Joel Meyers
Associate Professor of Physics, Southern Methodist University

Research

My research is organized around three themes: neutrinos and the dark sector, enhancing the value of CMB surveys, and gravitational waves across frequencies. Much of this work uses the cosmic microwave background, either alone or in combination with other data sets including galaxy surveys, primordial light element abundances, and gravitational wave observations.

You can see my publications on INSPIRE.

Neutrinos and the Dark Sector

The Cosmological Preference for Negative Neutrino Mass

The inferred sum of neutrino masses, with negative values allowed: CMB data alone and CMB with DESI both prefer a value below zero, below the minimum allowed by neutrino oscillations

Neutrinos have mass, and massive neutrinos change both the expansion history of the universe and the growth of structure, which makes cosmology sensitive to the sum of their masses. Surprisingly, the most precise current data from the cosmic microwave background and baryon acoustic oscillations prefer a value below zero, which is unphysical for ordinary neutrinos and may be a hint of new physics. With collaborators, I have shown how this preference arises in the data and explored physical mechanisms that could explain it, including long-range interactions in the dark matter, primordial non-Gaussianity, and non-standard neutrino interactions. My group is now testing this further: Ph.D. student Joshua Perez is studying whether more general primordial fluctuations alter the preference, and Christopher Cook, now a Ph.D. student at UT Austin, and I are studying how early dark energy affects the cosmological measurement of neutrino mass.

Light Relics and Dark Sector Physics

Light relics are light particles that were produced abundantly in the hot early universe, like the three species of neutrinos of the Standard Model. The neutrinos fell out of equilibrium about one second after the Big Bang, but they remain extremely numerous today, making up the cosmic neutrino background. The gravitational influence of light relics leaves observable signatures in the cosmic microwave background and large-scale structure, so measuring their density probes neutrino physics and very weakly interacting particles beyond the Standard Model. I showed that free-streaming radiation shifts the phase of the acoustic peaks in the CMB, a robust signature that has since been measured in CMB and DESI BAO data. I have also studied how comparing the light relic density measured in the CMB with the abundances of light elements produced in big bang nucleosynthesis tests new physics in the early universe. Simons Observatory will substantially improve these measurements, sharpening tests of dark sector physics and of whether light relics play a role in the preference for negative neutrino mass.

Enhancing the Value of CMB Surveys

CMB Delensing

Gravitational lensing by the large-scale structure between us and the surface of last scattering distorts our view of the primary CMB fluctuations. Lensing smooths the acoustic peaks, converts E-mode polarization into B modes, and introduces correlations between scales, all of which weaken constraints on the parameters best measured through the primary CMB. Delensing reverses these effects: the lensing deflection is reconstructed from the CMB maps themselves, or traced with other observations of large-scale structure, and its effects are then removed. Removing lensing B modes is essential in the search for primordial gravitational waves, and delensing the temperature and E-mode polarization sharpens measurements of light relics, of inflation, and of models proposed to resolve the Hubble tension. Because delensing uses the same data a survey collects anyway, it comes at no additional experimental cost; in some cases, the improvement is comparable to halving the noise level of the experiment. Delensing of temperature and E modes has now been carried out on real data by ACT and SPT-3G.

New Probes of Cosmic Structure

Illustration of the moving lens effect: several galaxy clusters moving across the sky in different directions, each leaving a temperature dipole, cooler ahead of the cluster and hotter behind

Lensing is also challenging to measure on the smallest angular scales: at multipoles above a few thousand, standard lensing reconstruction methods become suboptimal for low-noise surveys. The Small Correlated Against Large Estimator (SCALE) instead measures how lensing ties the local small-scale power of the CMB temperature to the large-scale temperature gradient. For upcoming surveys, SCALE can outperform standard methods, and the small-scale lensing amplitude it measures is sensitive to neutrino mass and to models such as warm or fuzzy dark matter that suppress structure on small scales.

The CMB also carries other imprints of the structure it passes on its way to us. When a massive structure such as a galaxy cluster moves across our line of sight, its lensing of the CMB leaves a small temperature dipole aligned with its motion: cooler ahead of the cluster and hotter behind. Combined with galaxy surveys, this moving lens effect gives access to the transverse velocities of cosmological structures, and together with the kinetic Sunyaev-Zel’dovich effect it could map the three-dimensional velocity field of the universe. The moving lens effect has now been detected for the first time, at 4.8σ, in ACT and DESI Legacy Surveys data.

Just after recombination, CMB photons also scatter off newly formed neutral hydrogen and helium. This Rayleigh scattering grows steeply with frequency (as ν4), giving a second, partly independent view of the last-scattering surface that can significantly improve constraints on light relics and the sum of neutrino masses. Atmospheric effects and astrophysical foregrounds make it hard to detect with experiments planned for this decade, but a future CMB satellite should be able to measure it with high significance.

Machine Learning Tools

Architecture of the ResUNet-CMB network: maps of CMB polarization (Q and U) pass through convolutional and residual blocks, and the network outputs maps of cosmic polarization rotation and lensing convergence, among others

Machine learning offers ways to extract information from CMB maps that standard estimators leave behind. ResUNet-CMB, a convolutional neural network, takes maps of CMB polarization and simultaneously reconstructs gravitational lensing, patchy reionization, and cosmic polarization rotation, along with the primary CMB fluctuations. It outperforms the standard quadratic estimator and matches the performance expected from iterative reconstruction methods. Applied to the search for primordial gravitational waves, the same network can estimate and remove several sources of secondary B-mode polarization at once, and its output can be used in a likelihood analysis to produce nearly optimal, unbiased estimates of the amplitude of primordial gravitational waves.

Gravitational Waves Across Frequencies

The gravitational-wave landscape from 10^-19 to 10^4 hertz: the spectrum of gravitational waves from inflation at today's bound, the signal seen by pulsar timing arrays, the LIGO-Virgo-KAGRA limit and the planned LISA sensitivity, and the frequency ranges covered by CMB B modes, pulsar timing arrays, CMB time-dependent deflection, a gravitational-wave timing array, LISA, and LIGO-Virgo-KAGRA

Gravitational waves offer a window onto fundamental physics across an enormous range of frequencies. At the lowest frequencies, gravitational waves produced during cosmic inflation would leave a distinctive pattern of B-mode polarization in the CMB. Detecting it would reveal the energy scale of inflation, but the signal is faint and lies beneath the B modes produced by gravitational lensing, which makes delensing essential. B-mode measurements from Simons Observatory, together with constraints on primordial non-Gaussianity, will sharpen tests of the physics that drove inflation.

At nanohertz frequencies, pulsar timing arrays have found evidence for a background of gravitational waves. Supermassive black hole binaries are the expected source, but the signal could also come from the early universe: for example, gravitational waves from inflation, amplified by a period of kination in which the expansion was dominated by the kinetic energy of a scalar field, can also fit current data. Gravitational waves can also be found in data collected for other purposes. A gravitational wave passing the Earth shifts the apparent positions of distant sources in a correlated, time-dependent pattern, and time-dependent deflection reconstruction searches for this pattern in CMB maps, extending sensitivity to frequencies both below and above those covered by pulsar timing arrays. Precise timing of strongly lensed repeating fast radio bursts, or of the continuous gravitational waves from Galactic binaries seen by a space-based detector like LISA, can likewise reveal long-wavelength gravitational waves.

Collaborations

I am a member of the Simons Observatory, CCAT-prime, CMB-HD, and PICO collaborations. I was a member of CMB-S4 from 2015 until the collaboration wound down, serving as Co-Chair of its Science Council from 2020 to 2024.