Search NASA⌕ Search

SEARCH · Search NASA

Results for “deep data”

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 487 records · Page 27

Increasing phosphorus loss despite widespread concentration decline in US rivers

The loss of phosphorous (P) from the land to aquatic systems has polluted waters and threatened food production worldwide. Systematic trend analysis of P, a nonrenewable resource, has been challenging, primarily due to sparse and inconsistent historical data. Here, we leveraged intensive hydrometeorological data and the recent renaissance of deep learning approaches to fill data gaps and reconstruct temporal trends. We trained a multitask long short-term memory model for total P (TP) using data from 430 rivers across the contiguous United States (CONUS). Trend analysis of reconstructed daily records (1980–2019) shows widespread decline in concentrations, with declining, increasing, and insignificantly changing trends in 60%, 28%, and 12% of the rivers, respectively. Concentrations in urban rivers have declined the most despite rising urban population in the past decades; concentrations in agricultural rivers however have mostly increased, suggesting not-as-effective controls of nonpoint sources in agriculture lands compared to point sources in cities. TP loss, calculated as fluxes by multiplying concentration and discharge, however exhibited an overall increasing rate of 6.5% per decade at the CONUS scale over the past 40 y, largely due to increasing river discharge. Results highlight the challenge of reducing TP loss that is complicated by changing river discharge in a warming climate.

Science & Technology - Other Topics↗

Unified description of diffractive deep inelastic scattering with saturation

We propose a new description of inclusive diffraction in deep inelastic scattering. The diffractive structure functions are expressed in the dipole picture and contain heavy-quark contributions. The dipole scattering amplitude, a saturation model fitted on inclusive deep inelastic scattering data, features a saturation scale Q{sub s}(x) larger than 1 GeV for x=10{sup -5}. The qqg contribution to the diffractive final state is modeled in such a way that both the large-Q{sup 2} and small-{beta} limits are implemented. In the regime x{sub P}<0.01 in which saturation is expected to be relevant, we obtain a parameter-free description of the HERA data with {chi}{sup 2}/points=1.2.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Deeply virtual Compton scattering via color dipoles: Nonperturbative effects

We study the deeply virtual Compton scattering amplitude within the color dipole approach. The light-cone wave function of a real photon is evaluated in the instanton vacuum model. Our parameter-free calculations are able to describe H1 data, both the absolute values and the t dependences, at medium-high values of Q{sup 2}. The Q{sup 2} dependence is found to be sensitive to the choice of the phenomenological cross section fitted to deep-inelastic scattering data.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

OPTICAL IMAGES AND SOURCE CATALOG OF AKARI NORTH ECLIPTIC POLE WIDE SURVEY FIELD

We present the source catalog and the properties of the B-, R-, and I-band images obtained to support the AKARI North Ecliptic Pole Wide (NEP-Wide) survey. The NEP-Wide is an AKARI infrared imaging survey of the north ecliptic pole covering a 5.8 deg{sup 2} area over 2.5-6 {mu}m wavelengths. The optical imaging data were obtained at the Maidanak Observatory in Uzbekistan using the Seoul National University 4k x 4k Camera on the 1.5 m telescope. These images cover 4.9 deg{sup 2} where no deep optical imaging data are available. Our B-, R-, and I-band data reach the depths of {approx}23.4, {approx}23.1, and {approx}22.3 mag (AB) at 5{sigma}, respectively. The source catalog contains 96,460 objects in the R band, and the astrometric accuracy is about 0.''15 at 1{sigma} in each R.A. and decl. direction. These photometric data will be useful for many studies including identification of optical counterparts of the infrared sources detected by AKARI, analysis of their spectral energy distributions from optical through infrared, and the selection of interesting objects to understand the obscured galaxy evolution.

79 ASTRONOMY AND ASTROPHYSICS↗

Metabolic potential of fatty acid oxidation and anaerobic respiration by abundant members of Thaumarchaeota and Thermoplasmata in deep anoxic peat

To probe the metabolic potential of abundant Archaea in boreal peats, we reconstructed two near-complete archaeal genomes, affiliated with Thaumarchaeota group 1.1c (bin Fn1, 8% abundance), which was a genomically unrepresented group, and Thermoplasmata (bin Bg1, 26% abundance), from metagenomic data acquired from deep anoxic peat layers. Each of the near-complete genomes encodes the potential to degrade long-chain fatty acids (LCFA) via β-oxidation. Fn1 has the potential to oxidize LCFA either by syntrophic interaction with methanogens or by coupling oxidation with anaerobic respiration using fumarate as a terminal electron acceptor (TEA). Fn1 is the first Thaumarchaeota genome without an identifiable carbon fixation pathway, indicating that this mesophilic phylum encompasses more diverse metabolisms than previously thought. Furthermore, we report genetic evidence suggestive of sulfite and/or organosulfonate reduction by Thermoplasmata Bg1. In deep peat, inorganic TEAs are often depleted to extremely low levels, yet the anaerobic respiration predicted for two abundant archaeal members suggests organic electron acceptors such as fumarate and organosulfonate (enriched in humic substances) may be important for respiration and C mineralization in peatlands.

59 BASIC BIOLOGICAL SCIENCES↗

Quantifying the robustness of deep multispectral segmentation models against natural perturbations and data poisoning

In overhead image segmentation tasks, including additional spectral bands beyond the traditional RGB channels can improve model performance. However, it is still unclear how incorporating this additional data impacts model robustness to adversarial attacks and natural perturbations. For adversarial robustness, the additional in-formation could improve the model’s ability to distinguish malicious inputs, or simply provide new attack avenues and vulnerabilities. For natural perturbations, the additional information could better inform model decisions and weaken perturbation effects or have no significant influence at all. In this work, we seek to characterize the performance and robustness of a multispectral (RGB and near infrared) image segmentation model subjected to adversarial attacks and natural perturbations. While existing adversarial and natural robustness research has focused primarily on digital perturbations, we prioritize on creating realistic perturbations designed with physical world conditions in mind. For adversarial robustness, we focus on data poisoning attacks whereas for natural robustness, we focus on extending ImageNet-C common corruptions for fog and snow that coherently and self-consistently perturbs the input data. Overall, we find both RGB and multispectral models are vulnerable to data poisoning attacks regardless of input or fusion architectures and that while physically-realizable natural perturbations still degrade model performance, the impact differs based on fusion architecture and input data.

Deep learning, multispectral images, multimodal fu↗

Morphological Characters Can Strongly Influence Early Animal Relationships Inferred from Phylogenomic Data Sets

There are considerable phylogenetic incongruencies between morphological and phylogenomic data for the deep evolution of animals. This has contributed to a heated debate over the earliest-branching lineage of the animal kingdom: the sister to all other Metazoa (SOM). Here, we use published phylogenomic data sets ($\sim $45,000–400,000 characters in size with $\sim $15–100 taxa) that focus on early metazoan phylogeny to evaluate the impact of incorporating morphological data sets ($\sim $15–275 characters). We additionally use small exemplar data sets to quantify how increased taxon sampling can help stabilize phylogenetic inferences. We apply a plethora of common methods, that is, likelihood models and their “equivalent” under parsimony: character weighting schemes. Our results are at odds with the typical view of phylogenomics, that is, that genomic-scale data sets will swamp out inferences from morphological data. Instead, weighting morphological data 2–10$\times $ in both likelihood and parsimony can in some cases “flip” which phylum is inferred to be the SOM. This typically results in the molecular hypothesis of Ctenophora as the SOM flipping to Porifera (or occasionally Placozoa). However, greater taxon sampling improves phylogenetic stability, with some of the larger molecular data sets ($>$200,000 characters and up to $\sim $100 taxa) showing node stability even with $\geqq100\times $ upweighting of morphological data. Accordingly, our analyses have three strong messages. 1) The assumption that genomic data will automatically “swamp out” morphological data is not always true for the SOM question. Morphological data have a strong influence in our analyses of combined data sets, even when outnumbered thousands of times by molecular data. Morphology therefore should not be counted out a priori. 2) We here quantify for the first time how the stability of the SOM node improves for several genomic data sets when the taxon sampling is increased. 3) The patterns of “flipping points” (i.e., the weighting of morphological data it takes to change the inferred SOM) carry information about the phylogenetic stability of matrices. The weighting space is an innovative way to assess comparability of data sets that could be developed into a new sensitivity analysis tool.

59 BASIC BIOLOGICAL SCIENCES↗

ESM data downscaling: a comparison of super-resolution deep learning models

Abstract Climate projections at fine spatial resolutions are required to conduct accurate risk assessment for critical infrastructure and design adaptation planning. Generating these projections using advanced Earth system models (ESM) requires significant computational resources. To address this issue, various statistical downscaling techniques have been introduced to generate fine-resolution data from coarse-resolution simulations. In this study, we evaluate and compare five deep learning-based downscaling techniques, namely, super-resolution convolutional neural networks, fast super-resolution convolutional neural network ESM, efficient sub-pixel convolutional neural network, enhanced deep residual network (EDRN), and super-resolution generative adversarial network (SRGAN). These techniques are applied to a dataset generated by the Energy Exascale Earth System Model (E3SM), focusing on key surface variables such as surface temperature, shortwave heat flux, and longwave heat flux. Models are trained and validated using paired fine-resolution (0.25 $$^{\circ }$$ ∘ ) and coarse-resolution (1 $$^{\circ }$$ ∘ ) monthly data obtained from a 9-year simulation. Next, blind testing is performed using monthly data obtained from two different years outside of the training and validation set. To evaluate the efficiency of each technique, different statistical metrics are used, including mean squared error (MSE), peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), and learned perceptual image patch similarity (LPIPS). The results show that EDRN outperforms other algorithms in terms of PSNR, SSIM, and MSE, but struggles to capture fine-scale features in the data. In contrast, SRGAN, a generative model that uses perceptual loss, excels in capturing fine details at boundaries and internal structures, resulting in lower LPIPS than other methods.

Pawar, Nikhil M. (ORCID:0000000211613289)↗

Ionizing radiation from AGNs at z > 3.3 with the Subaru Hyper Suprime-Cam Survey and the CFHT Large Area U-band Deep Survey (CLAUDS)

ABSTRACT We use deep and wide imaging data from the CFHT Large Area U-band Deep Survey (CLAUDS) and the Hyper Suprime-Cam Subaru Strategic Program (HSC-SSP) to constrain the ionizing radiation (Lyman Continuum, LyC) escape fraction from active galactic nuclei (AGNs) at z ∼ 3–4. For 94 AGNs with spectroscopic redshifts at 3.3 < z < 4.0, we use their U/i-band flux ratios to estimate LyC transmission of individual AGNs. The distribution of their LyC transmission shows values lower than the range of LyC transmission values for IGM of the same redshift range, which suggests that LyC escape fraction of AGNs at z > 3.3 is considerably lower than unity in most cases. We do not find any trend in LyC transmission values depending on their UV luminosities. Based on the photometry of stacked images we find the average flux ratio of LyC and non-ionizing UV photons escaping from the objects (fLyC/fUV)out = 0.182 ± 0.043 for AGNs at 3.3 < z < 3.6, which corresponds to LyC escape fraction fesc = 0.303 ± 0.072 if we assume a fiducial intrinsic SED of AGN. Based on the estimated LyC escape fraction and the UV luminosity function (UVLF) of AGNs, we argue that UV-selected AGNs’ contribution to the LyC emissivity at the epoch is minor, although the size of their contribution largely depends on the shape of the UVLF.

79 ASTRONOMY AND ASTROPHYSICS↗

Synthetic galaxy clusters and observations based on Dark Energy Survey Year 3 Data

ABSTRACT We develop a novel data-driven method for generating synthetic optical observations of galaxy clusters. In cluster weak lensing, the interplay between analysis choices and systematic effects related to source galaxy selection, shape measurement, and photometric redshift estimation can be best characterized in end-to-end tests going from mock observations to recovered cluster masses. To create such test scenarios, we measure and model the photometric properties of galaxy clusters and their sky environments from the Dark Energy Survey Year 3 (DES Y3) data in two bins of cluster richness $\lambda \in [30; 45)$, $\lambda \in [45; 60)$ and three bins in cluster redshift ($z\in [0.3; 0.35)$, $z\in [0.45; 0.5)$ and $z\in [0.6; 0.65)$. Using deep-field imaging data, we extrapolate galaxy populations beyond the limiting magnitude of DES Y3 and calculate the properties of cluster member galaxies via statistical background subtraction. We construct mock galaxy clusters as random draws from a distribution function, and render mock clusters and line-of-sight catalogues into synthetic images in the same format as actual survey observations. Synthetic galaxy clusters are generated from real observational data, and thus are independent from the assumptions inherent to cosmological simulations. The recipe can be straightforwardly modified to incorporate extra information, and correct for survey incompleteness. New realizations of synthetic clusters can be created at minimal cost, which will allow future analyses to generate the large number of images needed to characterize systematic uncertainties in cluster mass measurements.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Dynamic Behavior of Natural Seep Vents: Analysis of Field and Laboratory Observations and Modeling (Final Scientific/Technical Report)

In this project, we have analyzed data collected by the U.S. Department of Energy (DOE), National Energy Technology Laboratory (NETL) in a high pressure water tunnel (HPWT) and data from two research cruises to natural seeps in the Gulf of Mexico to adapt and validate a numerical model to predict the dynamics of natural seeps in the deep oceans. The HPWT data include video observations of the shrinkage rate of individual methane and natural gas bubbles under simulated deep-water conditions. Field data were collected during two cruises by the Gulf Integrated Spill Research (GISR) Consortium led by Texas A&M University and funded by the Gulf of Mexico Research Initiative (GoMRI). These data included in situ observations from a remotely operated vehicle (ROV) of gas bubbles at two natural seep sites in the Gulf and acoustic observations of the natural seep bubble flares in the ocean water column. The acoustic data were from multibeam echosounders, one mounted in a forward-looking orientation on the ROV and another mounted down-looking in the haul of the ship. All of these laboratory and field data were focused on the dynamics of natural gas bubbles at temperatures and pressures favorable for clathrate hydrate formation between the gas and water. Our analyses of this data focused on understanding the mechanisms responsible for gas bubble dissolution within the hydrate stability zone (HSZ) of the oceans. Ice-like hydrate shells may form on the bubble-water interface under these conditions, and it was unknown how this might affect the mass transfer of gas into the ocean. We were able to extract bubble shrinkage rates from the HPWT datasets. Using this data we determined that mass transfer coefficients with and without a hydrate shell match empirical values for bubbles in contaminated systems (so-called dirty bubbles contaminated by naturally occurring surfactants). We also showed that free gas, and not gas hydrate, is the dominant dissolving phase when the hydrate sub-cooling is below 11 degree Celsius (temperature difference between hydrate the hydrate formation temperature and ambient temperator) or the pressure is reducing as bubbles rise through the ocean water column. Using this mass transfer model, our numerical model of bubble dissolution matched the over 200 HPWT experiments with an average error of 10% for predicting the bubble size at the end of an experiment. From field data in the literature, we also observed that gas bubbles dissolve faster when they are initially released, following mass transfer coefficients for so-called clean-bubbles (those not yet contaminated by surfactants). Shortly after release within the HSZ, a hydrate shell forms on the bubble-water interface, and the mass transfer reduces to rates matching those of dirty bubbles. We correlated this transition time from clean to dirty bubble behavior with the initial bubble surface area and the hydrate sub-cooling. With this model for hydrate formation time and using the mass transfer coefficients deduced from the HPWT data, we validated our numerical model for predicting the rise heights of natural seep flares in the oceans. Flare heights are commonly observed in haul-mounted acoustic multibeam data. The numerical model predicts bubbles to rise high in the ocean water column owing to the slower mass transfer rates for dirty bubbles that accompany the majority of their rise time. We found that the numerical model predictions matched the observed flare heights within 5% to 10% accuracy when we compared the rise heights of the largest bubbles released from the seafloor with the bubbles acoustically visible in the multibeam data. Bubbles become acoustically transparent as they shrink to sizes of order 1 mm in diameter for the multibeam frequencies used in the field. The forward-looking multibeam on the ROV also provided data on the lateral spreading of bubbles in natural seep flares. Our analysis of this data showed that spreading follows a diffusion process, with the effective diffusivity correlating with the wobbling length scale of these ellipsoidal bubbles. When we apply this diffusivity in a random displacement model of bubble spreading, our numerical simulations match closely the lateral spread observed by the M3 in the ocean water column. Finally, we compared the seep model predictions for the acoustic properties of these natural seep plumes with that observed by the acoustic instruments in the field. The M3 and EM 302 observations were converted to relative values of target strength using a calibration we obtained in the laboratory for the M3 and using an algorithm from the manufacturer for the EM 302. Comparing the numerical seep model to these data, we obtain good agreement over the whole height of rise of these bubble flares. This further validates the numerical model. Overall, our validated seep model captures the key dynamics of gas bubbles released from natural seeps in the oceans and helps to predict the fate of methane in the water column.

03 NATURAL GAS↗

OpenSolar: Promoting the Openness and Accessibility of Diverse Public Solar Datasets

Observational solar data is the foundation of data-driven research in solar power grid integration and power system operations. Compared to other fields in data science, the openness and accessibility of solar data is lacking, which prevents solar data science from catching up with the emerging trends of data science (e.g., deep learning). In this paper, OpenSolar, a package with both R and Python versions, is developed to enhance the openness and accessibility of publicly available solar datasets. The OpenSolar package provides access to multiple types of solar data, primarily from four datasets: (1) the National Renewable Energy Laboratory (NREL) Solar Power Data for Integration Studies dataset, (2) the NREL Solar Radiation Research Laboratory dataset, (3) the Sheffield Solar-Microgen database, and (4) the Dataport database. Unlike other open solar datasets that only contain meteorological data, the four datasets in the OpenSolar package also contain behind-the-meter power data, sky images, and solar power data with satisfactory temporal and spatial resolution and coverage. The overview, quality-control methods, and potential usage of the datasets, in conjunction with sample code implementing the OpenSolar functions, are described in this paper. The package is expected to assist in bridging the gaps between the research fields of solar energy, power systems, and data science.

14 SOLAR ENERGY↗

CANDELS MULTI-WAVELENGTH CATALOGS: SOURCE DETECTION AND PHOTOMETRY IN THE GOODS-SOUTH FIELD

We present a UV to mid-infrared multi-wavelength catalog in the CANDELS/GOODS-S field, combining the newly obtained CANDELS HST/WFC3 F105W, F125W, and F160W data with existing public data. The catalog is based on source detection in the WFC3 F160W band. The F160W mosaic includes the data from CANDELS deep and wide observations as well as previous ERS and HUDF09 programs. The mosaic reaches a 5{sigma} limiting depth (within an aperture of radius 0.''17) of 27.4, 28.2, and 29.7 AB for CANDELS wide, deep, and HUDF regions, respectively. The catalog contains 34,930 sources with the representative 50% completeness reaching 25.9, 26.6, and 28.1 AB in the F160W band for the three regions. In addition to WFC3 bands, the catalog also includes data from UV (U band from both CTIO/MOSAIC and VLT/VIMOS), optical (HST/ACS F435W, F606W, F775W, F814W, and F850LP), and infrared (HST/WFC3 F098M, VLT/ISAAC Ks, VLT/HAWK-I Ks, and Spitzer/IRAC 3.6, 4.5, 5.8, 8.0 {mu}m) observations. The catalog is validated via stellar colors, comparison with other published catalogs, zero-point offsets determined from the best-fit templates of the spectral energy distribution of spectroscopically observed objects, and the accuracy of photometric redshifts. The catalog is able to detect unreddened star-forming (passive) galaxies with stellar mass of 10{sup 10} M{sub Sun} at a 50% completeness level to z {approx} 3.4 (2.8), 4.6 (3.2), and 7.0 (4.2) in the three regions. As an example of application, the catalog is used to select both star-forming and passive galaxies at z {approx} 2-4 via the Balmer break. It is also used to study the color-magnitude diagram of galaxies at 0 < z < 4.

79 ASTRONOMY AND ASTROPHYSICS↗

Applying 3D Geologic Modeling Workflows to the Argillite Reference Case (Rev. 1)

The objective of this short report is to document the application of our 3D geologic modeling workflow to an argillite (shale) host rock. Over the past four years, our team at Los Alamos National Laboratory has developed a geologic modeling workflow that can be applied to generic alluvial basins such as those found in the western United States. In “frontier” or “exploratory” basins where data are sparse, the first steps are to collect, evaluate and integrate available subsurface data into conceptual geologic models. Those models form the basis for constructing the geologic framework model, a 3D geocellular model ideally constrained by seismic and borehole data. To date we have constructed our models using “synthetic” well data derived from conceptual models, without the prospect of validating our workflow using “real” subsurface data. We were tasked to investigate whether our workflow designed for alluvial basin sediments could be applied to other potential repository host rocks. This task also provided the opportunity to work with high-quality subsurface data collected specifically for siting and evaluating a nuclear waste repository. Nagra, the Swiss governmental agency responsible for the disposal of the nation’s radioactive waste, generously provided us with data from two deep boreholes drilled through their argillaceous target formation. The aim of our proof-of-concept demonstration is to evaluate whether geostatistical methods offer a viable approach to property modeling in argillaceous rocks. Nagra provided us with the well data on the condition that we maintain confidentiality with all transferred information and results. Fortunately, Nagra posts numerous technical reports on its public website that describe the subsurface geology in great detail. All of the information and illustrations in this report related to the Swiss repository enterprise are taken from the Nagra public website.

58 GEOSCIENCES↗

FPGA Acceleration of GCN in Light of the Symmetry of Graph Adjacency Matrix

Graph Convolutional Neural Networks (GCNs) are widely used to process large-scale graph data. Different from deep neural networks (DNNs), GCNs are sparse, irregular, and unstructured, posing unique challenges to hardware acceleration with regular processing elements (PEs). In particular, the adjacency matrix of a GCN is extremely sparse, leading to frequent but irregular memory access, low spatial/temporal data locality and poor data reuse. Furthermore, a realistic graph usually consists of unstructured data (e.g., unbalanced distributions), creating significantly different processing times and imbalanced workload for each node in GCN acceleration. To overcome these challenges, we propose an end-to-end hardware-software co-design to accelerate GCNs on resource-constrained FPGAs with the features including: (1) A custom dataflow that leverages symmetry along the diagonal of the adjacency matrix to accelerate feature aggregation for undirected graphs. We utilize either the upper or the lower triangular matrix of the adjacency matrix to perform aggregation in GCN to improve data reuse. (2) Unified compute cores for both aggregation and transform phases, with full support to the symmetry-based dataflow. These cores can be dynamically reconfigured to the systolic mode for transformation or as individual accumulators for aggregation in GCN processing. (3) Preprocessing of the graph in software to rearrange the edges and features to match the custom dataflow. This step improves the regularity in memory access and data reuse in the aggregation phase. Moreover, we quantize the GCN precision from FP32 to INT8 to reduce the memory footprint without losing the inference accuracy. We implement our accelerator design in Intel Stratix10 MX FPGA board with HBM2, and demonstrate 1.3x-110.5x improvement in end-to-end GCN latency as compared to the state-of the-art FPGA implementations, on the graph datasets of Cora, Pubmed, Citeseer and Reddit.

Nair, Gopikrishnan R.↗