Upcoming and Previous Seminars (Past months or Previous years)

Jul 30, Thursday
4 pm (LBL RPM)
Jose Bernal, IFCA
LBL 50-5132 and ZOOM
Line-Intensity Mapping: progress and lessons learned
Line-Intensity mapping (LIM) uses the integrated flux along the line of sight as tracer of the LSS. Using relatively low-aperture telescopes, it recovers radial information targeting known spectral lines discarding the continuum emission, which makes the experiments cheaper compared with galaxy surveys. Mapping the intensity fluctuations of an array of lines from HI 21cm to optical-UV lines offers a unique opportunity to probe redshifts well beyond the reach of other cosmological observations, access regimes that cannot be explored otherwise, and exploit the enormous potential of cross-correlations with other measurements. This promises to deepen our understanding of various questions related to galaxy formation and evolution, cosmology, and fundamental physics. In this talk I will cover the current status of LIM, focusing on recent developments of the HETDEX and MeerKLASS collaborations and the lessons learned in this pathfinder stage that must be applied for future, larger surveys.

Sep 1, Tuesday
1:10 pm (BCCP/Cosmology seminar)
1 minute intro slides
Campbell 131

Sep 8, Tuesday
1:10 pm (BCCP/Cosmology seminar)
Sam Goldstein, Columbia
Campbell 131
Squeezing New Physics out of Cosmological Surveys: From Axions to Cosmological Colliders
Cosmic microwave background (CMB) and large-scale structure (LSS) surveys offer an unprecedented opportunity to search for physics beyond the Standard Model. Fully realizing their potential, however, requires new methods to extract subtle signals from increasingly complex cosmological datasets, where astrophysical foregrounds, nonlinear structure formation, and observational systematics can obscure or mimic the signatures of interest. In the first half of the talk, I will describe the resonant conversion of CMB photons into axions in the magnetized plasma within galactic halos. This effect modifies the blackbody spectrum of the CMB and can be isolated using multifrequency cleaning techniques. I will present results from the first search for this effect, using unWISE galaxies and Planck CMB maps, with a particular focus on foreground mitigation strategies. I will then present results from a follow-up search that incorporates ACT data via a new component-separation pipeline, yielding significantly improved constraints on the axion-photon coupling, and comment on its application to future Simons Observatory data. In the second half of the talk, I will turn to galaxy surveys as a powerful probe of inflation, focusing on signatures of ultra-high-energy particles accessible via the cosmological collider. I will present new methods for extracting these signatures with LSS data and conclude with the first LSS constraint on tauNL, the amplitude of the primordial four-point function generated in multi-field inflation models, via its imprint on scale-dependent bias. This analysis, which uses DESI DR2 quasars, constitutes the largest-scale three-dimensional power-spectrum measurement to date, underscoring the unique sensitivity of galaxy clustering to fundamental physics while highlighting several challenges of using LSS observations on ultra-large scales. Together, these results demonstrate the potential of ongoing and upcoming CMB and LSS surveys to probe physics beyond the Standard Model.

Sep 15, Tuesday
1:10 pm (BCCP/Cosmology seminar)
Carmen Embil-Villagra, Cambridge
Campbell 131
Unveiling the Late-Time Universe with CMB Secondary Anisotropies
As cosmic microwave background (CMB) observations achieve higher angular resolution, secondary anisotropies, generated by the interaction of CMB photons with the intervening matter, are becoming some of the richest science targets of current and future surveys. Since these secondary signals predominantly appear on small angular scales, high-resolution CMB experiments are uniquely positioned to probe in detail the imprint of large-scale structure (LSS) on the CMB, providing a powerful complementary tracer to traditional LSS surveys.
In this talk, I will describe my work extracting cosmological information from the late-time Universe using CMB secondary anisotropies and their cross-correlations with large-scale structure surveys. I will focus on measurements of the large-scale velocity field using the kinetic Sunyaev–Zel'dovich effect, including the development of a foreground-immune velocity reconstruction estimator and its application to data. I will then present measurements of structure growth at high redshift through the cross-correlation of CMB lensing and the Quaia quasar catalogue. Finally, I will briefly discuss ongoing work using machine learning to recover non-Gaussian cosmological information from the thermal Sunyaev–Zel'dovich field.

Sep 22, Tuesday
1:10 pm (BCCP/Cosmology seminar)
Kritti Sharma, Caltech
Campbell 131
Backlighting the Cosmic Web with Fast Radio Bursts: First Measurements and Constraints
The dispersion measures (DMs) of fast radio bursts (FRBs) have emerged as a powerful new probe of cosmic baryons, offering a key advantage over traditional methods: they are largely unbiased with respect to gas density and temperature. In this talk, I will present a unified view of recent advances in FRB cosmology, from feedback inference using the DM-redshift relation to cross-correlations with large-scale structure. Using a sample of localized FRBs with robust host associations, we infer the baryon distribution across galaxy groups and cluster-scale halos in the local Universe - measurements that are complementary to the thermal and kinematic Sunyaev-Zel'dovich (tSZ/kSZ) effects. Extending this approach, cross-correlations of FRB DMs with a wide range of large-scale structure tracers map baryon over- and under-densities along FRB sightlines, establishing FRBs as effective backlights of the cosmic web. Looking ahead, next-generation FRB experiments - most notably the Deep Synoptic Array - will deliver orders-of-magnitude larger localized samples, sharply improving both DM-redshift inference and cross-correlation measurements, and enabling precision constraints on baryonic feedback and its evolution with cosmic time. These measurements will be complementary to SZ measurements from the Simons Observatory and will demonstrate powerful synergies with upcoming large-scale structure surveys such as Rubin and Roman, jointly disentangling the imprint of feedback from that of the underlying cosmology.
Relevant papers:

1. https://arxiv.org/abs/2604.17162
2. https://arxiv.org/abs/2604.22105
3. https://arxiv.org/abs/2605.01994

Sep 29, Tuesday
1:10 pm (BCCP/Cosmology seminar)
Lurdes Ondaro, Cambridge
Campbell 131
Probing baryonic feedback with stacked kinetic Sunyaev-Zeldovich effect
Understanding how baryons trace the underlying matter distribution remains one of the main challenges in precision cosmology. Recent high signal-to-noise measurements of the kinematic Sunyaev-Zeldovich (kSZ) effect now provide a powerful way to probe these diffuse baryons directly, and appear to imply very strong baryonic feedback. However, it is still unclear whether the theoretical modelling used so far is sufficiently detailed to describe the signal accurately.
In this talk, I will present a comprehensive analysis of mock stacked kSZ measurements on realistic DESI-like galaxy samples based on hydrodynamic simulations that include non-linear physics and complex astrophysical processes. By separating the signal into its different components and isolating the processes that affect the density and velocity fields, I will identify the physical mechanisms that contribute to the kSZ signal and therefore need to be included in an accurate model. In particular, I will discuss the baryonic effects on density and velocity fields, the impact of the satellites in the galaxy sample, non-linear effects in the velocity field, and velocity reconstruction. I will show that at the precision of current data, ignoring satellites in the modeling biases low the inferred halo gas fractions; while with upcoming data, several other higher-order effects will become statistically significant. Accounting for these effects will be essential, given the precision of upcoming measurements, to extract reliable information about the gas distribution and robustly assess the strength of baryonic feedback in the Universe.

Oct 6, Tuesday
1:10 pm (BCCP/Cosmology seminar)
Jared Siegel, Princeton
Campbell 131
Astrophysical barriers to small-scale cosmology
It remains uncertain how baryonic feedback processes modify the distribution and thermodynamics of gas on Mpc scales. This is a key open question in galaxy evolution, and it is now a limiting systematic for small-scale weak lensing cosmology. I will review constraints on baryonic feedback from a joint analysis of the kinetic Sunyaev-Zel'dovich (kSZ) effect, X-ray gas mass fractions, and galaxy-galaxy lensing, which together reveal evidence of efficient gas depletion from group mass halos. When combined with direct measurements of galaxy intrinsic alignments with DESI, this data-driven prior on baryonic feedback significantly improves cosmological constraining power on small scales. To reveal the mechanism of efficient gas depletion and calibrate probes of small-scale structure with the thermal Sunyaev-Zel'dovich (tSZ) effect, I will present new measurements of the tSZ effect around the DESI LRG sample, directly modeling the effects of dust and radio emission. Across group mass halos at 0.5 < z < 1.0, hydrodynamical simulations significantly overpredict the observed tSZ signal within approximately twice the virial radius. This discrepancy is difficult to explain with gas depletion alone, and may instead point toward missing non-thermal pressure support or significant departures from hydrostatic equilibrium.

Oct 12, Monday
12:10 (TAC seminar)
Tara Dacunha, Stanford
Campbell 131
Enabling precision inference on dark matter with robust subhalo population modeling
The small scale structure of dark matter is an emerging frontier for precision cosmological inference on dark matter physics. The near-field universe can provide powerful constraints through tracers of the underlying population of low mass dark matter “subhalos”: faint dwarf galaxies, gravitationally-impacted cold stellar streams, perturbed strong lenses, and annihilation/decay flux sources. However, in order to turn these tracer populations into robust dark matter constraints, it is necessary to model the connection between the theoretically understood population of dark matter halos falling into the potential well of a galaxy and the uncertain tidally evolved subhalo population probed by these observables. Any constraint on deviation from the standard cold dark matter (CDM) paradigm implies a thorough and precise understanding of what CDM predicts at these small nonlinear scales, an understanding that currently lacks consensus and remains rife with unquantified modeling uncertainties. We present SubMERGE, a physically motivated analytical subhalo evolution model tested on cosmological simulations with and without disks, accounting for the numerical limits of the simulations and propagating learned uncertainties. This model builds on our new measurements of a universal profile shape for tidally evolving subhalos across simulations. Applying the model to cosmological simulations allows us to leverage their realism to generate various realizations of subhalo populations by modeling numerically unreliable subhalos. Crucially, without running new simulations, SubMERGE can flexibly explore the impact of altering tidal mass loss for different halo or disk growth histories as well as different density profiles for cases of dark matter self-interactions or baryonic effects. SUBMERGE has broad applicability to analyses that require fast, flexible, and realistic evolved dark matter subhalo populations. We present predictions for present day subhalo properties of Milky Way-mass hosts with implications for dark matter inference with dwarf galaxies, stellar streams, strong lenses, and annihilation/decay searches.

Oct 13, Tuesday
1:10 pm (BCCP/Cosmology seminar)
Sofia Chiarenza, Waterloo
Campbell 131
No Mode Left Behind: combining experiments to measure inflation
Primordial non-Gaussianity is one of the cleanest windows into the physics of inflation: a detection of local f_ NL beyond the single-field expectation would rule out entire classes of models. Large-scale structure offers a route to this signal through the scale-dependent bias of galaxies and quasars on the largest scales, but those scales are also where imaging systematics are hardest to control. Cross-correlating spectroscopic tracers with CMB lensing provides a complementary way to access these scales, using a lensing map that is independent of the imaging systematics affecting the galaxy sample. In this talk, I’ll present constraints on f_NL from DESI, starting with the DR1 quasar x Planck PR4 lensing analysis and moving to the ongoing DR2 analysis, which extends to LRGs and includes auto-correlations. I’ll also describe Blast.jl, a differentiable framework that makes accurate angular power spectrum inference fast enough for these analyses.

Oct 20, Tuesday
1:10 pm (BCCP/Cosmology seminar)
Cliff Burgess, McMaster and Perimeter
Campbell 131
Thinking Effectively About the Dark
We live in remarkable times: our Core Theory (the Standard Model plus General Relativity) describes very accurately almost all observations that have been made. Yet there are a handful of observations it does NOT describe - so it cannot be the entire story - but these do not yet determine what the correct alternative must be. Though we have plausible candidates for physics at the highest energies that reconciles General Relativity with Quantum Mechanics, the phenomenon of decoupling usually obstructs using this information in practice. In this talk I argue why a few UV features can penetrate the fog of decoupling in interesting ways and these seem to play important roles in what we see in cosmology. The ideas to which one is led point to a rich (possibly supersymmetric) dark sector and a class of late-universe cosmologies that are relatively poorly explored (but which seem to help with recent cosmological puzzles despite not being designed to do so). Theories likely also have many other implications for terrestrial observations, at colliders and underground experiments and for precision tests of gravity and particle properties.

Oct 27, Tuesday
1:10 pm (BCCP/Cosmology seminar)
Kai Lehman, LMU
Campbell 131
Beyond the Explicit Likelihood: Novel Methods for Simulation-Based Cosmology
Modern cosmology has presented us with an immense volume of data, but we still lack the ability to take full advantage of it. Some current results hint at physics beyond the standard model, but we are not yet able to draw definitive conclusions. To resolve these questions, we need to exploit the full statistical power of our datasets with increasingly sophisticated models of complex physical and observational effects. While such models can often be simulated numerically, their likelihoods may be intractable to evaluate. Simulation-based inference (SBI) allows us to perform inference in this regime by using neural networks to learn the quantities needed for inference directly from simulations. Yet SBI also presents a new set of challenges: How do we retain a physical understanding of our inference pipelines when neural networks take over much of the inference machinery? How can we compare different models when the likelihood cannot be evaluated? I will present new methods addressing both challenges. First, I will show how physically informed machine-learning architectures can incorporate domain knowledge directly into SBI pipelines, retaining a connection between the learned inference and the underlying physics. Second, I will introduce novel methods for Bayesian model inference on the basis of the evidence without evaluating the likelihood. I will demonstrate the latter on DESI DR2 data, enabling rapid comparison of a range of cosmological models. Together, these approaches exploit the advantages of SBI and enable analyses that are intractable with traditional likelihood-based methods.

Nov 2, Monday
4:15 pm (Physics Colloquium)
Chung-Pei Ma, Berkeley
Lecture Hall #1, Physics North
Gravity's Fatal Attraction: Black Holes at the Centers of Galaxies
Black holes are spectacular end products of the fatal attraction of gravity. Residing at the centers of galaxies, supermassive black holes have masses up to 20 billion suns and directly impact the evolution of their host galaxies. I will highlight recent progress in studying black holes in the giant elliptical galaxy M87 and other massive galaxies enabled by integral-field spectroscopy on the James Webb Space Telescope and the Keck Telescopes. I will also discuss the implications of these results for the low-frequency gravitational waves from merging supermassive black hole binaries probed by the ongoing Pulsar Timing Array experiments.

Nov 3, Tuesday
1:10 pm (BCCP/Cosmology seminar)
Daniel Gilman, Chicago
Campbell 131
Inferences on dark matter properties from the JWST lensed quasar dark matter survey
Strong gravitational lensing provides a direct probe of dark matter halos across cosmological distances, irrespective of whether halos contain enough baryons to form stars. JWST has recently surveyed a sample of 29 quadruply imaged quasars, systems that probe (sub)halo populations down to 10^6 solar masses. I will present measurements of the abundance and internal structure of dark matter (sub)halos from these data, and discuss their interpretation in terms of the fundamental nature of dark matter. In particular, I will present constraints on the free-streaming length of warm dark matter, and on the minimum mass of cold dark matter halos consistent with the data. I will also show new results suggesting that low-mass halos have central densities consistent with gravothermal core collapse, a key prediction of self-interacting dark matter (SIDM), and the interpretation of these results in terms of SIDM models with long-range dark forces. I will conclude with the outlook for strong lensing inferences on fundamental physics, which will soon be revolutionized by large cosmological surveys forecast to discover tens of thousands of new lens systems, including hundreds of quadruply imaged quasars.

Nov 6, Friday
1 pm (CMB group lunch)
Martine Lokken, IFAE
LBL 50A-5132 (Sessler Room)
"When gas goes beyond: exploring the power of oriented SZ stacking"

Nov 10, Tuesday
1:10 pm (BCCP/Cosmology seminar)
Chenze Dong, IPMU
Campbell 131
Tracking Baryon Redistribution Beyond Galactic Halos Across Cosmic Time with Field-Level Environments
Although feedback is central to models of galaxy evolution, its effects on the temperature and distribution of gas beyond galactic halos remain difficult to measure. A key challenge is separating such signatures from the effects of the underlying environment. In this talk, I will discuss how combining gas observations with reconstructed cosmic environments can help address this problem. I will begin at Cosmic Noon with COSTCO-I, a massive protocluster at z \sim 2.3 that exhibits unexpectedly weak Ly\alpha absorption compared with constrained simulations. Comparisons with hydrodynamical simulations favor strong kinetic AGN feedback over gravitational heating alone as an explanation for its transparency. I will discuss the implications for heating intergalactic gas and the motivation for complementary simulations of AGN jets. I will then turn to low-redshift baryon redistribution, using fast radio burst dispersion measures and foreground galaxy spectroscopy from FLIMFLAM. I will introduce the basic rationale of the FLIMFLAM project, then present the density reconstruction framework I built for the foreground mapping test. Alongside simulation studies of environment-dependent baryon redistribution, these reconstructions provide a basis for interpreting FRB sightlines and testing feedback models. Finally, I will outline future work connecting galaxy properties, multi-tracer reconstructions, and differentiable gas models at different redshifts.

Nov 17, Tuesday
1:10 pm (BCCP/Cosmology seminar)
held,
Campbell 131

Nov 24, Tuesday
No Talk

Dec 1, Tuesday
1:10 pm (BCCP/Cosmology seminar)
held,
Campbell 131