that one can isolate the background signal from that of nonprimordial compact object mergers.

C-Q4b. Standard Sirens as a New Probe of the Cosmic Distance Scale

With the discovery in 2017 of both gravitational waves and electromagnetic signatures from a binary neutron star merger, scientists were able to make the first “standard siren” measurement of the cosmic distance scale. This measurement was enabled by the exquisite predictive power of GR, which allows the gravitational luminosity distance to be directly measured for individual compact object mergers. This method has notable

advantages, such as its reliance on laboratory calibration and independence from the effects of astrophysical dust, although, like standard candle methods, it suffers at high redshift from magnification uncertainties from gravitational lensing. In the coming decades, standard siren samples will increase enormously in size and quality. Using low-redshift mergers, LIGO and Virgo may provide an independent assessment of the current Hubble constant tension to the 1 percent level. Using supermassive black hole binaries, LISA will reach out to redshift ~10, providing a means to build a single distance scale over a remarkable span of cosmic history. Eventually, using the individual events to build maps may enable novel cosmological tests.

Electromagnetic counterparts will be crucial for many applications and will require extensive observing resources to locate and study. The panel expects that the gravitational wave network sensitivity will soon place substantial demands on the capacity of the observatories needed to find the electromagnetic counterparts and acquire the source redshifts. Continued access to follow-up resources will be important to speed progress using this probe.

C-Q4c. Light Fields and Other Novel Phenomena

The increased precision and expanded frequency range of future gravitational wave facilities will offer numerous opportunities to uncover novel phenomena that are now only theoretical speculations. To list some cosmological examples: (1) Gravitational wave emission from rapidly spinning black holes may reveal light bosonic particles whose Compton wavelength matches the horizon size. These bosons need not be a major component of dark matter, but their existence would be an intriguing clue to other light states. (2) Mergers might be found from sub-stellar mass black holes or from extreme redshift, suggesting a new cosmological source of compact objects. (3) Waveforms of merger events might display signatures of gravitational lensing by dark matter substructure. (4) Study of compact object mergers might even reveal a new aspect of strong-field gravitational physics that could perhaps be connected to cosmic acceleration. We should be prepared to be surprised when looking through the gravitational wave window into the unseen relativistic universe.

Summary of Capabilities Needed for C-Q4

Capabilities needed include improvements in gravitational wave detection, particularly through deployment of LISA, coupled with increasing efforts in multiwavelength electromagnetic and multi-messenger studies to characterize the population of merger events.

DISCOVERY AREA: THE DARK AGES AS A COSMOLOGICAL PROBE

Our understanding of how the universe began is measured through the fingerprints that the Hot Big Bang left in matter density fluctuations. Unfortunately, these fingerprints are often smudged. At small scales, Silk damping in the CMB and the nonlinear astrophysics of galaxies hides and confuses the primordial density fluctuations, while at large scales, measurements are limited by the volume of space that can be observed with galaxy surveys or on the surface of last scattering. At the end of the Dark Ages, the neutral hydrogen pervading the universe became visible against the CMB backlight, enabling observations of the primordial density fluctuations over a vastly larger range of scales than for any other cosmological probe. The panel sees 21 cm and molecular line intensity mapping of the Dark Ages and reionization era as both the discovery area for the next decade and as the likely future technique for measuring the initial conditions of the universe in the decades to follow.

C-DA1. The End of the Dark Ages

As the first luminous objects formed, it is expected that the vast majority of the baryons were very cold and that the hydrogen spin temperature was in equilibrium with the CMB. The first Lyman-alpha photons then coupled the hydrogen spin temperature to the gas temperature, highlighting the neutral hydrogen against the CMB. As illustrated by the theoretical work interpreting the surprisingly large global absorption signal detected by the EDGES experiment, the time evolution and spatial fluctuations in this spin temperature provides a wealth of cosmological information.

During the Dark Ages, astrophysical structure formation responded dramatically to the dark matter power spectrum at the smallest scales. Models of dark matter that suppress small-scale power (e.g., warm dark matter or dark sector interactions) delay the formation of the first luminous objects, while theories that enhance small-scale power (e.g., primordial black holes, dark sector interactions that boost early black hole formation, or models that create compositional fluctuations on small scales) advance this timing. By using the emergence of luminous systems as a timestamp, cosmologists can leverage the astrophysics of early structure formation to probe the primordial power spectrum at currently inaccessible scales.

Unlocking this cosmological window will require advances in both measurements and theory, but it appears attainable in the coming decade. A goal for the coming decade is reconnaissance across a wide range of redshift, primarily with next-generation interferometric mapping supported by global single-receiver measurements, in order to map the temperature history of the intergalactic gas. While small changes in the timing of galaxy formation can be caused by astrophysical details, large changes in when structure formed would be a hallmark of new physics in the dark sector. As our understanding of reionization and the late Dark Ages improves, we will increasingly be able to disentangle the astrophysics of reionization from effects of cosmology.

C-DA2. The Future of Primordial Density Mapping

As described earlier, understanding the initial conditions of our universe requires observing the primordial density fluctuations. Intensity mapping of neutral hydrogen has the potential to measure these fluctuations with unprecedented precision and reach. Prior to the onset of nonlinear collapse and galaxy formation, the primordial density fluctuations can be measured to much smaller scales than possible at later times. Further, one can measure far more modes with hydrogen intensity mapping than in the CMB primary anisotropies, as one is no longer limited by the Silk damping of the CMB and one is using a three-dimensional map rather than an angular map. A window for this direct mapping of pre-collapse structure is predicted to exist at redshift 50. Here, increases by factors of over 100 in scale and a billion in number of modes might be available, giving intensity mapping the potential to provide the next major leap in the understanding of the initial conditions imprinted on the primordial density fluctuations by inflation.

But intensity mapping measurements are in their infancy, and the most ambitious program at redshift 50 requires space-based measurements. While significant technical progress has been made toward the first line measurements of the power spectrum at high redshift, the state of the art is still decades away from superseding the CMB in scientific reach. As in the cases of the CMB, gravitational waves, and weak lensing, the development of intensity mapping from concept to a robust cosmological tool will take several decades of steady support.

In the coming decade, the panel anticipates that neutral hydrogen intensity mapping will mature to the point that it can make the first anisotropy measurements of reionization. This is a crucial milestone, and measuring the process of reionization and the CMB optical depth will improve the current understanding of cosmology. The panel also hopes to see the first measurements of the BAO scale using either the 21-cm or other atomic or molecular emission lines. As these techniques mature, the panel expects the precision,

angular scale, and redshift of the measurements to steadily improve. A 30- to 40-year goal would be to map the density fluctuations in the pre-reionization universe with an unprecedented number of modes traceable to the primordial density fluctuations, using the power spectrum and non-Gaussianity to measure the statistical initial conditions of the universe.

Summary of Capabilities Needed for the Discovery Area

Needed capabilities include next-generation 21-cm interferometers targeting both the reionization epoch and lower redshifts, along with planning toward very high redshift mapping. Progress will require both higher sensitivity and a better understanding of instrumental systematics and astrophysical couplings.

CROSSCUTTING CAPABILITIES

This appendix identifies new observational capabilities needed to address the science questions and discovery area. In addition, the panel identified the following crosscutting capabilities needed to support the overall cosmological research enterprise.

Tremendous opportunities will be offered by the facilities currently nearing completion. These facilities will produce vast data sets that will be useful across a wide range of efforts, especially when these data sets are combined. Fully leveraging the cosmological utility of these observations will require elaborate analyses and extensive collaboration beyond the scope of an individual investigator grant. The panel is concerned that potential scientific output could be unrealized owing to a lack of available human resources to fully and, where appropriate, collaboratively analyze and exploit data sets. The panel urges that attention be given to the support of these larger analysis efforts.

Computational and theoretical studies of cosmology are critical to support the field. The impact of modern computing on cosmology is ubiquitous—ranging from ambitious data reduction methods, to detailed statistical analyses, to high-performance simulations—while theoretical research continues to contribute important new physical hypotheses to be tested as well as calculational and statistical opportunities to extend methods for interpreting the complex data sets derived from both observations and simulations. Cosmology is particularly remarkable within astrophysics, as it is often concerned with testing theoretical models that provide specific and realistic physical initial conditions, such that forward simulation to compare to observational results is an essential aspect of the interpretation of cosmological data. Advances in theory and scientific computing (the latter enhanced by advances from the data science and machine learning communities) directly enable analyses from current experiments and help to guide the design of future ones. A commitment to the public release of both data and analysis software from next-generation projects, as well as to the development of software that makes good use of the computational power provided by new computer hardware architectures and facilities, will continue to push the state of the art in these areas.

Last, today’s cosmological experiments and facilities rely on technology far beyond what was available in past decades, and the health of the field surely depends on continuing this technological growth. Whether for detectors, correlators, robotic mechanisms, novel optics, or technologies enabling cheaper and more capable spaceflight, pushing the state of the art in cosmology requires strategic support for technology development projects.

CONCLUSION

The origin, composition, and physical laws of the universe are ancient sources of wonder and ever-present drivers in our study of astronomy. The coming decade will be a bold new chapter in that cos-

mological story, with ambitious facilities and unprecedented data sets uniting with powerful statistical, computational, and analytical methods to explore many different frontiers.

This panel has identified four critical science questions and one discovery area that it believes are ripe for substantial progress this decade: (1) What set the Hot Big Bang in motion? (2) What are the properties of dark matter and the dark sector? (3) What physics drives the cosmic expansion and large-scale evolution of the universe? (4) How will measurements of gravitational waves reshape our cosmological view? The discovery area is the Dark Ages as a cosmological probe. These items and their parts are summarized in Table C.1 at the end of this appendix. Table C.2 presents the highest profile yet-unfunded capabilities needed to address the cosmology science questions and discovery area. These questions build on the successful framework of the standard cosmological model to search for distinctive signatures from the dark sector, the early universe, the cosmic expansion history, the gravitational wave window, and the Dark Ages, all of which can reveal rich new phenomena in realistic physical theories. Through the exquisite experiments, observations, and computations now possible, we can explore domains in energy, space, and time previously inaccessible yet critical to understanding our place in the universe.

TABLE C.1 Summary of Science Questions

TABLE C.2 Capabilities and Future Needs to Address the Cosmology Science Questions and Discovery Area

NOTES: This table of capabilities focuses on the highest profile yet-unfunded items, echoed from the summaries at the end of each section. Other capabilities are listed in the text. Needs provided by facilities existing or currently under construction are not included. Some capabilities are described in terms of the observational goal, agnostic to wavelength, as there are multiple plausible paths. The panel stresses that neither these capabilities nor the science questions are presented in priority order.