Search NASA⌕ Search

SEARCH · Search NASA

Results for “statistical selection”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 181 records · Page 10

Upper Subcritical Limit Calculations with Correlated Integral Experiments

The American National Standards Institute (ANSI) and American Nuclear Society (ANS) standard for Validation of Neutron Transport Methods for Nuclear Criticality Safety Calculations defines the Upper Subcritical Limit (USL) as “a limit on the calculated k effective value established to ensure that conditions calculated to be subcritical will actually be subcritical.” Often, USL calculations are based on statistical techniques that infer information about a nuclear system of interest (application) from a set of known/well characterized similar systems (experiments). In determining a USL, it is common to statistically determine a calculation bias and bias uncertainty. Bias is “the systematic difference between calculated results and experimental data.” The bias uncertainty is “the uncertainty that accounts for the combined effects of uncertainties in the benchmarks, the calculational models of the benchmarks, and the calculational methods.” Both the bias and the bias uncertainty are often inferred through statistical methods applied to the selected set of experiments.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Proper orthogonal decomposition analysis of parallel twin jets measured using particle image velocimetry - 18685

The mixing of turbulent jets has been found in several locations of High Temperature Gas-cooled Reactors (HTGRs) including upper and lower plenums, where jets impinge into open space from coolant channels. In the present work, a parallel twin jets experimental facility was used to generate velocity field of mixing jets by means of Particle Image Velocimetry (PIV), which can be eventually used for validation of Computational Fluid Dynamics (CFD) simulations. The facility has two fixed parallel rectangular jet nozzles (5.8 mm in width, 87.8 mm in length) submerged in the middle of the water tank, which allows an assumption of free jet mixing. In order to compare the experimental data produced to existing data, Re = 4000 {approx} 5000 at the jet outlet was selected. In the past, several statistically meaningful data have been reported using Laser Doppler Velocimetry (LDV) and Hot Wire Anemometer (HWA), which can support Reynolds Averaged Navier-Stokes turbulent models. However, recent CFD simulations using Large Eddy Simulation (LES) require high resolution experimental data in time and space. Thus, in this study, in addition to statistically enough number of data sets for RANS models validation, high temporal and spatial resolution data were collected for the validation of LES calculations. In order to study the turbulent structure Reynolds decomposition and Proper Orthogonal Decomposition (POD) were applied to the data obtained. These decomposition analyses visualized multi-scale vortices and the transportation of eddies. In terms of twin jets structure, the present experimental data showed good agreement with other techniques for the important characteristic parameters including the length of the converging region and the location of the merging point. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

NEUTRON MULTIPLICITY AND ACTIVE WELL NEUTRON COINCIDENCE VERIFICATION MEASUREMENTS PERFORMED FOR MARCH 2009 SEMI-ANNUAL DOE INVENTORY

The Analytical Development (AD) Section field nuclear measurement group performed six 'best available technique' verification measurements to satisfy a DOE requirement instituted for the March 2009 semi-annual inventory. The requirement of (1) yielded the need for SRNL Research Operations Department Material Control & Accountability (MC&A) group to measure the Pu content of five items and the highly enrich uranium (HEU) content of two. No 14Q-qualified measurement equipment was available to satisfy the requirement. The AD field nuclear group has routinely performed the required Confirmatory Measurements for the semi-annual inventories for fifteen years using sodium iodide and high purity germanium (HpGe) {gamma}-ray pulse height analysis nondestructive assay (NDA) instruments. With appropriate {gamma}-ray acquisition modeling, the HpGe spectrometers can be used to perform verification-type quantitative assay for Pu-isotopics and HEU content. The AD nuclear NDA group is widely experienced with this type of measurement and reports content for these species in requested process control, MC&A booking, and holdup measurements assays Site-wide. However none of the AD HpGe {gamma}-ray spectrometers have been 14Q-qualified, and the requirement of reference 1 specifically excluded a {gamma}-ray PHA measurement from those it would accept for the required verification measurements. The requirement of reference 1 was a new requirement for which the Savannah River National Laboratory (SRNL) Research Operations Department (ROD) MC&A group was unprepared. The criteria for exemption from verification were: (1) isotope content below 50 grams; (2) intrinsically tamper indicating or TID sealed items which contain a Category IV quantity of material; (3) assembled components; and (4) laboratory samples. Therefore all (SRNL) Material Balance Area (MBA) items with greater than 50 grams total Pu or greater than 50 grams HEU were subject to a verification measurement. The pass/fail criteria of reference 7 stated 'The facility will report measured values, book values, and statistical control limits for the selected items to DOE SR...', and 'The site/facility operator must develop, document, and maintain measurement methods for all nuclear material on inventory'. These new requirements exceeded SRNL's experience with prior semi-annual inventory expectations, but allowed the AD nuclear field measurement group to demonstrate its excellent adaptability and superior flexibility to respond to unpredicted expectations from the DOE customer. The requirements yielded five SRNL items subject to Pu verification and two SRNL items subject to HEU verification. These items are listed and described in Table 1.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Model averaging approaches to data subset selection

Model averaging is a useful and robust method for dealing with model uncertainty in statistical analysis. Often, it is useful to consider data subset selection at the same time, in which model selection criteria are used to compare models across different subsets of the data. Two different criteria have been proposed in the literature for how the data subsets should be weighted. We compare the two criteria closely in a unified treatment based on the Kullback-Leibler divergence and conclude that one of them is subtly flawed and will tend to yield larger uncertainties due to loss of information. Here, analytical and numerical examples are provided.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Functional protein mining with conformal guarantees

Molecular structure prediction and homology detection offer promising paths to discovering protein function and evolutionary relationships. However, current approaches lack statistical reliability assurances, limiting their practical utility for selecting proteins for further experimental and in-silico characterization. To address this challenge, we introduce a statistically principled approach to protein search leveraging principles from conformal prediction, offering a framework that ensures statistical guarantees with user-specified risk and provides calibrated probabilities (rather than raw ML scores) for any protein search model. Our method (1) lets users select many biologically-relevant loss metrics (i.e. false discovery rate) and assigns reliable functional probabilities for annotating genes of unknown function; (2) achieves state-of-the-art performance in enzyme classification without training new models; and (3) robustly and rapidly pre-filters proteins for computationally intensive structural alignment algorithms. Our framework enhances the reliability of protein homology detection and enables the discovery of uncharacterized proteins with likely desirable functional properties.

59 BASIC BIOLOGICAL SCIENCES↗

Using the Monte Carlo Method to Evaluate the Reliability of Screening Multifamily Housing for Radon

When screening for radon in a multifamily housing complex using a fixed sample density (e.g., testing 1 in 10 (10%) or 1 in 4 units (25%)), the statistical confidence is dependent upon the assumed elevated radon frequency. For example, testing 25% of units in a complex estimated to have eight units with elevated radon levels will provide 90% confidence. However, if it is assumed that, in the same complex, there are only three units with elevated radon levels, the confidence drops to around 58% for the same sample density. Furthermore, in the previous example, if elevated radon levels are not found during the screening, all that can be stated is that the screening provides 90% confidence that there are no more than seven units in the complex with elevated radon levels. To more fully illustrate this uncertainty, ten separate multifamily housing radon data sets with 1 to 10 units with radon levels ≥4 pCi/L (Table 1) were selected for analysis using the Monte Carlo statistical method. Unlike other mathematically based statistical approaches, the Monte Carlo statistical method relies on repeated analysis of randomly selected data. from a 100% sampled complex at various sample densities. Success for each simulation is defined as finding at least one unit with elevated radon levels. In this statistical method, confidence in the overarching conclusion can be greatly enhanced by repeating the simulated screening hundreds or even thousands of times. For this study, each of the ten data sets was simulated 1,000 times. For each simulation, the data set was randomized three times before the fixed percentage of data was selected.

38 RADIATION CHEMISTRY, RADIOCHEMISTRY, AND NUCLEA↗

Testing for X-Ray–SZ Differences and Redshift Evolution in the X-Ray Morphology of Galaxy Clusters

We present a quantitative study of the X-ray morphology of galaxy clusters, as a function of their detection method and redshift. We analyze two separate samples of galaxy clusters: a sample of 36 clusters at 0.35<z<0.9 selected in the X-ray with the ROSAT PSPC 400 deg{sup 2} survey, and a sample of 90 clusters at 0.25<z<1.2 selected via the Sunyaev–Zel’dovich (SZ) effect with the South Pole Telescope. Clusters from both samples have similar-quality Chandra observations, which allow us to quantify their X-ray morphologies via two distinct methods: centroid shifts (w) and photon asymmetry (A{sub phot}). The latter technique provides nearly unbiased morphology estimates for clusters spanning a broad range of redshift and data quality. We further compare the X-ray morphologies of X-ray- and SZ-selected clusters with those of simulated clusters. We do not find a statistically significant difference in the measured X-ray morphology of X-ray and SZ-selected clusters over the redshift range probed by these samples, suggesting that the two are probing similar populations of clusters. We find that the X-ray morphologies of simulated clusters are statistically indistinguishable from those of X-ray- or SZ-selected clusters, implying that the most important physics for dictating the large-scale gas morphology (outside of the core) is well-approximated in these simulations. Finally, we find no statistically significant redshift evolution in the X-ray morphology (both for observed and simulated clusters), over the range of z∼0.3 to z∼1, seemingly in contradiction with the redshift-dependent halo merger rate predicted by simulations.

79 ASTRONOMY AND ASTROPHYSICS↗

Low surface brightness galaxies selected from the 40% sky area of the ALFALFA H i Survey. I. Sample and statistical properties

The population of low surface brightness (LSB) galaxies, which are objects with central surface brightnesses at least one magnitude fainter than the night sky, is crucial for understanding the extremes of galactic formation and evolution of the universe. As LSB galaxies are mostly rich in gas (H i), the α.40 Sloan Digital Sky Survey (SDSS) Data Release 7 (DR7) sample is one of the best survey combinations to select a sample of them in the local universe. Since the sky backgrounds are systematically overestimated for galactic images by the SDSS photometric pipeline, particularly for luminous galaxies or galaxies with extended LSB outskirts, in this paper, we above all estimated the sky backgrounds of SDSS images accurately in both the g and r bands for each galaxy in the α.40 SDSS DR7 sample, using a precise method of sky subtraction. Once subtracting the sky background, we did surface photometry with the Kron elliptical aperture using the SExtractor software and fitted geometric parameters with an exponential profile model using the Galfit software for each galactic image in the α.40 SDSS DR7 sample. Based on the photometric and geometric results, we further calculated the B-band central surface brightness, μ{sub 0}(B), for each galaxy and ultimately defined a sample of LSB galaxies consisting of 1129 galaxies with μ{sub 0}(B) > 22.5 mag arcsec{sup −2} and the axis ratio b/a > 0.3 from the 12,423 α.40 SDSS DR7 galaxies. This H i-selected sample of LSB galaxies is a relatively unbiased sample of gas-rich and disk-dominated LSB galaxies, which is complete both in H i observation and the optical magnitude within the limit of the SDSS DR7 photometric survey. This LSB galactic sample spans from 22.5 to 28.3 in μ{sub 0}(B) with a fraction of 4% fainter than 25.0 mag arcsec{sup −2} in B-band central surface brightness and distributes from −27.0 to −12.3 mag in absolute magnitude in the B band (M(B)), including the 43 faintest galaxies (M(B) > −17.3 mag). This sample is a blue LSB galactic sample, of which 98% are bluer than 0.75 mag in B–V color. As for 21 cm H i properties, our LSB galactic sample has a median M{sub H i}/L{sub B} = 0.87, and a large portion (95%) has high mass of H i (M(H i) > 10{sup 7.7}M⊙), which supports galaxies in this LSB galactic sample being mostly gas-rich. Additionally, we statistically investigated the environment of our LSB galaxies and found that up to 92% of the total LSB galaxies have less than eight neighboring galaxies, which provides strong evidence that LSB galaxies prefer to reside in the low-density environment.

79 ASTRONOMY AND ASTROPHYSICS↗

The galaxy–halo connection of emission-line galaxies in IllustrisTNG

ABSTRACT We employ the hydrodynamical simulation IllustrisTNG-300-1 to explore the halo occupation distribution (HOD) and environmental dependence of luminous star-forming emission-line galaxies (ELGs) at z ∼ 1. Such galaxies are key targets for current and upcoming cosmological surveys. We select model galaxies through cuts in colour–colour space allowing for a direct comparison with the Extended Baryon Oscillation Spectroscopic Survey and the Dark Energy Spectroscopic Instrument (DESI) surveys and then compare them with galaxies selected based on specific star formation rate (sSFR) and stellar mass. We demonstrate that the ELG populations are twice more likely to reside in lower density regions (sheets) compared with the mass-selected populations and twice less likely to occupy the densest regions of the cosmic web (knots). We also show that the colour-selected and sSFR-selected ELGs exhibit very similar occupation and clustering statistics, finding that the agreement is best for lower redshifts. In contrast with the mass-selected sample, the occupation of haloes by a central ELG peaks at ∼20 per cent. We furthermore explore the dependence of the HOD and the autocorrelation on environment, noticing that at fixed halo mass, galaxies in high-density regions cluster about 10 times more strongly than low-density ones. This result suggests that we should model carefully the galaxy–halo relation and implement assembly bias effects into our models (estimated at ∼4 per cent of the clustering of the DESI colour-selected sample at z = 0.8). Finally, we apply a simple mock recipe to recover the clustering on large scales (r ≳ 1 Mpc h−1) to within 1 per cent by augmenting the HOD model with an environment dependence, demonstrating the power of adopting flexible population models.

79 ASTRONOMY AND ASTROPHYSICS↗

Testing general relativity using Bayesian model selection: Applications to observations of gravitational waves from compact binary systems

Second-generation interferometric gravitational-wave detectors, such as Advanced LIGO and Advanced Virgo, are expected to begin operation by 2015. Such instruments plan to reach sensitivities that will offer the unique possibility to test general relativity in the dynamical, strong-field regime and investigate departures from its predictions, in particular, using the signal from coalescing binary systems. We introduce a statistical framework based on Bayesian model selection in which the Bayes factor between two competing hypotheses measures which theory is favored by the data. Probability density functions of the model parameters are then used to quantify the inference on individual parameters. We also develop a method to combine the information coming from multiple independent observations of gravitational waves, and show how much stronger inference could be. As an introduction and illustration of this framework-and a practical numerical implementation through the Monte Carlo integration technique of nested sampling-we apply it to gravitational waves from the inspiral phase of coalescing binary systems as predicted by general relativity and a very simple alternative theory in which the graviton has a nonzero mass. This method can (and should) be extended to more realistic and physically motivated theories.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Audit Report on "Management Controls over the Department's Excess Weapons Inventories and Selected Sensitive Equipment used by Protective Forces"

Since September 11, 2001, the Department of Energy has, on several occasions, revised its security posture based on identified threats and adversaries. These revisions in security posture have driven Departmental sites to upgrade their defensive and tactical equipment. Subsequent changes in the perceived threats have, in some cases, led to a reduction in the need for certain types of weapons, thus creating a pool of surplus equipment. These surplus weapons could potentially be used by other Department sites and Federal law enforcement agencies. Recent Office of Inspector General reports have raised concerns with the adequacy of controls related to defensive and tactical equipment. For example, our report on Management Controls Over Defense Related High Risk Property (OAS-M-08-06, April 2008) found that administrative controls over certain defense related high risk property were not sufficient for providing accountability over these items. Because of prior reported weaknesses in controls over defensive and tactical equipment, we initiated this audit to determine whether the Department and its contractors were properly managing excess weapons inventories and selected sensitive equipment used by protective forces. Our review disclosed that the Department was not always properly managing its inventories of excess weapons and selected sensitive equipment. We identified issues with the retention of unneeded weapons at many locations and with the identification and tracking of sensitive items. More specifically: Sites maintained large inventories of weapons that were no longer needed but had not been made available for use by either other Departmental sites or other Federal law enforcement agencies. For instance, at six of the locations included in our review we identified a total of 2,635 unneeded weapons with a total acquisition value of over $2.8 million that had not been officially declared as excess - an action that would have made them available for others to use. In addition; Sites were not always identifying, tracking and properly disposing of potentially high risk and sensitive equipment. In particular, we identified control weaknesses in this area related to weapons sights and scopes. These issues occurred because the Department did not have processes in place to properly manage excess inventories of weapons. In particular, the Department does not have requirements for ensuring timely declaration of excess weapons. Additionally, certain sites indicated that they were unwilling to give up excess weapons because of the possibility that they may be needed in the future. However, other sites had a need for some of these weapons and could have avoided purchasing them had they been made available through the excess screening process. Also, we found that the Department lacks clear guidance on the identification of high risk/sensitive equipment. Except for immaterial differences, we were able to locate and verify accountability over the items of defensive and tactical equipment we selected for review. Specifically, we took statistical samples of weapons, ammunition, and other related equipment and were able to verify their existence. While these accountability measures were noteworthy, additional action is necessary to strengthen controls over weapon and sensitive equipment management. Untimely declaration of excess weapons may result in an inefficient use of scarce Government resources. Similarly, if selected high risk/sensitive equipment is not properly categorized and tracked, accountability issues may occur. To address these issues, we made recommendations aimed at improving the management of these categories of defensive and tactical equipment.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

A RADIO-SELECTED SAMPLE OF GAMMA-RAY BURST AFTERGLOWS

We present a catalog of radio afterglow observations of gamma-ray bursts (GRBs) over a 14 year period from 1997 to 2011. Our sample of 304 afterglows consists of 2995 flux density measurements (including upper limits) at frequencies between 0.6 GHz and 660 GHz, with the majority of data taken at 8.5 GHz frequency band (1539 measurements). We use this data set to carry out a statistical analysis of the radio-selected sample. The detection rate of radio afterglows has stayed unchanged almost at 31% before and after the launch of the Swift satellite. The canonical long-duration GRB radio light curve at 8.5 GHz peaks at three to six days in the source rest frame, with a median peak luminosity of 10{sup 31} erg s{sup -1} Hz{sup -1}. The peak radio luminosities for short-hard bursts, X-ray flashes, and the supernova-GRB classes are an order of magnitude or more fainter than this value. There are clear relationships between the detectability of a radio afterglow and the fluence or energy of a GRB, and the X-ray or optical brightness of the afterglow. However, we find few significant correlations between these same GRB and afterglow properties and the peak radio flux density. We also produce synthetic light curves at centimeter and millimeter bands using a range of blast wave and microphysics parameters derived from multiwavelength afterglow modeling, and we use them to compare to the radio sample. Finding agreement, we extrapolate this behavior to predict the centimeter and millimeter behavior of GRBs observed by the Expanded Very Large Array and the Atacama Large Millimeter Array.

79 ASTRONOMY AND ASTROPHYSICS↗

Selecting registration schemes in case of interstitial lung disease follow-up in CT

Purpose: Primary goal of this study is to select optimal registration schemes in the framework of interstitial lung disease (ILD) follow-up analysis in CT. Methods: A set of 128 multiresolution schemes composed of multiresolution nonrigid and combinations of rigid and nonrigid registration schemes are evaluated, utilizing ten artificially warped ILD follow-up volumes, originating from ten clinical volumetric CT scans of ILD affected patients, to select candidate optimal schemes. Specifically, all combinations of four transformation models (three rigid: rigid, similarity, affine and one nonrigid: third order B-spline), four cost functions (sum-of-square distances, normalized correlation coefficient, mutual information, and normalized mutual information), four gradient descent optimizers (standard, regular step, adaptive stochastic, and finite difference), and two types of pyramids (recursive and Gaussian-smoothing) were considered. The selection process involves two stages. The first stage involves identification of schemes with deformation field singularities, according to the determinant of the Jacobian matrix. In the second stage, evaluation methodology is based on distance between corresponding landmark points in both normal lung parenchyma (NLP) and ILD affected regions. Statistical analysis was performed in order to select near optimal registration schemes per evaluation metric. Performance of the candidate registration schemes was verified on a case sample of ten clinical follow-up CT scans to obtain the selected registration schemes. Results: By considering near optimal schemes common to all ranking lists, 16 out of 128 registration schemes were initially selected. These schemes obtained submillimeter registration accuracies in terms of average distance errors 0.18 ± 0.01 mm for NLP and 0.20 ± 0.01 mm for ILD, in case of artificially generated follow-up data. Registration accuracy in terms of average distance error in clinical follow-up data was in the range of 1.985–2.156 mm and 1.966–2.234 mm, for NLP and ILD affected regions, respectively, excluding schemes with statistically significant lower performance (Wilcoxon signed-ranks test, p < 0.05), resulting in 13 finally selected registration schemes. Conclusions: Selected registration schemes in case of ILD CT follow-up analysis indicate the significance of adaptive stochastic gradient descent optimizer, as well as the importance of combined rigid and nonrigid schemes providing high accuracy and time efficiency. The selected optimal deformable registration schemes are equivalent in terms of their accuracy and thus compatible in terms of their clinical outcome.

60 APPLIED LIFE SCIENCES↗

Electron Neutrino Selection for Neutrino Oscillation Searches in the SBN Program

The Short- Baseline Neutrino (SBN) Pprogram at Fermilab employs a near detector, SBND, and a far detector, ICARUS, to search for baseline-dependent changes to the measure muon neutrino disappearance and electron neutrino appearance and disappearance from the muon neutrino-dominated Booster Neutrino Beam flux. The use of the same liquid argon time projection chamber (LArTPC) technology in both detectors and the same neutrino beam enables a systematically well-constrained oscillation measurement, by leveraging the high statistics of O(10 million) neutrino interactions expected at SBND through 2027 to constrain flux and cross-section systematics. In this poster, we present high-statistics, high-purity electron neutrino selections in SBND and ICARUS and related electron shower energy calibration measurements in SBND enabled by the SPINE deep learning-based particle reconstruction framework. Furthermore, we discuss how these samples can enable high sensitivity to '3+1' sterile neutrino oscillations in a combined fit accounting for systematic uncertainties and correlations.

Oza, Nupur [Nevis Labs, Columbia U.; Columbia U.]↗

ICARUS at the Short-Baseline Neutrino Program: First Results

First results from ICARUS experiment are presented at FNAL. The selection of nu_mu CC events with 1muon+ N Protons from BNB targeted at numu disappearance analysis is presented for a subset of the collected statistics, , compared with MC predictions. A similar selection of nu_mu CC events 1muon+ N Protons + 0 pions in the NuMI beam aiming at the neutrino-Argon cross section measurement is also presented, together with a control sideband requiring in addition at least a pion candidate. Finally the result of a BSM search for a new particle decaying into two muons is also presented, showing no evidence within the studied sample of new physics.

43 PARTICLE ACCELERATORS↗

Offset active galactic nuclei as tracers of galaxy mergers and supermassive black hole growth

Offset active galactic nuclei (AGNs) are AGNs that are in ongoing galaxy mergers, which produce kinematic offsets in the AGNs relative to their host galaxies. Offset AGNs are also close relatives of dual AGNs. We conduct a systematic search for offset AGNs in the Sloan Digital Sky Survey by selecting AGN emission lines that exhibit statistically significant line-of-sight velocity offsets relative to systemic. From a parent sample of 18,314 Type 2 AGNs at z < 0.21, we identify 351 offset AGN candidates with velocity offsets of 50 km s{sup –1} < |Δv| < 410 km s{sup –1}. When we account for projection effects in the observed velocities, we estimate that 4%-8% of AGNs are offset AGNs. We designed our selection criteria to bypass velocity offsets produced by rotating gas disks, AGN outflows, and gravitational recoil of supermassive black holes, but follow-up observations are still required to confirm our candidates as offset AGNs. We find that the fraction of AGNs that are offset candidates increases with AGN bolometric luminosity, from 0.7% to 6% over the luminosity range 43 < log (L{sub bol}) [erg s{sup –1}] <46. If these candidates are shown to be bona fide offset AGNs, then this would be direct observational evidence that galaxy mergers preferentially trigger high-luminosity AGNs. Finally, we find that the fraction of AGNs that are offset AGN candidates increases from 1.9% at z = 0.1 to 32% at z = 0.7, in step with the growth in the galaxy merger fraction over the same redshift range.

79 ASTRONOMY AND ASTROPHYSICS↗

Full-stack Quantification of Variability in Predicting Ion Transport Properties using Machine-learned Interatomic Potentials

Machine-learned interatomic potentials (MLIPs) have become the state-of-the-art for performing accurate, scalable molecular dynamics (MD) simulations. It is therefore crucial to understand and quantify the reliability of MLIPs for downstream property predictions. Uncertainty in predicted properties can arise from limitations in first-principles training data, intrinsic MLIP model errors in representing the data, and the statistical noise introduced during subsequent MD simulations. Using ion transport in Li7P3S11 as a case study, we systematically assess the impact of training set size and selection, neural network stochasticity, and MD sampling statistics on predicted diffusivity and activation energy. We find that when using equivariant MLIP architectures with standard MD protocols, uncertainty arising from MD sampling dominates over model-induced errors. In contrast, MLIP errors relative to the underlying first-principles data are consistently minor. Given this, there are two main routes to improving the accuracy of predictions based on MLIP potentials: adopting higher accuracy reference data generation methods, and improving the MD sampling statistics.

36 MATERIALS SCIENCE↗

The Galaxy–Halo Connection of DESI Luminous Red Galaxies with Subhalo Abundance Matching

Abstract We use subhalo abundance and age distribution matching to create magnitude-limited mock galaxy catalogs at z ∼ 0.43, 0.52, and 0.63 with z -band and 3.4 μ m W 1-band absolute magnitudes and r − z and r − W 1 colors. From these magnitude-limited mocks, we select mock luminous red galaxy (LRG) samples according to the ( r − z )-based (optical) and ( r − W 1)-based (infrared) selection criteria for the LRG sample of the Dark Energy Spectroscopic Instrument (DESI) survey. Our models reproduce the number densities, luminosity functions, color distributions, and projected clustering of the DESI Legacy Surveys that are the basis for DESI LRG target selection. We predict the halo occupation statistics of both optical and IR DESI LRGs at fixed cosmology and assess the differences between the two LRG samples. We find that IR-based SHAM modeling represents the differences between the optical and IR LRG populations better than using the z band and that age distribution matching overpredicts the clustering of LRGs, implying that galaxy color is uncorrelated with halo age in the LRG regime. Both the optical and IR DESI LRG target selections exclude some of the most luminous galaxies that would appear to be LRGs based on their position on the red sequence in optical color–magnitude space. Both selections also yield populations with a nontrivial LRG–halo connection that does not reach unity for the most massive halos. We find that the IR selection achieves greater completeness (≳90%) than the optical selection across all redshift bins studied.

Berti, Angela M. (ORCID:0000000335826649)↗