SEARCH · Search NASA
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.
Our goal for this summer is to use various neural network models to classify gravitational lensing in simulated LSST images. We focus on maximizing the efficiency and accuracy of our models for training and testing using fast inference techniques on FPGA and TPU hardware.
Not provided.
Saturn is the first giant planet to be visited by an orbiting spacecraft that can transmit large amounts of data to Earth. Crater counts on satellites from Phoebe inward to the regular satellites and ring moons are providing unprecedented insights into the origin and time histories of the impacting populations. Many Voyager-era scientists concluded that the satellites had been struck by at least two populations of impactors. In this view, the Population I impactors, which were generally judged to be comets orbiting the Sun, formed most of the larger and older craters, while Population II impactors, interpreted as Saturn-orbiting ejecta from impacts on satellites, produced most of the smaller and younger craters. Voyager data also implied that all of the ring moons, and probably some of the midsized classical moons, had been catastrophically disrupted and reaccreted since they formed. We examine models of the primary impactor populations in the Saturn system. At the present time, ecliptic comets, which likely originate in the Kuiper belt/scattered disk, are predicted to dominate impacts on the regular satellites and ring moons, but the models require extrapolations in size (from the observed Kuiper belt objects to the much smaller bodies that produce the craters) or in distance (from the known active Jupiter family comets to 9.5 AU). Phoebe, Iapetus, and perhaps even moons closer to Saturn have been struck by irregular satellites as well. We describe the Nice model, which provides a plausible mechanism by which the entire Solar System might have experienced an era of heavy bombardment long after the planets formed. We then discuss the three cratering chronologies, including one based upon the Nice model, that have been used to infer surface ages from crater densities on the saturnian satellites. After reviewing scaling relations between the properties of impactors and the craters they produce, we provide model estimates of the present-day rate at which comets impact, and catastrophically disrupt, the saturnian moons. Finally, we present crater counts on the satellites from two different groups. Many of the heavily cratered terrains appear to be nearly saturated, so it is difficult to infer the provenance of the impactors from crater counts alone. More large craters have been found on Iapetus than on any other satellite. Enceladus displays an enormous range of surface ages, ranging from the old mid-latitude plains to the extremely young South Polar Terrain. Cassini images provide some evidence for the reality of Population II. Most of the observed craters may have formed in one or more cataclysms, but more work is needed to determine the roles of heliocentric and planetocentric bodies in creating the craters.
Post-training model compression can reduce the inference costs of deep neural networks, but uncompressed training still consumes enormous hardware resources and energy. To enable low-energy training on edge devices, it is highly desirable to directly train a compact neural network from scratch with a low memory cost. Low-rank tensor decomposition is an effective approach to reduce the memory and computing costs of large neural networks. However, directly training low-rank tensorized neural networks is a very challenging task because it is hard to determine a proper tensor rank a priori, and the tensor rank controls both model complexity and accuracy. Here, this paper presents a novel end-to-end framework for low-rank tensorized training. We first develop a Bayesian model that supports various low-rank tensor formats (e.g., CANDECOMP/PARAFAC, Tucker, tensor-train, and tensor-train matrix) and reduces neural network parameters with automatic rank determination during training. Then we develop a customized Bayesian solver to train large-scale tensorized neural networks. Our training methods shows orders-of-magnitude parameter reduction and little accuracy loss (or even better accuracy) in the experiments. On a very large deep learning recommendation system with over 4.2 ×10 9 model parameters, our method can reduce the parameter number to 1.6 ×10 5 automatically in the training process (i.e., by 2.6 ×10 4 times) while achieving almost the same accuracy. Code is available at https://github.com/colehawkins/bayesian-tensor-rank-determination.
Optical collective Thomson scattering provides precise density and temperature measurements in numerous plasma-physics experiments. The accuracy of such measurements depends on the core assumption that the underlying electron distribution functions in under-dense laser-produced plasmas are Maxwellian. A statistically based, quantitative analysis of the errors in the measured electron density and temperature is presented when synthetic data calculated using a non-Maxwellian electron distribution function is fit assuming a Maxwellian electron distribution. Such analysis can lead to errors of up to 50% in temperature and 30% in density, in the specific case of super-Gaussian distributions characteristic of inverse bremsstrahlung heating. Furthermore including the proper family of non-Maxwellian electron distribution functions, as a fitting parameter, in Thomson-scattering analysis removes the model-dependent errors in the inferred parameters at a minimal cost to the statistical uncertainty.
Freeway crash prediction models are the basic of traffic safety research, yet crash occurrence and the influencing factors change over time. In order to make sure the implemented safety models fit the current traffic environment, this study conducts a comparative analysis of 2017 and 2020 datasets collected from freeways in Suzhou, China. Herein, considering the spatial correlation among analysis units and the hierarchical data structure, a Bayesian conditional autoregressive negative binomial (CAR-NB) model and a Bayesian hierarchical CAR-NB (HCAR-NB) model were used to explore the safety influencing factors, and a traditional NB model was developed for further comparison. To update the HCAR-NB model from 2017 to 2020, Bayesian inference with informative priors was used to improve its goodness of fit and efficiency. Preliminary results showed that 1) the HCAR-NB model outperformed the NB model and CAR-NB model in prediction accuracy, and 2) the number of crashes was significantly correlated with average speed, speed variance, road segment length, number of lanes, and presence of ramps. The potential for safety improvement (PSI) method was applied to the modeling results to identify hotspots for the two years. The results confirmed that the hotspots spatiotemporally shifted among the freeways. The proposed crash prediction model and updating method are expected to assist implementation of informed countermeasures for freeway safety improvement.
During the Subsonic Contrail and Cloud Effect Special Study (SUCCESS) a Solar Spectral Flux Radiometer was deployed at the surface in a zenith observing position. The instrument measured the solar spectral downwelling irradiance between 350 and 2500 nm with 10 nm resolution. From April 12 through April 29 approximately 18000 spectra were acquired, under a variety of meteorological conditions including cloud free, cirrus, Stearns, and cumulonimbus clouds. This study focuses on the effect of cirrus and cirrus contrails on the spectral distribution of solar irradiance at the surface and on inferring cirrus properties from their spectral transmittance. The observations have also proven to be useful for comparing the solar spectral irradiance measurements with model predictions, and in particular, for inferring the amount of solar radiation absorbed in the clear and cloudy atmosphere.
Employing deep-learning and artificial-intelligence (AI) techniques onboard spacecraft can dramatically improve priority data selection to ensure more effective use of the available downlink. However, deployment of effective deep-learning models requires significant training on the ground, which may not be feasible, due to limited data available in an unexplored environment. Therefore, this research explores building robust classification models for onboard data processing where training data is highly limited using transfer-learning techniques. In this paper, we focus on the use case of hyperspectral imaging for remote sensing, a domain where the high dimensionality of the data from the sensor can rapidly saturate the downlink bandwidth. With this bottleneck, there is an impending need to autonomously and robustly classify data onboard to optimize downlink of high-impact measurements, thus maximizing the scientific utility per bit transmitted to the ground. This paper examines the use of deep neural networks onboard for hyperspectral image classification in a communication-limited scenario to analyze how the models perform with limited training data. The use of transfer learning can ameliorate the issue of poor generalization by transferring features learned from training on a large source dataset for one classification task to the target classification task with limited training data. For two deep-learning models from literature, we compare the accuracy of the models trained using transfer learning to models trained from scratch using a random weight initialization with varying amounts of training data. We demonstrate the feasibility and performance of running inference of the deep-learning models on representative flight-like hardware.
As a result of their formation via massive single and binary stellar evolution, the masses of stellar-remnant black holes (BH) are subjects of great interest in this era of gravitational-wave detection from binary black hole (BBH) and binary neutron star merger events. In this work, we present new developments in the stellar-remnant formation and related schemes of the current N-body evolution program NBODY7. We demonstrate that the newly implemented stellar-wind and remnant-formation schemes in the stellar-evolutionary sector or BSE of the NBODY7 code, such as the “rapid” and the “delayed” supernova (SN) schemes along with an implementation of pulsational-pair-instability and pair-instability supernova (PPSN/PSN), now produce neutron star (NS) and BH masses that agree nearly perfectly, over large ranges of zero-age-main-sequence (ZAMS) mass and metallicity, with those from the widely recognised StarTrack population-synthesis program. We also demonstrate the new, recipe-based implementations of various widely debated mechanisms of natal kicks on NSs and BHs, such as “convection-asymmetry-driven”, “collapse-asymmetry-driven”, and “neutrino-emission-driven” kicks, in addition to a fully consistent implementation of the standard, fallback-dependent, momentum-conserving natal kick. All the above newly implemented schemes are also shared with the standalone versions of SSE and BSE. Furthermore, all these demonstrations are performed with both the updated standalone BSE and the updated NBODY7/BSE. When convolved with stellar and primordial-binary populations as observed in young massive clusters, such remnant-formation and natal-kick mechanisms crucially determine the accumulated number, mass, and mass distribution of the BHs retained in young massive, open, and globular clusters (GCs); these BHs would eventually become available for long-term dynamical processing. Among other conclusions, we find that although the newer, delayed SN remnant formation model gives birth to the largest number (mass) of BHs, the older remnant-formation schemes cause the largest number (mass) of BHs to survive in clusters, when incorporating SN material fallback onto the BHs. The SN material fallback also causes the convection-asymmetry-driven SN kick to effectively retain similar numbers and masses of BHs in clusters as for the standard, momentum-conserving kick. The collapse-asymmetry-driven SN kick would cause nearly all BHs to be retained in clusters irrespective of their mass, remnant-formation model, and metallicity, whereas the inference of a large population of BHs in GCs would potentially rule out the neutrino-driven SN kick mechanism. Pre-SN mergers of massive primordial binaries would potentially cause BH masses to deviate from the theoretical, single-star ZAMS to mass-remnant mass relation unless a substantial of the total merging stellar mass of up to ≈40% is lost during a merger process. In particular, such mergers, at low metallicities, have the potential to produce low-spinning BHs within the PSN mass gap that can be retained in a stellar cluster and be available for subsequent dynamical interactions. As recent studies indicate, the new remnant-formation modelling reassures us that young massive and open clusters would potentially contribute to the dynamical BBH merger detection rate to a similar extent as their more massive GC counterparts.
Microbial communities assemble through a complex set of interactions between microbes and their environment, and the resulting metabolic impact on the host ecosystem can be profound. Microbial activity is known to impact human health, plant growth, water quality, and soil carbon storage which has lead to the development of many approaches and products meant to manipulate the microbiome. In order to understand, predict, and improve microbial community engineering, genome-scale modeling techniques have been developed to translate genomic data into inferred microbial dynamics. However, these techniques rely heavily on simulation to draw conclusions which may vary with unknown parameters or initial conditions, rather than more robust qualitative analysis. To better understand microbial community dynamics using genome-scale modeling, we provide a tool to investigate the network of interactions between microbes and environmental metabolites over time. Using our previously developed algorithm for simulating microbial communities from genome-scale metabolic models (GSMs), we infer the set of microbe-metabolite interactions within a microbial community in a particular environment. Because these interactions depend on the available environmental metabolites, we refer to the networks that we infer as metabolically contextualized , and so name our tool MetConSIN: Met abolically Con textualized S pecies I nteraction N etworks.
Thermally emitting X-ray isolated neutron stars (NSs) represent excellent targets for testing cooling surface emission and atmosphere models, which are used to infer the physical parameters of the NS. Among the seven known members of this class, RXJ1605.3+3249 is the only one that still lacks confirmation of its spin period. Here we analyze NICER and XMM-Newton observations of RXJ1605.3+3249, in order to address its timing and spectral behavior. Contrary to a previous tentative detection, but in agreement with the recent work by Pires et al., we find no significant pulsation with a pulsed fraction higher than 1.3% (3σ) for periods above 150 ms. We also find a limit of 2.6% for periods above 2 ms, despite searches in different energy bands. The X-ray spectrum can be fit by either a double-blackbody model or by a single-temperature magnetized atmosphere model, both modified by a Gaussian absorption line at ∼0.44 keV. The origin of the absorption feature as a proton cyclotron line or as an atomic transition in the NS atmosphere is discussed. The predictions of the best-fit X-ray models extended to IR, optical, and UV bands, are compared with archival data. Our results are interpreted in the framework of a fallback disk scenario.
Water resources model development and simulation efforts have seen rapid growth in recent decades to aid evaluations and planning around water scarcity and allocation. Models are typically developed by two distinct communities: (1) large-scale hydrologic modelers emphasizing hydroclimatological processes, and (2) water systems modelers emphasizing environmental, infrastructural, and institutional features that shape water scarcity at the local basin level. This study assesses whether two representative models from these communities produce consistent insights when evaluating the water scarcity vulnerabilities in the Upper Colorado River Basin within the state of Colorado. Results showed that although the regional-scale model [model for scale adaptive river transport (MOSART)—water management (WM)] can capture the aggregate effect of all water operations in the basin, it underestimates the subbasin-scale variability in specific user’s vulnerabilities. The basin-scale water systems model [State of Colorado’s Stream Simulation Model (StateMod)] suggests a larger variance of scarcity across the basin’s water users due to its more detailed accounting of local water allocation infrastructure and institutional processes. This model intercomparison highlights potentially significant limitations of large-scale studies in seeking to evaluate water scarcity and actionable adaptation strategies, as well as ways in which basin-scale water systems model’s information can be used to better inform water allocation and shortage when used in tandem with larger-scale hydrological modeling studies.
Not Available
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Scattering calculations using the discrete dipole approximation and vector radiative transfer calculations were performed to model multiparameter radar return and passive microwave emission for a simple model of a winter storm. The issue of dendrite riming was addressed by computing scattering properties of thin ice disks with varying bulk density. It was shown that C-band multiparameter radar contains information about particle density and the number concentration of the ice particles. The radiative transfer modeling indicated that polarized multifrequency passive microwave emission may be used to infer some properties of ice hydrometers. Detailed radar modeling and vector radiative transfer modeling is in progress to enhance the understanding of simultaneous radar and radiometer measurements, as in the case of the proposed TRMM field program. A one-dimensional cloud model will be used to simulate the storm structure in detail and study the microphysics, such as size and density. Multifrequency polarized radiometer measurements from the SSMI satellite instrument will be analyzed in relation to dual-frequency and dual-polarization radar measurements.
Concentrations of chloropigments and particulate organic carbon (POC) in large-volume in-situ pump samples from the Mediterranean Sea were used to estimate rate constants of processes that control the fate of particles, and specifically chloropigments, in the water column. In this paper we introduce a Bayesian statistical inversion method that combines the data with a new box model and has the capacity to infer rate constants for POC respiration/dissolution, chlorophyll and pheopigment degradation, and particle aggregation and disaggregation. We use first-order kinetics to model disaggregation, and use both first-order and second-order kinetics to model aggregation. Using these methods, the estimated small-particle (1–70 μm) POC respiration rate constant was 2.44 -1.00 +1.69 yr -1 (0.41 yr). The estimated disaggregation and second-order aggregation rate constants were 85.6 -36.4 +63.4 yr -1 (1.17×10 -2 yr) and 2.78 -1.17 +2.01 μM -1 yr -1 , respectively. Using the optimal rate constants and the corresponding particle concentrations, disaggregation is ~4.2 times faster than the small-size POC dissolution rate, which indicates that disaggregation is a dominant processes at the time of sampling. More importantly, by comparing our results with those of previous studies, we conclude that sampling methods have less influence than tracers themselves on inferring particle dynamic rate constants. We previously introduced a somewhat similar approach to modeling SV sediment trap data, but large volume pumps are a much more common sample collection method in oceanographic surveys than SV sediment traps, and thus our new model should have a wider applicability.
Here, we propose a new method for inferring the governing stochastic ordinary differential equations (SODEs) by observing particle ensembles at discrete and sparse time instants, i.e., multiple “snapshots.” Particle coordinates at a single time instant, possibly noisy or truncated, are recorded in each snapshot but are unpaired across the snapshots. By training a physics-informed generative model that generates “fake” sample paths, we aim to fit the observed particle ensemble distributions with a curve in the probability measure space, which is induced from the inferred particle dynamics. We employ different metrics to quantify the differences between distributions, e.g., the sliced Wasserstein distances and the adversarial losses in generative adversarial networks. We refer to this method as generative “ensemble-regression” (GER), in analogy to the classic “point-regression,” where we infer the dynamics by performing regression in the Euclidean space. We illustrate the GER by learning the drift and diffusion terms of particle ensembles governed by SODEs with Brownian motions and Lévy processes up to 100 dimensions. We also discuss how to treat cases with noisy or truncated observations. Apart from systems consisting of independent particles, we also tackle nonlocal interacting particle systems with unknown interaction potential parameters by constructing a physics-informed loss function. Finally, we investigate scenarios of paired observations and discuss how to reduce the dimensionality in such cases by proving a convergence theorem that provides theoretical support.