January 2022
Jan 14, Friday
12 Noon (INPA, LBL)
Matthew Ho, CMU
By ZOOM
Galaxy Cluster Mass Estimation Using Deep Learning
In this talk, I will discuss how we use modern deep learning models to infer galaxy cluster masses with high precision, reliable uncertainty, and computational efficiency. I will describe our work in using Convolutional Neural Networks (CNNs) to mitigate systematics in the virial scaling relation to produce dynamical mass estimates of galaxy clusters, using projected galaxies, with remarkably low bias and scatter. I will also discuss how we can recover and empirically verify Bayesian uncertainties on deep learning mass predictions using variational weight distributions. I will describe how we've validated our methods on real observational systems like the Coma, CLASH, and HeCS clusters as well as projections for how we can use these models to study cluster cosmology using data from current and upcoming sky surveys. Lastly, I will mention results from our ongoing work on combining multi-wavelength observables to produce fully informed observational probes of cluster dark matter.
Jan 18, Tuesday
1:10 pm (BCCP/Cosmology Seminar)
Raul Monsalve, UCB
By ZOOM
Studying the Early Universe with Radio Measurements of the Global 21-cm Signal
In this talk I will discuss how we can study the first billion of the Universe, specifically the Dark Ages, Cosmic Dawn, and Epoch of Reionization, using the sky-averaged 21-cm signal emitted by neutral hydrogen in the intergalactic medium. I will first describe some of the connections between this signal, expected at radio frequencies below 300 MHz, and the properties of the first stars, galaxies, and black holes, as well as the potential to use it to constrain dark matter and exotic physics in the early Universe. Then, I will describe two ongoing ground-based experiments in which I participate, EDGES and MIST, which attempt to measure this signal from very remote radio-quiet sites. I will discuss their design, characterization in the lab, performance in the field, scientific outcomes to date, and projections for the near future. Finally, I will discuss ideas and current concepts to measure this signal from the pristine environment of space, motivated by the significant challenges associated with observations from the ground.
Jan 21, Friday
12 noon (INPA, LBL)
Lingyuan Ji, JHU
By ZOOM
Cosmological Neutrino Perturbations Without The Boltzmann Hierarchy
We present a formulation of cosmological perturbation theory where the Boltzmann hierarchies that evolve the neutrino phase-space distributions are replaced by integrals that can be evaluated easily with Fast Fourier transforms. The simultaneous evaluation of these integrals combined with the differential equations for the rest of the system (dark matter, photons, baryons) are then solved with an iterative scheme that converges quickly. The formulation is particularly powerful for massive neutrinos, where the phase space is three-dimensional rather than two-dimensional, and even more so for three different neutrino mass eigenstates. Therefore, it has the potential to significantly speed up the computation times of cosmological perturbation calculations. This approach should also be applicable to models with other non-cold collisionless relics.
Jan 25, Tuesday
1:10 pm (BCCP/Cosmology Seminar)
Andrina Nicola, Princeton
By ZOOM
Cosmological probe combination for current and future surveys
Recent progress in observational cosmology and the establishment of ΛCDM have relied on the combination of different cosmological probes.
These probes are not independent, and their cross-correlations allow for robust tests of the cosmological model and constraining theoretical and
observational systematics, making them a promising analysis method for both current and future data.
In this talk, I will outline possible ways in which to extend joint analyses to optimally benefit from upcoming data, and present results on photometric
galaxy clustering in the Hyper Suprime Cam DR1 data. I will then discuss the potential of using two-point function statistics of baryon tracers of the
matter density to robustly constrain both cosmology and baryonic feedback.
Jan 25, Tuesday
4 pm (special LBL seminar)
Paulo Montero Camacho, Tsinghua University
By ZOOM
Unveiling the astrophysics of reionization in the post-reionization era
Cosmic reionization is the last major phase transition that our Universe goes through. In this milestone, the intergalactic medium transitions from primarily neutral, dark, and cold into mainly ionized, luminous, and warm. However, as exciting as the reionization process surely was, many aspects remain unknown, e.g. what are the sources of the ionizing photons? what is the timeline of reionization? Upcoming telescopes (e.g. JWST and HERA) will help to directly probe cosmic reionization. Nonetheless, interpretation of said measurements will strongly rely on modeling the surroundings of luminous objects (JWST) or on foreground removal (21 cm). In this talk, I will describe a promising novel method to constrain reionization through its impact in the post-reionization intergalactic medium. I will focus on the ability of the Lyman-alpha forest (using DESI) to unveil the astrophysics of reionization in the first part of my talk. In the second part, I will describe how we can further exploit this new methodology through cross-correlations of different cosmological probes of the post-reionization era. Furthermore, I will showcase the potential of this new methodology to constrain the nature of dark matter via its exquisite small-scale sensitivity.
Jan 26, Wednesday
6 pm (special LBL seminar)
Priyanka Jalan, ARIES, India
By ZOOM
Probing the physical state of the Intergalactic-Medium and Quasar environments
The proximity effect in closely separated quasars pairs is one of the unique tools to probe the environment of quasars both in the transverse and the longitudinal directions. Therefore, we used a large sample of 181 closely separated quasars pairs (with a separation smaller than 1.5 arc-minutes) to study the proximity effect both in the longitudinal as well as in the transverse directions of foreground quasars, by carrying out a statistical comparison of the Lyman-alpha absorption lines in the vicinity of quasars to those seen far away in the general Intergalactic-Medium. We have constrained the density profile and found a significant difference in the longitudinal and the transverse directions. Such observed discrepancy among these two directions has many interesting interpretations such as anisotropic radiation, finite episodic light time of AGN etcetera. Among them, our analysis shows that the anisotropic obscuration seems to be the most plausible. Throughout our analysis, any systematic observational biases have been eliminated by using the observed control data sample, while the detail simulation has been used in the validation of our analysis methodology. We also extend this technique on high quality (SNR ~30, R~5100) dataset of 88 quasars at z~4, by studying the longitudinal proximity effect, to constrain their density profile. Furthermore, we also investigated the main parameters which can possibly affect this overdensity profile using high-resolution (R~40000) spectra. In addition to the available archival data, for further improvements, we also used long-slit spectroscopy at 2-4m class telescope to discover 10 new quasar pairs even at angular separation smaller than 1 arcmin, where both the proximity effect and over-density must be profound. In this search, we have used the available photometric catalogue to list the high confidence probable AGN pairs candidates, and then carry out their long-slit spectroscopy using the 2m class telescope. Moreover, the discovery of gravitationally lensed quasars will be also discussed.
Jan 28, Friday
9 am (special LBL seminar)
Roger de Belsunce , Kavli Institute for Cosmology, Cambridge
By ZOOM
Cosmological inference from CMB & Lyman-a forest
The CMB and Lyman-a forest have provided us with exquisite constraints on the standard model of cosmology. Upcoming data from the DESI survey, will provide us with unprecedented statistics on dark matter clustering, containing imprints from neutrinos created in the early Universe on small scales and from dark energy driving the expansion of the Universe on large scales. Extracting the full potential of this data set is an ongoing challenge because of its size and highly non-linear nature.
In this seminar, I will discuss methods to make inference from the CMB and Lyman-a forest and how to cross-correlate both. This allows us to probe large scales with the CMB, small scales with the one-dimensional Lyman-a power spectrum (P1D), and intermediate scales from their cross-correlation. However, the Lya-P1D is sensitive to small-scale astrophysical effects and fails at jointly extracting large- and small-scale information because of the amplitude-slope degeneracy. As an outlook, I will present exciting opportunities with the measurement of the three-dimensional Lyman-a power spectrum (P3D) and how we can access the information stored in the broadband shape.
February 2022
Feb 1, Tuesday
1:10 pm (BCCP/Cosmology Seminar)
Corentin Cadiou, UCL
By ZOOM
On the causal origin of properties of dark matter halos and galaxies
(also presentation with videos here)
Large-cosmological simulations, such as Illustris-TNG or Horizon-AGN, allows to draw relations between the formation history of galaxies and their host haloes (number of mergers, formation time) and their properties (galaxy morphology, halo concentration). These relations are however statistical in nature, as no two galaxies share the same formation history. In this seminar, I will present an alternative approach to draw causal links instead of statistical ones. Indeed, by carefully modifying the initial conditions of cosmological simulations, one can control the subsequent formation of halos and galaxies in the simulation, thus allowing one to conduct controlled numerical experiments. Using this approach, I will present fundamental results on the origin of the angular momentum and concentration parameter of dark matter halos, before presenting preliminary results on the properties of galaxies, and in particular their spin.
Feb 8, Tuesday
1:10 pm (BCCP/Cosmology Seminar)
Dan Green, San Diego
By ZOOM
Non-Gaussianity in Maps of the Large Scale Structure
Equilateral non-Gaussianity is a compelling observational target for CMB and LSS surveys. Unfortunately, progress in measuring non-Gaussianity is expected to be limited by our ability to accurately model the distribution of galaxies in the universe. In this talk, I will discuss how these challenges could be overcome using the full maps of large scale structure. I will explain how the special nature of the equilateral signal protects it from being corrupted by nonlinear evolution. This suggests that a map level approach to LSS could dramatically impact the search for primordial non-Gaussianity.
Feb 15, Tuesday
1:10 pm (BCCP/Cosmology Seminar)
Anthony Pullen, NYU
By ZOOM
A galaxy property census with line intensity mapping
Uncovering the physics of galaxy evolution has been a longstanding problem for astronomers. Physical galaxy properties like the star formation rate, stellar mass, and metallicity can tell us not only how these properties change over generations of galaxies, but they also give us a window into the conditions of the universe at the time of galaxy formation. Galaxy surveys, which are predominantly used to measure these properties, only observe the brightest galaxies, which are a biased sample. Line intensity mapping (LIM) observes the aggregate emission from brighter and fainter galaxies over much larger volumes. The EXCLAIM survey is a pathfinder for this technique, and has the potential to aid in constructing a full census of galaxy emission while minimizing sample variance. In this talk I will present how we can use LIM surveys to construct a galaxy property census. Specifically, I will discuss currently planned methods using LIM to uncover galaxy properties like molecular hydrogen density, star formation, and metallicity properties, as well as their limitations. I will also discuss potential ways we could get around these hurdles using hydrodynamic simulations along with semi-analytic star formation models. Finally we will discuss how these methods could be applied to EXCLAIM and other upcoming LIM surveys.
Feb 22, Tuesday
1:10 pm (BCCP/Cosmology Seminar)
Shadab Alam, Edinburgh
By ZOOM
Are we ready for the challenges of Redshift Space Distortion?
I will introduce redshift space distortions, talk about the kind of measurements we typically use and then show some of the measurements from the DESI Survey Validation. I will then highlight that typically models we use for RSD are approximate and assume that the galaxies are a fair sample of the universe independent of environment. I will show some examples of how this is not true and how the details of galaxies can affect non-linear clustering to fairly large scales. Hence there is the need to perform robustness tests of the model. I will then talk about some effort in this direction focusing on star-forming galaxies, and finally comment on how to get to the precision we need for future surveys.
Feb 24, Thursday
4 pm (LBL RPM)
Kirit Karkare, Chicago
By ZOOM
“Pushing the Limits of Cosmology with Next-Generation Millimeter-Wave Telescopes”
While the Lambda-CDM cosmological model is remarkably effective at describing the Universe and its evolution as a whole, foundational questions remain. Where did the primordial fluctuations come from and why were they so uniform? What is causing the present-day accelerated expansion? What is the mass of the neutrino? Maps of large-scale structure at different epochs can address these questions. I will present the latest results from the BICEP/Keck cosmic microwave background experiment, which is searching for B-mode polarization from an inflationary expansion at the earliest moments of time. I will focus on my work calibrating the instrument and controlling systematics, and how current experiments are informing the design of the future CMB-S4 experiment which will test large classes of physically-motivated inflation models. I will then discuss mm-wave line intensity mapping — a new cosmological probe for measuring large-scale structure well into the first billion years of the Universe — and how future instruments using on-chip spectrometers such as SPT-SLIM could transform our understanding of cosmological physics.
March 2022
Mar 1, Tuesday
1:10 pm (BCCP/Cosmology Seminar)
Kfir Blum, Weizmann Institute
By ZOOM
Gravitational lensing measurements of the Hubble parameter: challenges and opportunities
The time delay between the images of gravitationally lensed quasars probes a combination of angular diameter distances and the mass profile of the lens galaxy. Some analyses have reported a determination of the Hubble parameter H0 in tension with determinations based on the cosmic microwave background. A central challenge facing the method is the mass sheet degeneracy (MSD). To mitigate the MSD, lensing analyses must constrain cosmological weak lensing, which can bias the inferred value of H0 at the few percent level. I will show that in current H0 measurements the weak lensing corrections have not been dealt with in full. Another related challenge is due to uncertainties in the intrinsic density profile of the lens, notably the possibility of an extended core feature. I will discuss how an account of the core/MSD effect provides possible resolution for a previous claim of a lensing H0 tension by the TDCOSMO collaboration. The core/MSD effect is also an interesting opportunity, as it could suggest a non-minimal model of dark matter. Based on
2001.07182,2105.10873,
2201.05111.
Mar 3, Thursday
3:30 pm (Astronomy Colloquium)
Seiji Fujimoto, Niels Bohr Inst.
By ZOOM
Linking visible and obscured sides of the early universe
Understanding the galaxy formation and evolution in the Epoch of Reionization (EoR; z>~6) is essential, as they are the key probe for fundamental cosmological questions (e.g., cosmic reionization, large-scale structure formation). In this context, the baryon cycle is one of the most important mechanisms regulating galaxy evolution, but its direct observation is still difficult at EoR. I conduct deep ALMA and HST observations for EoR galaxies and find 10-kpc scale cold carbon gas halos surrounding them. The structure is ~5 times larger than the stellar continuum, evidence of outflow remnants in the EoR galaxies but challenging current galaxy evolution models. To further understand the early baryon cycle and directly address the "formation" of galaxies, I will also present my vision in the next few to 10s years. A unique ensemble of ~70 hrs PI JWST, ALMA, and MUSE programs targeting strongly lensed galaxies at EoR forms a cornerstone of my plan. I will comprehensively characterize 1) chemical enrichment, 2) mass (star, dust, gas, and dark matter) assembly, and 3) kinematics down to physical scales of ~20-100 pc in the EoR lensed galaxies. The insight and experience of JWST x ALMA joint analysis will be smoothly applied to other JWST-observed galaxies at z > 6-12 by leading immediate ALMA follow-up in multiple co-I JWST treasury programs, where I will search for the first emergence of stellar light and dust in the universe. Bringing exclusive data accesses and large collaboration networks of JWST and ALMA to Berkeley, I will lead to a dramatic transformation in our understanding of the primary formation and evolution mechanisms from the first galaxies in the first billion years after the Big Bang.
Mar 4, Friday
12 pm (LBL RPM/INPA)
Chelsea Bartram, Washington
By ZOOM
“Combing the Cosmos: A Deep Dive into the Nature of (Dark) Matter and Fundamental Symmetries”
One of the most striking mysteries in modern astrophysics manifests itself as an absence of matter needed to form a narrative consistent with our current understanding of the universe. A promising candidate for the so-called “dark matter” is the axion, whose wave-like nature requires unique technology to detect. The Axion Dark Matter eXperiment (ADMX) leverages the inverse Primakoff effect to search for axions using an ultra-low noise receiver chain and microwave cavity in a strong magnetic field. If discovered, the axion could not only solve the dark matter problem but also resolve the Strong CP problem. Fundamental symmetries have been at the center of many deep astrophysical questions. As such, they have served as a driving force behind my research, motivating my graduate search for CP- and CPT-violation in the lepton sector, as well as my post-doctoral work in axion dark matter. I present my prior research and explain how it fits into my larger ambition to understand the large-scale structure and composition of the universe with CMB-S4.
Mar 8, Tuesday
1:10 pm (BCCP/Cosmology Seminar)
Carlos Garcia Garcia, Oxford
By ZOOM
The last 10 billion years of cosmic structure growth
The current constraints on the S_8 parameter are subject of debate.
Cosmic shear observations show a lower value than that predicted by Planck. For
instance, KiDS finds results 3\sigma away from Planck's value and data from
DESY1 also points in the same direction. In this talk I will show the data
driven reconstruction of the evolution of the S_8(z) parameter from a
combination of 6 different data sets that include galaxy clustering, weak
lensing and CMB lensing (with DESY1 and KiDS-1000 among them). I will show that
these data constrain the amplitude of fluctuations in the range 0.2 <~ z
<~ 2 and give consistent growth histories. Furthermore, I will show that
in the range 0.2 <~ z <~ 0.7 current data prefer a lower value than
that predicted by Planck and that it is mostly driven by cosmic shear
observations.
Mar 10, Thursday
4 pm (LBL RPM)
Kevin Crowley, UCB
By ZOOM
"The View From Down Here: Ground-Based Measurements of the Cosmic Microwave Background from the Atacama Desert"
The study of the Cosmic Microwave Background (CMB) anisotropies has become a percent-level science constraining the accepted cosmological model ΛCDM. The most powerful measurements thus far have been made from space, but observations from the Earth’s surface can be made at higher resolution, with higher detector count, and with iterative upgrades to the instruments. One key goal of recent upgrades has been recovering the large angular-scale polarization signal that could reveal the signature of inflation in the early universe. The current generation of ground-based telescopes have developed new technologies, including polarization-modulator systems, kilopixel focal plane arrays, and highly-multiplexed detector readout, all with the goal of enabling sensitivity while controlling systematic errors in the challenging environment of polarization measurements under the atmosphere. In this talk, I will describe my efforts in these areas of instrumentation development and my experiences pushing for sensitivity while pondering just what lies below the noise in CMB experiments observing from the Atacama Desert. I will also highlight the ways that future experiments, including the under-construction Simons Observatory and the exciting prospect of CMB-S4, will lay a path toward the ultimate goal of fully exploring the physics in the CMB from terra firma.
Mar 11, Friday
12 noon (INPA)
Marco Gatti, Penn
By ZOOM
Cosmology from Weak Lensing Non-Gaussian Statistics
The statistical analysis of lensed galaxies is a powerful tool to study the dark matter distribution of the Universe. For instance, the distortion of galaxy shapes induced by the large scale structure of the Universe can be used to reconstruct the projected matter density along the line-of-sight (mass maps). Mass maps are useful as they provide a wealth of information that goes beyond and complements the more traditional two-point statistics used in Cosmology.
During this talk, I will present the mass map obtained using the first three years of data (Y3) of the Dark Energy Survey (DES), which is the largest curved-sky galaxy weak lensing mass map to date. I will then show the constraints on cosmological parameters from two independent analyses using non Gaussian statistics applied to the DES Y3 mass maps: moments and peaks. The constraints from these analyses are compatible with and tighter than the ones from the fiducial DES Y3 cosmic shear analysis, which only relies on two-point statistics. These results showcase the potential of non Gaussian statistics in terms of cosmological constraining power and as independent consistency check with ordinary two-point statistics.
Mar 15, Tuesday
1:10 pm (BCCP/Cosmology Seminar)
Naomi Robertson, Cambridge
By ZOOM
Consistent lensing and clustering in a low-S8 Universe with BOSS, DES Year 3, HSC Year 1 and KiDS-1000
I will discuss our recent study were we have evaluated the consistency between lensing and clustering probes of large-scale structure based on measurements of projected galaxy clustering from the Baryon Oscillation Spectroscopic Survey (BOSS) combined with overlapping galaxy–galaxy lensing from three surveys: the Dark Energy Survey Year 3 (DES Y3), the Hyper-Suprime Cam survey (HSC) Year 1, and the Kilo- Degree Survey (KiDS-1000). As part of this work we have investigated small scale systematics in modelling lensing and clustering measurements and how they limit our ability to improve our cosmological constraining power. Also in this work, we have performed an intra-lensing-survey study. I will present joint fits to both the clustering and lensing measurements and show how this analysis demonstrates the statistical power of these small-scale measurements, but also indicates that caution is still warranted given current uncertainties in modelling baryonic effects, assembly bias, and selection effects in the foreground sample.
Mar 22, Tuesday
spring break
Mar 29, Tuesday
1:10 pm (BCCP/Cosmology Seminar)
Pauline Zarrouk, LPNHE Sorbonne
By ZOOM
Precision cosmology at all scales with DESI brightest galaxies
Stage IV Dark Energy Spectroscopic Instrument (DESI) is currently observing a complete magnitude-limited survey of low-redshift galaxies (z < 0.5) called the Bright Galaxy Survey (BGS). The exceptionally high sampling density, low-redshift sample and the wide range of galaxy bias represented in the BGS offer new opportunities for innovative analysis techniques, such as multi-tracer method, small-scale clustering and environment-dependent clustering. In this talk, I will first review the BGS target selection pipeline, their clustering properties and the tests we performed during the Survey Validation period. Then I will present the standard redshift-space 2-pt correlation function (2PCF) analysis we are currently developing and testing using DESI mocks for the Y1 sample (which is know as Key Project 5) and the multi-tracer pipeline I am developing with my PhD student. I will also present an ongoing work in collaboration with my colleagues at Durham that aims at building a flexible neural network-based emulator that reproduces the dark matter halo clustering properties and relative motions produced by gravitational collapse in N-body simulations. Such an emulator is combined with empirical models of the galaxy-halo connection such as HOD in order to obtain accurate predictions for the galaxy 2PCF at small scales and extract unbiased cosmological information from the non-linear regime. This approach can also be extended to more complex galaxy-halo connection models that account for assembly bias and to other summary statistics such as cross-correlation functions between different tracers (multi-tracer) or density environments with the entire redshift space density field (density split).
April 2022
Apr 5, Tuesday
1:10 pm (BCCP/Cosmology Seminar)
Stefania Amodeo, Strasbourg
By ZOOM
Constraining baryonic processes in galaxy halos with the SZ effects
The thermodynamic properties of the ionized baryons in galaxies, groups, and clusters encode the effects of the assembly history and feedback processes that shape structure formation. These properties can be studied through the thermal and kinematic Sunyaev-Zel’dovich (SZ) effects imprinted on high resolution maps of the cosmic microwave background (CMB), measured from stacking analyses of galaxy surveys. I will present recent SZ cross-correlation measurements from the Atacama Cosmology Telescope (ACT) DR5 and Planck, from which we achieve high signal-to-noise measurements of the electron density, temperature and pressure distribution around the CMASS galaxy groups. I will discuss the impact of the modeling uncertainties in the inference of the properties of the intergalactic medium from SZ observations, and I will present our constraints of the baryonic processes like the feedback and the non-thermal pressure support, as well as the effect of including baryons in the modeling of galaxy-galaxy lensing measurements. These measurements provide novel tests of current and future hydrodynamical simulations with sub-grid physics models. I will finally outline the rapid growth of SZ cross-correlation measurements expected over the next decade with upcoming CMB experiments, like Simons Observatory and CMB-S4, and large-scale structure surveys like DESI, the Rubin Observatory or Euclid.
Apr 7, Thursday
3:45 pm (Astronomy Colloquium)
Vanessa Boehm, UCB
By ZOOM/in person
Machine Learning for Astrophysical Data Analysis
The recent rise of artificial intelligence, fuelled by advances in parallel computing on Graphical Processing Units (GPUs), promises to revolutionize more than just automatic image classification. Astrophysicists have good reason to be excited: With datasets ever growing in size and complexity, machine learning promises to be a panacea for tackling the challenge of harvesting information from them. However, machine learning algorithms are often not designed with scientific applications in mind and can struggle with critical tasks such as uncertainty estimation. In my talk, I will explore how we can make the most out of novel machine learning algorithms and related concepts while meeting the needs of rigorous scientific data analysis.
Apr 12, Tuesday
1:10 pm (BCCP/Cosmology Seminar)
Francisco Villaescusa-Navarro, CCA
By ZOOM
The CAMELS project
In this talk I will present and describe the CAMELS project, whose aim is to build bridges between cosmology and galaxy formation by combining numerical simulations and machine learning. Containing a set of 4,233 simulations, both N-body and state-of-the-art hydrodynamic simulations, it is currently the largest dataset of cosmological simulations designed to train artificial intelligence algorithms. I will present some of the results the CAMELS collaboration has obtained recently, such as the finding of an universal relation in subhalo properties, how neural networks can extract cosmological information while marginalizing over baryonic effects at the field level, the first constrain on the mass of the Milky Way and Andromeda from AI, and the prospects of doing cosmology with individual galaxies. I will conclude presenting a potential strategy the scientific community may pursue to extract the maximum amount of information from cosmological surveys, highlighting the multiple challenges associated to it.
Apr 18, Monday
11 am (BIDS ML and Science Forum)
Tomasz Kacprzak, ETH Zurich
By ZOOM
Large Scale Structure Cosmology with Artificial Intelligence
In large scale structure cosmology, the information about the cosmological parameters governing the evolution of the universe is contained in the complex and rich structure of dark matter density field.
To date, this information was probed using simple human-designed statistics, which are not guaranteed, or even expected, to extract the full information content of the data. Recently we proposed an alternative analysis with deep learning that automatically designs the relevant features in order to maximise the information gain from the maps. In our results we observe a significant information gain for the deep analysis as compared to the power spectrum: I will present our latest results on the KiDS-1000 dataset as well as a forecast for combined probes analysis with AI, which we call DeepLSS. I will discuss the sources of the information gain from AI and describe the insights we can learn from it.
Cancelled--Apr 19, Tuesday
1:10 pm (BCCP/Cosmology Seminar)
CANCELED,
By ZOOM
Apr 26, Tuesday
1:10 pm (BCCP/Cosmology Seminar)
Minh Nguyen, Michigan
131 Campbell (not ZOOM)
Bayesian forward modeling of
galaxy clustering: Applications to astrophysics (and cosmology)
The forward modeling of galaxy clustering program aims at modeling the galaxy data directly at the field level. Given a three-dimensional map of galaxies from a galaxy redshift survey, this framework jointly infers the underlying Dark Matter field, and relevant astrophysical-cosmological processes. With high signal-to-noise galaxy surveys like DESI, the information gain offered by the former will be significant for many analyses. In this talk, I will present, as applications of forward modeling, two astrophysical measurements of: 1) the kinematic Sunyaev-Zel'dovich effect and 2) the galaxy intrinsic alignment process. I will discuss in details how these measurements can benefit from the forward modeling framework. Lastly, I will highlight preliminary results for a cosmological application: constraining the clustering amplitude of matter, sigma8.
Apr 28, Thursday
4pm (LBL RPM)
Zeljko Ivezic, Washington
LBL 50A-5132 (ZOOM option available)
"Cosmology with the Greatest Movie of All Time”
The Legacy Survey of Space and Time (LSST), the first project to be undertaken at the new Vera C. Rubin Observatory, will be the most comprehensive optical astronomical survey ever undertaken. Starting in 2024, Rubin Observatory will obtain panoramic images covering the sky visible from its location in Chile every clear night for ten years. The resulting hundreds of petabytes of imaging data, essentially a digital color movie of the night sky, will include about 40 billion stars and galaxies, and will be used for investigations ranging from cataloging dangerous near-Earth asteroids to fundamental physics such as characterization of dark matter and dark energy. I will start the presentation with an overview of LSST science drivers and system design, and continue with a construction status report for the Rubin Observatory. I will conclude with a brief discussion of synergies between Rubin, Euclid, and Roman Space Telescope.
May 2022
May 17, Tuesday
1:10 pm (special seminar)
Ido Ben-Dayan, UCB/Ariel University
131 Campbell Hall
Bouncing Cosmologies with Sourced Gauge fluctuations
Detection of primordial gravitational waves (GW) via measurement of B-mode polarization of the CMB is considered as a proof of Inflation.
Verily, a robust prediction of inflation is a scale invariant GW spectrum as opposed to alternatives that predict a very blue spectrum.
We show that coupling a scalar field to a gauge field in a bouncing cosmology model changes this distinction.
The coupling between the scalar and the gauge field generates additional sourced scalar and GW spectra.
We take into account all effects including the crossing of the bounce.
The predictions match the observed CMB scalar spectrum and predict a chiral GW spectrum that may be observable on CMB and Laser Interferometer scales.
May 31, Tuesday
4 pm (RPM, LBL) 50-5012/ZOOM
David Weinberg, OSU
LBL 50-5012/ ZOOM
Decoding the Origin of the Elements
The elemental abundance patterns of stars encode a wealth of information
about the history of the Milky Way galaxy and the astrophysical processes
that create atomic nuclei. I will describe some of what we have learned
about the origin of elements in the Milky Way from the SDSS APOGEE survey,
which has measured detailed abundance patterns (typically 15-18 elements
per star) for half a million stars. The average chemical enrichment and
the relative contributions from core collapse (massive star) supernovae
and Type Ia (white dwarf) supernovae changes systematically with position
in the Galaxy. However, the observed median trends of element abundance
ratios are nearly independent of position, and the abundance pattern of
a typical APOGEE star can be described with two free parameters and residuals
at the few-percent level. By isolating the enrichment pattern of core
collapse supernovae we can test theoretical models of massive star evolution
and black hole formation. Other elements such as nitrogen and cerium
provide clues to the mixing of nuclear-processed material within intermediate
mass stars. The most important degeneracy in Galactic chemical evolution
models is the tradeoff between the overall yield of heavy elements (sensitive
to black hole formation) and the ejection of elements in galactic winds.
Measurement of deuterium and helium-3 can help break this degeneracy
because these isotopes originate primarily or entirely in the big bang.
|
