November 2017
Nov 1, Wednesday
12 pm LBL lunch
Miguel Zumalacarregui, UCB
LBL 504025
Constraining the abundance of primordial black holes with
magnification of type Ia SNe
Black hole mergers detected by the Laser Interferometer
GravitationalWave Observatory (LIGO) have revived dark matter models
based on primordial black holes (PBH) or other massive compact halo
objects (MACHO) with masses in the range 1100 M_Sun, where previous
bounds were the weakest. I will present new constraints on the PBH
abundance and mass using the gravitational lensing magnification of
type Ia supernovae (Sne) using current datasets (JLA, Union 2.1). Our
results rule out the hypothesis of MACHO/PBH comprising the totality
of the dark matter at high significance in the mass range M > 0.01
M_Sun. Eliminating the possibility of a LIGOmass MACHO further
strengthens the case for microscopic dark matter.
Nov 6, Monday
12:10 pm (TAC)
Brian O'Shea, Michigan State
Campbell 131
"Better living through computation: Connecting the first stars in the universe to the Milky Way using cosmological simulations"
Galaxies are complicated beasts  many physical processes operate simultaneously, and over a huge range of scales in space and time. As a result, creating accurate models of the formation and evolution of galaxies over the lifetime of the universe presents tremendous technical challenges. In this talk I will discuss these challenges and their solutions, and will explain how largescale computational models can be used to gain insights into the very first galaxies that formed in the universe (over 13 billion years ago!), and how we can use both these computational models and observations of the Milky Way and its neighbors to infer how galaxies have grown and evolved in the intervening time.
Nov 7, Tuesday
1:10 pm (Cosmology/ BCCP)
Justin Alsing, CCA
Campbell 131
"Towards scalable likelihoodfree inference for cosmology"
Many statistical models for cosmological data analysis can be simulated forwards but have intractable likelihoods, for example because of complicated reduction pipelines and systematics, or nonlinear structure formation and astrophysics impacting small scales. Likelihoodfree inference provides an alternative paradigm for doing Bayesian inference using forward simulations only, eliminating the need to make uncomfortable likelihood approximations and allowing us to extract information from previously inaccessible statistics and scales. Traditional Approximate Bayesian Computation (ABC) methods involve drawing parameters and forward simulating mock data, accepting the parameters if the simulated data are within some small distance \epsilon of the real data (recovering the true posterior in the limit \epsilon > 0). These methods scale poorly with the size of the dataset and critically slow down as \epsilon > 0, often leading to expensive analyses with overly conservative error bars. I will introduce a new approach to likelihoodfree inference that is “epsilonfree”, bypassing the limitations of ABC, and use massive lossless data compression to dramatically reduce the size of the dataspace; together, these advances provide a framework for performing scalable likelihoodfree inference from cosmological surveys, using reasonable numbers of forward simulations.
Nov 14, Tuesday
1:10 pm (Cosmology/ BCCP)
Anze Slosar, BNL
Campbell 131
21cm cosmology in the 20s
Intensity mapping of the neutral hydrogen 21cm line is in principle an
extremely cheap and simple method, but has in practice proven to be a
devilishly difficult way of mapping the universe. I will argue that it
nevertheless offers the best way to complete the programmatic goal of
measuring the linear scales of the universe deep into the
preacceleration era after current crop of optical experiments (DESI,
LSST, Euclid, WFIRST). It will give us expansion history measurements
to z=6, which can constrain early dark energy and modified gravity,
break the w/mnu degeneracy, improve curvature constraints and help
with radiation density measurements. It could in principle also
measure the largest scales in the universe through tidal
reconstruction, reconstruct the weaklensing field and probe the
primordial nonGaussianity, but robust forecasts for these promises
have yet to be developed. I will discuss the scientific, technological
and programmatic issues that need to be overcome to make such
experiment a reality.
Nov 21, Tuesday
no talk
Nov 28, Tuesday
1:10 pm (Cosmology/ BCCP)
Francois Lanusse, CMU
Campbell 131
December 2017
Dec 5, Tuesday
1:10 pm (informal talk, term seminars are over)
Michael Kopp, CEICO
Campbell 131
Solving the Vlasov equation in two spatial dimensions with the Schrödinger method
I present recent progress (Kopp, Vattis & Skordis 1711.00140) to solve the Vlasov equation via the socalled Schrödinger method (Widrow & Kaiser 1993). This technique consists of solving the SchrödingerPoisson equation, together with a prescription to construct a phase space distribution from the wave function.
For the first time, we implemented the Schrödinger method in two spatial dimensions, extending the onedimensional previous studies.
We performed a quantitative comparison of our code and the stateoftheart Vlasov solver ColDICE (Sousbie & Colombi 1509.07720), finding excellent agreement.
I review how the fully fledged phase space dynamics is encoded in a wave function with its mere 2 spatial degrees of freedom, and how the cumulants of the phase space distribution, including vorticity and velocity dispersion, can be easily decoded from the wave function in a way that completely avoids the cumbersome highdimensional phase space.
As an application we evaluate the background pressure induced by the nonlinearity of large scale structure formation, thereby estimating the magnitude of cosmological backreaction. We find that it is negligibly small and has time dependence and magnitude compatible with expectations from the effective field theory of large scale structure.
January 2018
Jan 16, Tuesday
1:10 pm (Cosmology/ BCCP)
tba,
Campbell 131
Jan 23, Tuesday
1:10 pm (Cosmology/ BCCP)
tba,
Campbell 131
Jan 30, Tuesday
1:10 pm (Cosmology/ BCCP)
tba,
Campbell 131
February 2018
Feb 6, Tuesday
1:10 pm (Cosmology/ BCCP)
tba,
Campbell 131
Feb 13, Tuesday
1:10 pm (Cosmology/ BCCP)
Tom Collett, Portsmouth
Campbell 131
Feb 20, Tuesday
1:10 pm (Cosmology/ BCCP)
tba,
Campbell 131
Feb 27, Tuesday
1:10 pm (Cosmology/ BCCP)
tba,
Campbell 131
March 2018
Mar 6, Tuesday
1:10 pm (Cosmology/ BCCP)
tba,
Campbell 131
Mar 13, Tuesday
1:10 pm (Cosmology/ BCCP)
tba,
Campbell 131
Mar 20, Tuesday
1:10 pm (Cosmology/ BCCP)
Boris Leistedt, NYU
Campbell 131
Mar 27, Tuesday
Spring Break, no talk
April 2018
Apr 3, Tuesday
1:10 pm (Cosmology/ BCCP)
tba,
Campbell 131
Apr 10, Tuesday
1:10 pm (Cosmology/ BCCP)
tba,
Campbell 131
Apr 17, Tuesday
1:10 pm (Cosmology/ BCCP)
Benedikt Diemer, Harvard
Campbell 131
Apr 24, Tuesday
1:10 pm (Cosmology/ BCCP)
tba,
Campbell 131
Past Months
August 2017
Aug 29, Tuesday
1:10 pm (Cosmology/ BCCP)
Elisabeth Krause, Stanford
Campbell 131
Cosmology Results from the Dark Energy Survey Year 1
This talk presents cosmology constraints from a combined analysis of galaxy clustering and weak gravitational lensing, using 1321 deg2 of griz imaging data from the first year of the Dark Energy Survey (DES Y1). The analysis combines (i) the cosmic shear correlation function of 26 million source galaxies in four redshift bins, (ii) the galaxy angular autocorrelation function of 650,000 luminous red galaxies in five redshift bins, and (iii) the galaxyshear crosscorrelation of luminous red galaxy positions and source galaxy shears. These three measurements yield consistent cosmological results, and provide constraints on the amplitude of density fluctuations (S8 = 0.794+0.0290.027) and dark energy equation of state (w = 0.80+0.200.22) that are competitive with those from Planck cosmic microwave background measurements.
I will describe the validation of measurements and modeling from catalogs to cosmology, and highlight cosmology constraints from the combination of DES Y1 with external data sets.
Based on DES Collaboration 2017 (1708.01530) and supporting papers
September 2017
Sep 5, Tuesday
1:10 pm (Cosmology/ BCCP)
Jordan Mirocha, UCLA
Campbell 131
Metrics for Disentangling PopII and PopIII Contributions to the 21cm Background
The first generations of stars to form after the Big Bang are expected to have dramatically different properties than stars today. Most notably, in the absence of heavy elements, fragmentation in the Universe's first protostellar clouds was likely minimal, resulting in perhaps only one or a few massive stars. This idea has surfaced in many contexts over the last two decades, as very massive stars could kickstart reionization, provide the seeds of today's supermassive black holes, and give rise to abnormal abundance patterns in metalpoor stars in our own Galaxy. Unfortunately, directly detecting these socalled Population III stars (or clusters of them) may be impossible, even with the next generation of optical/nearIR spacebased observatories. In this talk, I will focus on the prospects for detecting PopIII sources indirectly using the skyaveraged ("global") 21cm background, which is currently being targeted by several experiments on the ground. I will describe first a new set of predictions for the global signal calibrated to match highz galaxy luminosity functions (i.e., "normal" PopII galaxies), and then move on to the subtle  but persistent  modulations of the shape of the global signal induced by the addition of PopIII sources. Such modulations may even be accessible to experiments operating at frequencies above 100 MHz (z < 13), at which point PopIII star formation has become subdominant to PopII in most models.
Sep 12, Tuesday
1:10 pm (Cosmology/ BCCP)
Tobias Schmidt, MPIA/UCSB
Campbell 131
Quasar Lifetime and Obscuration Constraints from the HeII Transverse
Proximity Effect
The reionization of helium at z~3 is the final phase transition of the
intergalactic medium and supposed to be driven purely by quasars. The HeII
transverse proximity effect  enhanced HeII transmission in a background
sightline caused by the ionizing radiation of a foreground quasar 
therefore offers a unique opportunity to probe the morphology of HeII
reionization and to investigate the emission properties of quasars, e.g.
ionizing emissivity, lifetime and beaming geometry. I will present results
from the first statistical analysis of the HeII transverse proximity
effect. We use the mostrecent HST/COS farUV dataset and conducted a
dedicated optical spectroscopic survey to find foreground quasars around
22 HeII sightlines. We find a large objecttoobject variance and a
surprising absence of strong transmission peaks for the strongest
foreground quasars. However, using a stacking analysis we find statistical
evidence for the HeII Transverse Proximity Effect which places a
geometrical constraint on the quasar lifetime of t_Q > 25 Myr. With a more
detailed modeling based on postprocessing of cosmological hydrodynamic
simulations we derive joint constraints on quasar age and obscuration for
the four strongest foreground quasars, suggesting that three of them are
highly obscured or relatively young.
Sep 14, Thursday
4 pm (RPM)
Cora Dvorkin, Harvard
LBL 505132
"Discovering New Physics Beyond the Standard Model with Cosmological Data Sets"
Cosmological observations have provided us with answers to ageold questions, involving the age, geometry, and composition of the universe. However, there are profound questions that still remain unanswered. I will describe ongoing efforts to shed light on some of these questions.
In the first part of this talk, I will explain how we can use measurements of the CMB and the largescale structure of the universe to reconstruct the detailed physics of much earlier epochs, when the universe was only a tiny fraction of a second old. In particular, I will show how we can probe the shape of the inflationary potential, extra degrees of freedom during inflation, and the signature of possible particles with mass and spin during this period.
In the last part of the talk, I will discuss how we can use observations at large scales and subgalactic scales (through strong gravitational lensing) to improve our understanding of another open question in fundamental physics: the particle nature of dark matter.
Sep 19, Tuesday
1:10 pm (Cosmology/ BCCP)
Julien Carron, Sussex
Campbell 131
Planck CMB delensing and beyond
The deflection of CMB photons by largescale structures smooths the observed CMB acoustic peaks and introduces a sizeable white noise component in the polarization Bmode.
In order to achieve best constraints on the inflationary perturbations tensor to scalar ratio, upcoming lownoise polarizationbased experiments must be able to undo these deflections, an operation called delensing.
I will present the first detection of Bmode delensing, and first internal delensing of the CMB that we recently performed on Planck public maps, as well as discuss methods and prospects for the future.
Sep 21, Thursday
4 pm (RPM)
Kev Abazajian, Irvine
LBL 505132
Candidate Signals and Stringent Constraints from Dark Matter in the Sky
I will discuss analyses leading to two recent candidate detections of photons from dark matter. Specifically, these are: first, gamma rays in a continuum “bump” at a few GeV which can be due to WIMPlike dark matter annihilation in the Galactic Center; and, second, Xrays from clusters of galaxies and Andromeda consistent with monoenergetic 3.55 keV photons from dark matter decay such as that predicted from sterile neutrino dark matter. Commensurately, there are also stringent constraints on these signals. I will discuss the particle and cosmological model implications of both.
Sep 21, Thursday
4:10 pm (Astronomy Colloquium)
Risa Wechsler, Stanford
LeConte Hall 1
Cosmology from the Dark Energy Survey and Beyond
A new generation of sky surveys are beginning to map the universe’s expansion history and evolution of structure over the last ~ 12 billion years, using statistical constraints from hundreds of millions of galaxies.
I will present cosmological constraints from a combined analysis of galaxy clustering and weak gravitational lensing, using 1321 deg^2 of griz imaging data from the first year of the Dark Energy Survey (DES Y1). The analysis combines (i) the cosmic shear correlation function of 26 million source galaxies in four redshift bins, (ii) the galaxy angular autocorrelation function of 650,000 luminous red galaxies in five redshift bins, and (iii) the galaxyshear crosscorrelation of luminous red galaxy positions and source galaxy shears. These three measurements yield consistent cosmological results, and provide constraints on the amplitude of density fluctuations (S8 = 0.794+0.0290.027) and dark energy equation of state (w = 0.80+0.200.22) that are competitive with those from Planck cosmic microwave background measurements. I will also describe the extensive validation of the measurements and modeling, with a particular focus on the role of cosmological simulations and modeling of the connection between galaxies and dark matter, highlighting where these efforts will need to go in the future to keep up with the exciting cosmological measurements expected from DES, DESI, LSST and other future galaxy surveys.
Sep 26, Tuesday
1:10 pm (Cosmology/ BCCP)
Andrina Nicola, ETH Zurich
Campbell 131
Integrated approach to cosmology
Recent progress in observational cosmology and the establishment of ΛCDM have relied on the combination of different cosmological probes. These probes are not independent, since they all measure the same physical fields. The resulting crosscorrelations allow for a robust test of the cosmological model through the consistency of different physical tracers and for the identification of systematics. Integrated analyses taking into account both the auto as well as the crosscorrelations between cosmological probes therefore present a promising analysis method for both current as well as future data.
In this talk, I will present an integrated analysis of CMB temperature anisotropies, CMB lensing, galaxy clustering and weak lensing as well as background probes. I will describe the cosmological probe combination framework, the obtained results and illustrate how this analysis has provided a confirmation of ΛCDM through the consistency of different probes. Furthermore, I will discuss possible tensions between the derived constraints on cosmological parameters and existing ones.
Sep 29, Friday
12 noon (INPA)
Stephen Portillo, Harvard
LBL 505132
Improved Source Detection in Crowded Fields using Probabilistic Cataloging
Cataloging is challenging in crowded fields because sources are extremely covariant with their neighbors and blending makes even the number of sources ambiguous. We present the first optical probabilistic stellar catalog, cataloging a crowded (~0.1 sources per pixel) Sloan Digital Sky Survey r band image from M2. Probabilistic cataloging returns an ensemble of catalogs inferred from the image and thus can capture sourcesource covariance and deblending ambiguities. By comparing to a traditional catalog of the same image and a Hubble Space Telescope catalog of the same region, we show that our catalog ensemble better recovers sources from the image. It goes more than a magnitude deeper than the traditional catalog while having a lower false discovery rate brighter than 20th magnitude. Future telescopes will be more sensitive, and thus more of their images will be crowded. We detail our efforts to extend probabilistic cataloging to galaxies, making the method applicable to the data that will be collected in the Large Synoptic Survey Telescope era.
October 2017
Oct 2, Monday
12:10 pm (TAC)
Xiancheng Ma, Caltech
Campbell 131
"Simulating galaxies at the epoch of reionization"
Galaxies at z>5 are thought to be the dominant sources for cosmic reionization, but current knowledge on their abundance, property, and contribution to reionization is still limited. With JWST to be launched next year, it allows us to study highredshift galaxies in much detail. I will introduce a new suite of highresolution cosmological zoomin simulations of z>5 galaxies, taking advantage of the realistic models of the multiphase ISM, star formation, and stellar feedback from the FIRE project. These simulations provide a more realistic sample of model galaxies, which can be very useful to predict and understand future observations. I will show the predicted galaxy scaling relations, mass functions, and luminosity functions at z>5. I will also discuss some applications using these simulations, including understanding the escape fraction of ionizing photons, highredshift galaxy morphologies and their implications for the Hubble Frontier Field observations, dust extinction and the brightend galaxy UV luminosity functions, etc.
Oct 3, Tuesday
1:10 pm (Cosmology/ BCCP)
Alexandra Amon, Edinburgh
Campbell 131
Weak Lensing with the ESO KiloDegree Survey
The Kilo Degree Survey, (KiDS) is an ongoing weak lensing survey that will span 1500 square degrees, on completion, in nine opticalNIR bands. I will summarize the recent cosmology results from our analysis of the first third of the survey area, and detail an important test for the robustness of our weak lensing analysis, where I compare our fiducial highquality KiDS multiband dataset with 815 square degrees of the overlapping, shallower KiDS ibandonly survey. I will conclude by presenting a new test of General Relativity, measuring the "gravitational slip" statistic, E_G, in a joint analysis of KiDS with the samesky spectroscopic surveys; BOSS and the recently completed 2dF Lensing Survey.
Oct 5, Thursday
4 pm (RPM)
Simeon Bird, JHU
LBL 505132
Strong Absorbers in the Lymanalpha Forest and Primordial Black Holes
Strong HI absorbers are essentially the largest foreground contamination for Lyman alpha forest surveys, and so a better understanding of them is necessary for achieving the goals of future Lyman alpha cosmology surveys. I will talk about a new automated technique for generating a probabilistic catalogue of strong absorbers for the entire survey, allowing more robust cleaning of the foreground. Since no technique can entirely remove strong absorbers, I will discuss new templates for characterising their effect on the flux power spectrum. A secondary systematic is induced by interpolation error between theoretical models, and I will discuss techniques to mitigate this error with refining Gaussian Process emulators. Lastly I will discuss the interesting possibility that the surprisingly common mergers of ~30 solar mass black holes observed by LIGO could be primordial black hole dark matter, which is intriguingly (still) not convincingly ruled out.
Oct 6, Friday
12 noon (INPA)
Marie Lau, Santa Cruz
LBL 505132
Quasars Probing Quasars: the Circumgalactic Medium Surrounding z ~ 2 Quasars
Understanding the circumgalactic mediumthe gaseous halo surrounding a galaxy, is an integral part to understanding galaxy evolution. The z ~ 23 universe is interesting as this is when the star formation rate and AGN activity peak. My work concludes the decadelong Quasars Probing Quasars survey designed for studying massive galaxy formation and quasar feedback. I use background quasar sightlines that pass close to foreground quasars to study the circumgalactic medium of quasarhost galaxies in absorption. My sample of 149 quasar pairs involve spectra taken with 17 different optical and near IR instruments. I present results on the statistical and physical properties of the quasar circumgalactic medium. My results pose challenges for cosmological hydrodynamic simulations to produce a substantial cool gas reservoir surrounding quasars, that is also enriched and exhibits extreme kinematics.
I will discuss other science goals that can be facilitated using the spectral databases and absorptionline analysis tools built. If there is interest, I will show preliminary results on a peculiar tidal disruption event and evidence for deep internal mixing in red giants.
Oct 10, Tuesday
1:10 pm (Cosmology/ BCCP)
Charlotte Mason, UCLA
Campbell 131
What Can Galaxies Tell Us About The Epoch of Reionization?
The reionization of neutral hydrogen in the intergalactic medium (IGM) in the universe's first billion years was likely driven by the first stars and galaxies, and its history encodes information about their properties. But the timeline of reionization is not wellmeasured and it is unclear whether galaxies alone can produce enough ionizing photons. I will describe two ways in which galaxies at our current observational frontiers can constrain reionization. One tool is the UV luminosity function (LF), which traces the evolution of starforming galaxies and their ionizing photons. I will describe a simple, but powerful, model for LF evolution and its implications for reionization and z>10 surveys with JWST. Secondly, Lyman alpha (Lya) emission from galaxies is a potential probe of the IGM, but requires disentangling physics on pc to Gpc scales. I will introduce a new forwardmodeling Bayesian framework which combines IGM simulations with models of interstellar medium conditions to infer the average neutral hydrogen fraction from Lya observations. I will present our new measurement of the neutral fraction at z~7 and place it in the context of other constraints on the reionization history. I will describe ongoing efforts to build larger samples of Lya emitting galaxies with the HST survey GLASS, and future prospects with JWST.
Oct 12, Thursday
4 pm (RPM)
Barnabas Poczos, CMU
LBL 505132
Distribution Regression and its Applications
The most common machine learning algorithms operate on
finitedimensional vectorial feature representations. In
many applications, however, the natural representation of the data
consists of more complex objects,
for example functions, distributions, and sets, rather than
ﬁnitedimensional vectors. In this talk
we will discuss machine learning algorithms that can operate directly
on these complex
objects. For this purpose, we use nonparametric statistical methods
that can consistently estimate the inner product, distance, and
certain kernel functions between distributions, sets, and other
objects. We will discuss applications in various scientific areas
including cosmology (e.g. estimating the mass of galaxy clusters,
finding anomalous galaxy clusters, estimating the cosmological
parameters of our Universe, accelerating cosmological simulations),
fluid dynamics (finding anomalous events in turbulence data),
neuroimaging, and agriculture.
Oct 13, Friday
12 noon (INPA)
Michael Walther, MPIA/UCSB
LBL 505132
New Constraints on Thermal Evolution in the IGM from the Small Scale Ly α Forest Power Spectrum
The lineofsight power spectrum (P_F(k)) of the Lyα forest has proven to be a valuable tool for doing cosmological observations. It also not only allows to constrain cosmological parameters, but enables us to measure the thermal state of the IGM at redshifts z>1.8. While at large scales (k<0.02 s/km) P_F(k) has been accurately measured using the large number (10^310^5) of quasar sightlines from SDSS and BOSS, there are much less spectra available at smaller scales (larger k). Prior power spectrum measurements from highresolution data only used several times less (QSO) spectra in our redshift range about 15 years ago whereas a few hundred became available in the meantime. We therefore performed a new measurement using 74 quasar sightlines with 1.8 < z< 3.4 significantly improving the precision of the smallscale P_F(k). Using this additional precision on small scales combined with the BOSS measurements on large scales enables us to accurately constrain the thermal cutoff scale of the IGM set by a combination of temperature broadening of Lyα forest lines, and 'Jeans' smoothing due to baryonic pressure support. We perform an MCMC analysis based on Gaussian process based techniques for interpolation between a grid of highresolution hydrodynamical simulations and using our new highresolution dataset, the BOSS data, a recent XSHOOTER analysis, and a previous HIRES/MIKE analysis at higher redshifts. This allows us to measure thermal evolution in the IGM from z=5.4 to z=1.8 showing a suggestive peak at z~3.3 that might be attributed to He reionization. These constraints will help solving the existing discrepancies in the IGM thermal evolution between different works using different techniques as existing degeneracies between different thermal parameters in the existing measurements can be broken in our analysis.and can be used to place limits on possible exotic sources of heating. Additionally a better knowledge of thermal evolution will also lead to better constraints of e.g. the nature of dark matter or neutrino masses by breaking degeneracies in those measurements and thereby improve our knowledge of the underlying cosmology.
Oct 17, Tuesday
1:10 pm (Cosmology/ BCCP)
Emanuela Dimastrogiovanni, CWRU/Perimeter
Campbell 131
Exploring the early Universe: possible signatures and tests
I will describe some interesting scenarios for the generation of gravitational waves from inflation and their characteristic imprints (at the level of both the power spectrum and the bispectrum), that can be tested with upcoming observations. I will then move on to discuss techniques that may help us learn more about the early Universe by gathering additional information on large scales. Specifically this is done with crosscorrelations of secondary CMB anisotropies from kinetic and polarized Sunyaev–Zel'dovich effects with tracers of the large scale structure.
Oct 19, Thursday
4 pm (RPM)
Anthony Pullen, NYU
LBL 505132
Revealing CII Emission with LSS Crosscorrelations
In this talk, I will present our joint measurement of cosmic infrared background (CIB) and CII line emission from large scales at redshift z=2.5 using an MCMC analysis of crosscorrelations of the 3 highfrequency Planck bands with both SDSSIII quasars at z=23.2 and SDSSIII CMASS galaxies at z=0.430.7. The CII emission is expected to correlate with the quasars and appear in the Planck 545 GHz band, while the other crosscorrelations are assumed to mostly be attributable to CIBLSS correlations. We report an excess emission inconsistent with the null result at 95% confidence, with an intensity value favoring the higher range of CII models. I will also forecast potential CII intensity constraints from Planck crosscorrelated with DESI quasars.
Oct 20, Friday
12 noon (INPA talk)
Prabhat, NERSC
LBL 505132
Deep learning for science
Deep Learning has revolutionized the fields of computer vision, speech recognition and control systems. Can Deep Learning (DL) work for scientific problems? This talk will explore a variety of DOE/LBL applications that are currently benefiting from Deep Learning. We will review classification and regression problems in astronomy, cosmology, neuroscience, genomics and highenergy physics. We will outline several short and longterm challenges at the frontier of DL research, and speculate about the future of DL for dataintensive science.
Oct 23, Monday
12:10 pm (TAC)
Sarah Loebman, Davis
Campbell 131
Simulating the Evolving Milky Way: From Disk to Halo
The Milky Way is the most observationally accessible galaxy in our Universe. In many ways it is also a "typical galaxy," making it an important benchmark for studying galaxy formation and testing LCDM. For this reason, uncovering the formation history of the Milky Way is the key goal of major ongoing surveys such as APOGEE and Gaia. However, observations of the Milky Way span a complex multidimensional space which necessitates sophisticated modeling to interpret. In this talk, I will highlight some recent achievements utilizing stateoftheart simulations to aid in our exploration of the Milky Way's formation and evolution. In particular, I will discuss recent APOGEE observations of the Milky Way's disk and the role radial migration has played in redistributing stars within it. I will also discuss observations and simulations of kinematics in the Milky Way's stellar halo, emphasizing how measurements of kinematic moments hold power for constraining the merger history of the Milky Way. Finally, I will discuss a strong prediction of LCDM  that stellar halos are radially anisotropic  and I will highlight the potential Gaia holds for testing this prediction in the Milky Way.
Oct 24, Tuesday
1:10 pm (Cosmology/ BCCP)
Phil Bull, UCB
Campbell 131
Cosmological tests of gravity with radio telescopes
GR is astoundingly welltested in the smallscale, weakfield limit, and constraints in the strong field limit are not far behind thanks to recent gravitational wave and binary pulsar observations. The cosmological regime remains a viable, wellmotivated hiding place for possible deviations from GR though. I discuss how radio telescopes are particularly wellsuited to the task of determining the nature of gravity on cosmological scales. As well as constraining background expansion at high redshift and providing new ways of measuring the weak lensing signal, radio surveys will be able to make the first practical measurements of clustering at the Hubble scale at late times. I will also describe several novel observational strategies that will make these measurements possible, and discuss how they can be supported experimentally.
Oct 27, Friday
12 noon (INPA talk)
Jason Prochaska, Santa Cruz
LBL 505132
Deep Learning of Quasar Spectra
I will describe our development of a convolutional
neural network (CNN) to learn to search for and characterize
absorption lines in quasar spectra. Specifically, the algorithm
discovers and measures the redshift and Hydrogen column
density of damped Lya systems (DLAs). These systems dominate
the neutral hydrogen gas of the universe, trace the interstellar
medium of distant galaxies, and offer cosmological constraints
on the build up of gas and heavy elements across cosmic time.
I will discuss the lessons learned employing CNN techniques
on large spectral datasets and the prospects for future analysis.
Oct 31, Tuesday
1:10 pm (Cosmology/ BCCP)
Marko Simonovic, IAS
Campbell 131
Analytical Tools for Largescale Structure of the Universe
The largescale structure of the universe is becoming the leading probe of cosmology. The ongoing and future LSS surveys combined with the next generation CMB experiments aim to provide the unprecedented constraints on the statistic of the initial conditions and physics of inflation, to measure the sum of neutrino masses and to shed new light on the nature of dark matter and dark energy. The main theoretical challenge in this program is understanding the nonlinear evolution of density fluctuations induced by gravitational collapse. In this talk I will describe several analytical tools to deal with this problem. In the first part I will describe perturbation theory approach to gravitational nonlinearities, its advantages, successes and limitations. I will also present some recent progress in evaluation of higher order/loop correlation functions using scattering amplitudes of a massless QFT. In the second part I will focus on some nonperturbative results such as consistency relations for the LSS and their applications to the modeling of the nonlinear evolution of the BAO peak.
