Lehman H. Garrison
A5084455874- Galaxies: Formation, Evolution, Phenomena
- Astronomy and Astrophysical Research
- Scientific Research and Discoveries
- Cosmology and Gravitation Theories
- Stellar, planetary, and galactic studies
- Astronomical Observations and Instrumentation
- Computational Physics and Python Applications
- Gamma-ray bursts and supernovae
- Radio Astronomy Observations and Technology
- Remote Sensing in Agriculture
- CCD and CMOS Imaging Sensors
- Astrophysics and Cosmic Phenomena
- History and Developments in Astronomy
- Dark Matter and Cosmic Phenomena
- Geophysics and Gravity Measurements
- Advanced Data Storage Technologies
- Algorithms and Data Compression
- Bayesian Methods and Mixture Models
- High-Energy Particle Collisions Research
- Parallel Computing and Optimization Techniques
- Neural Networks and Applications
- Gaussian Processes and Bayesian Inference
- Advanced Proteomics Techniques and Applications
- Plant Surface Properties and Treatments
- Solar and Space Plasma Dynamics
Flatiron Institute
2019-2024
Flatiron Health (United States)
2019-2024
The Ohio State University
2022
Simons Foundation
2021-2022
Center for Astrophysics Harvard & Smithsonian
2016-2020
Harvard University
2019
Abstract The Astropy Project supports and fosters the development of open-source openly developed Python packages that provide commonly needed functionality to astronomical community. A key element is core package astropy , which serves as foundation for more specialized projects packages. In this article, we an overview organization project summarize features in package, recent major release, version 2.0. We then describe infrastructure designed facilitate support a broader ecosystem...
The Dark Energy Spectroscopic Instrument (DESI) has embarked on an ambitious five-year survey to explore the nature of dark energy with spectroscopy 40 million galaxies and quasars. DESI will determine precise redshifts employ Baryon Acoustic Oscillation method measure distances from nearby universe z > 3.5, as well growth structure probe potential modifications general relativity. In this paper we describe significant instrumentation developed for survey. new includes a wide-field, 3.2-deg...
ABSTRACT We present the public data release of AbacusSummit cosmological N-body simulation suite, produced with Abacus code on Summit supercomputer Oak Ridge Leadership Computing Facility. achieves $\mathcal {O}(10^{-5})$ median fractional force error at superlative speeds, calculating 70M particle updates per second node early times, and 45M late times. The suite totals roughly 60 trillion particles, core which is a set 139 simulations mass $2\times 10^{9}\, h^{-1}\, \mathrm{M}_\odot$ in...
ABSTRACT The two-point correlation function (2PCF) is the most widely used tool for quantifying spatial distribution of galaxies. Since galaxies determined by galaxy formation physics as well underlying cosmology, fitting an observed yields valuable insights into both. calculation a 2PCF involves computing pair-wise separations and consequently, time-scales quadratically with number next-generation surveys are slated to observe many millions galaxies, such would be prohibitively...
Abstract The Dark Energy Spectroscopic Instrument (DESI) completed its 5 month Survey Validation in 2021 May. Spectra of stellar and extragalactic targets from constitute the first major data sample DESI survey. This paper describes public release those spectra, catalogs derived properties, intermediate products. In total, includes good-quality spectral information 466,447 objects targeted as part Milky Way Survey, 428,758 Bright Galaxy 227,318 Luminous Red sample, 437,664 Emission Line...
Abstract A key component of the Dark Energy Spectroscopic Instrument (DESI) survey validation (SV) is a detailed visual inspection (VI) optical spectroscopic data to quantify metrics. In this paper we present results from VI quasar using deep coadded SV spectra. We show that majority (≈70%) main-survey targets are spectroscopically confirmed as quasars, with ≈16% galaxies, ≈6% stars, and ≈8% low-quality spectra lacking reliable features. nonnegligible fraction quasars misidentified by...
Abstract The Dark Energy Spectroscopic Instrument (DESI) Survey has obtained a set of spectroscopic measurements galaxies to validate the final survey design and target selections. To assist in these tasks, we visually inspect DESI spectra approximately 2500 bright galaxies, 3500 luminous red (LRGs), 10,000 emission-line (ELGs) obtain robust redshift identifications. We then utilize inspected information characterize performance operation. Based on visual inspection (VI) catalogs, our...
We present a public data release of halo catalogs from suite 125 cosmological $N$-body simulations the Abacus project. The span 40 $w$CDM cosmologies centered on Planck 2015 cosmology at two mass resolutions, $4\times 10^{10}\;h^{-1}M_\odot$ and $1\times 10^{10}\;h^{-1}M_\odot$, in $1.1\;h^{-1}\mathrm{Gpc}$ $720\;h^{-1}\mathrm{Mpc}$ boxes, respectively. boxes are phase-matched to suppress sample variance isolate dependence. Additional volume is available via 16 fixed varied phase; few...
ABSTRACT We present abacus, a fast and accurate cosmological N-body code based on new method for calculating the gravitational potential from static multipole mesh. The analytically separates near- far-field forces, reducing former to direct 1/r2 summation latter discrete convolution over multipoles. achieves 70 million particle updates per second node of Summit supercomputer, while maintaining median fractional force error 10−5. express simulation time-step as an event-driven ‘pipeline’,...
We present a high-fidelity realization of the cosmological $N$-body simulation from Schneider et al. (2016) code comparison project. The was performed with our Abacus code, which offers high force accuracy, performance, and minimal particle integration errors. consists $2048^3$ particles in $500\ h^{-1}\mathrm{Mpc}$ box, for mass $1.2\times 10^9\ h^{-1}\mathrm{M}_\odot$ $10\ h^{-1}\mathrm{kpc}$ spline softening. executed 1052 global time steps to $z=0$ 107 hours on one dual-Xeon, dual-GPU...
ABSTRACT We introduce the AbacusHOD model and present two applications of AbacusSummit simulations to observations. is a Halo Occupation Distribution (HOD) framework written in Python that particle-based, multitracer, highly generalized, efficient. It designed specifically with multitracer/cosmology analyses for next-generation large-scale structure surveys mind, takes advantage volume precision offered by new state-of-the-art cosmological simulations. The also customizable should be broadly...
We describe a new method (\textsc{CompaSO}) for identifying groups of particles in cosmological $N$-body simulations. \textsc{CompaSO} builds upon existing spherical overdensity (SO) algorithms by taking into consideration the tidal radius around smaller halo before competitively assigning membership to particles. In this way, finder allows more effective deblending haloes close proximity as well formation on outskirts larger ones. This halo-finding algorithm is used \textsc{AbacusSummit}...
We present a novel simulation-based hybrid emulator approach that maximally derives cosmological and Halo Occupation Distribution (HOD) information from non-linear galaxy clustering, with sufficient precision for DESI Year 1 (Y1) analysis. Our first samples the HOD space on fixed simulation grid to constrain high-likelihood region of cosmology+HOD parameter space, then constructs within this constrained region. This significantly reduces volume emulated over, thus achieving much smaller...
Abstract The One-Percent survey of the Dark Energy Spectroscopic Instrument collected ∼ 270 k emission line galaxies (ELGs) at 0.8 < z 1.6. high completeness sample allowed clustering to be measured down scales never probed before, 0.04 Mpc/ h in r p for projected 2-point correlation function (2PCF) and 0.17 galaxy pair separation s 2PCF monopole quadrupole. most striking feature measurements is a strong signal smallest scales, below 0.2 1 . We analyse these data halo occupation...
In cosmological N-body simulations, the representation of dark matter as discrete 'macroparticles' suppresses growth structure, such that simulations no longer reproduce linear theory on small scales near kNyquist. Marcos et al. demonstrate this is due to sparse sampling modes kNyquist and often-assumed continuum growing are not proper particle system. We develop initial conditions (ICs) respect then rescale mode amplitudes account for suppression. These ICs also allow us take advantage our...
A measurement of a primordial non-Gaussianity (PNG) signal through late- or early-Universe probes has the potential to transform our understanding physics Universe. While large-scale structure observables in principle contain vital information, interpreting these measurements is challenging due poorly understood astrophysical effects. Luckily, $N$-body simulations, such as abacuspng set presented this study, consisting 9 boxes, each size ${L}_{\mathrm{box}}=2\text{ }\text{ }\mathrm{Gpc}/h$...
The combination of galaxy-galaxy lensing (GGL) with galaxy clustering is one the most promising routes to determining amplitude matter at low redshifts. We show that extending clustering+GGL analyses from linear regime down $\sim 0.5 \, h^{-1}$ Mpc scales increases their constraining power considerably, even after marginalizing over a flexible model non-linear bias. Using grid cosmological N-body simulations, we construct Taylor-expansion emulator predicts autocorrelation...
We present the GeneRalized ANd Differentiable Halo Occupation Distribution (GRAND-HOD) routine that generalizes standard five parameter halo occupation distribution (HOD) model with various halo-scale physics and assembly bias. describe methodology of four different generalizations: satellite generalization, velocity bias, closest approach distance showcase signatures these generalizations in 2-point correlation function (2PCF) squeezed 3-point (squeezed 3PCF). identify generalized HOD...
We describe our nonlinear emulation (i.e., interpolation) framework that combines the halo occupation distribution (HOD) galaxy bias model with $N$-body simulations of structure formation, designed to accurately predict projected clustering and galaxy-galaxy lensing signals from luminous red galaxies (LRGs) in redshift range $0.16 < z 0.36$ on comoving scales $0.6 r_p 30$ \hMpc. The interpolation accuracy is $\lesssim 1-2$ per cent across entire physically plausible parameters for all...
Abstract We present a deep machine learning (ML)–based technique for accurately determining σ 8 and Ω m from mock 3D galaxy surveys. The surveys are built the AbacusCosmos suite of N -body simulations, which comprises 40 cosmological volume simulations spanning range parameter values, we account uncertainties in formation scenarios through use generalized halo occupation distributions (HODs). explore trio ML models: convolutional neural network (CNN), power spectrum–based fully connected...
Analysis of large galaxy surveys requires confidence in the robustness numerical simulation methods. The simulations are used to construct mock catalogs validate data analysis pipelines and identify potential systematics. We compare three $N$-body codes, ABACUS, GADGET, SWIFT, investigate regimes which their results agree. run at different mass resolutions, $6.25\times10^{8}$, $2.11\times10^{9}$, $5.00\times10^{9}~h^{-1}$M$_{\odot}$, matching phases reduce noise within comparisons. find...
In order to address fundamental questions related the expansion history of Universe and its primordial nature with next generation galaxy experiments, we need model reliably large-scale structure observables such as correlation function power spectrum. Cosmological $N$-body simulations provide a reference through which can test our models, but their output suffers from sample variance on large scales. Fortunately, this is regime where accurate analytic approximations exist. To reduce...
Abstract We train graph neural networks on halo catalogs from Gadget N -body simulations to perform field-level likelihood-free inference of cosmological parameters. The contain ≲5000 halos with masses ≳10 10 h −1 M ⊙ in a periodic volume <?CDATA ${(25\,{h}^{-1}\,\mathrm{Mpc})}^{3}$?> <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" overflow="scroll"> <mml:msup> <mml:mrow> <mml:mo stretchy="false">(</mml:mo> <mml:mn>25</mml:mn> <mml:mspace width="0.25em" /> <mml:mi>h</mml:mi>...