Search NASA⌕ Search

SEARCH · Search NASA

Results for “sparse factorization”

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 73 records · Page 4
All sourcesNASA NTRSDOE OSTI
NASA NTRS1994

Asymmetry parameters of the phase function for densely packed scattering grains

Spatial correlation among densely packed particles can substantially change their single-scattering properties, thus making questionable the applicability of the independent scattering approximation in calculations of light scattering by planetary regoliths. The same problem arises in geophysics in light scattering computations for snow, frosts, and bare soil. In this paper, we use a dense-medium light-scattering theory based on the introduction of the static structure factor to calculate asymmetry parameters of the phase function for densely packed particles with real refractive indices 1.31 and 1.66, approximating water ice and soil particles, respectively, and imaginary refractive indices 0, 0.01, and 0.3. For sparsely distributed, independently scattering grains, the calculated asymmetry parameters are always positive and always larger than those for densely packed particles. For densely packed grains, the asymmetry parameters may be negative but only for radius-to-wavelength ratios from about 0.1 to about 0.4. With decreasing particle size, the calculated asymmetry parameters tend to zero independently of the compaction state. In the geometrical optics regime, the asymmetry parameters for densely packed scatterers are positive and very close to those for independently scattering grains. These results may have important implications for remote sensing of the Earth and solid planetary surfaces. In particular, it is demonstrated that negative asymmetry parameters derived with some approximate multiple-scattering theories may be physically irrelevant and can be the result of using an inaccurate bidirectional reflection function combined with the ill-conditionally of the inverse scattering problem.

Mishchenko, Michael I.↗
NASA NTRS1993

A spectral algorithm for envelope reduction of sparse matrices

A new algorithm for reducing the envelope of a sparse matrix is presented. This algorithm is based on the computation of eigenvectors of the Laplacian matrix associated with the graph of the sparse matrix. A reordering of the sparse matrix is determined based on the numerical values of the entries of an eigenvector of the Laplacian matrix. Numerical results show that the new reordering algorithm can in some cases reduce the envelope by more than a factor of two over the current standard algorithms such as Gibbs-Poole-Stockmeyer (GPS) or SPARSPAK's reverse Cuthill-McKee (RCM).

Barnard, Stephen T.↗
NASA NTRS2019

Predictive Modeling for Differential Diagnosis and Mortality Risk Assessment

The prevalence of electronic health record (EHR) systems has brought prodigious biomedical informatics opportunity. Automated machine learning methods can effectively utilize such data and have become common tools for healthcare predictive modeling. Researches in medical informatics have explored the potential of deep learning and classical models in emergent care scenarios. In particular, predicting differential diagnoses for admissions have proven useful in decreasing unnecessary lab tests and improving inpatient triage decision-making. Moreover, identification of high-risk patients for in-hospital mortality is vitally important to maximize allocation of medical resources.The Medical Information Mart for Intensive Care (MIMIC-III) database, containing de-identified critical care inpatient was used in our study. This data set captures hospital patient laboratory measurements, pharmacologic prescriptions, diagnostic data and procedure event recordings. When considering adult patients and discounting admissions with ICU length of stay less than 24 hours, there were 37,787 unique admissions and 30,414 total patients. We examined the top 25 most prevalent ICD-9 group-level disease specificities in MIMIC-III using a multi-label classification model. In-hospital mortality was modeled as binary classification with 4,155 (13%) adult patients that expired, of which 3,138 (75.5%) were in the ICU setting. The metrics AUC, F1 score, sensitivity and specificity values calculated for each disease label measured prediction performance.The usage of ICD-9 group codes reduced feature dimension from 14,567 to 942 and greatly improved distribution of patient diagnostic categories. Disease temporal patterns were captured by considering the most frequently sampled 6 vital signs and 13 laboratory values. Missing data were imputed at each time-stamp. Time-series raw hourly average values were converted into 5 summary features (mean, standard deviation, number of observations, min & max values). Patient demographic variables such as age, gender, marital status and ethnicity were also factored into the modeling. Choi et al showed that contextual embedding of medical data, diagnostic and procedural codes alone can predict future diagnoses with sensitivity as high as 0.79. We utilized an embedding technique called word2vec which allowed sparse representations of medical history to be transformed into dense word vectors. The mappings captured contextual information by treating each admission as a sentence and learning the most likely neighboring words in a sliding window fashion. Binary and multi-label classification was achieved via collapse models, which do not consider temporal information, as well as recurrent neural networks with regularization, Softmax output layer activation together with categorical cross-entropy as the loss function.

US Army collaboration↗
NASA NTRS2021

Discrete anisotropic radiative transfer modelling of solar-induced chlorophyll fluorescence: Structural impacts in geometrically explicit vegetation canopies

Solar-induced fluorescence (SIF) is a subtle but informative optical signal of vegetation photosynthesis. Remotely sensed SIF integrates environmental, physiological and structural changes that alter photosynthesis at leaf, plant and canopy scales. Radiative transfer models are ideally suited to investigate the complex sources of variability in the SIF signal to guide the interpretation of SIF retrievals from airborne and space-borne platforms. Here, we coupled the Fluspect-Cx model of leaf optical properties and chlorophyll-a fluorescence with the Discrete Anisotropic Radiative Transfer (DART) model to upscale SIF from individual leaves to three-dimensional (3D) structurally explicit canopies. For one-dimensional homogeneous (turbid-like) canopies, DART-SIF was nearly identical to SIF simulated in two existing models, SCOPE and mSCOPE (RMSE <0.221 W./sq.m.μm.sr). DART simulations in geometrically explicit 3D canopies offered four important insights regarding the influence of vegetation structure on the multi-angular top-of-canopy SIF signal. First, changes in the 3D canopy architecture of maize crops, represented by leaf density (leaf area index), and plant clumping (canopy closure) had a larger impact on SIF than the modelled photosynthetic efficiency distinction between sun-adapted and shade-adapted foliage. Second, clumping of leaves at the crop and stand levels was identified as one of the key driving factors of multi-angular anisotropy of red and far-red SIF (686 and 740 nm) for both maize and eucalyptus canopies. Third, non-photosynthetic woody material had a significant impact on top-of-canopy SIF in modelled 3D forest stands. Wood shadowing decreased the photosynthetically active radiation absorbed by green leaves, and consequently the SIF emissions, by 10% in sparse and 17% in dense eucalyptus stands. The wood obstruction (blocking) effect, quantified as a relative difference of SIF escape probabilities from canopies with and without wood in the nadir viewing direction, decreased far-red SIF by 4–6% but it had a smaller and sometimes positive influence (by less than 2%) on red SIF. Fourth, DART 3D radiative budget profiles revealed that the majority of the SIF signal from a dense eucalyptus stand originated from the top 25% of the simulated canopy. Interestingly, the introduction of bark-covered woody elements did not alter the simulated balance and omnidirectional escape factor of red SIF in this upper canopy part but did raise significantly both of them in case of far-red SIF. These results demonstrate the importance of 3D radiative transfer and radiative budget simulations for investigating SIF interactions in structurally complex plant canopies and for a better understanding of spatiotemporal and multi-angular remote sensing SIF observations.

Zbyněk Malenovský↗
NASA NTRS1991

Lanczos eigensolution method for high-performance computers

The theory, computational analysis, and applications are presented of a Lanczos algorithm on high performance computers. The computationally intensive steps of the algorithm are identified as: the matrix factorization, the forward/backward equation solution, and the matrix vector multiples. These computational steps are optimized to exploit the vector and parallel capabilities of high performance computers. The savings in computational time from applying optimization techniques such as: variable band and sparse data storage and access, loop unrolling, use of local memory, and compiler directives are presented. Two large scale structural analysis applications are described: the buckling of a composite blade stiffened panel with a cutout, and the vibration analysis of a high speed civil transport. The sequential computational time for the panel problem executed on a CONVEX computer of 181.6 seconds was decreased to 14.1 seconds with the optimized vector algorithm. The best computational time of 23 seconds for the transport problem with 17,000 degs of freedom was on the the Cray-YMP using an average of 3.63 processors.

Bostic, Susan W.↗
NASA NTRS1991

Implicit solvers for unstructured meshes

Implicit methods for unstructured mesh computations are developed and tested. The approximate system which arises from the Newton-linearization of the nonlinear evolution operator is solved by using the preconditioned generalized minimum residual technique. These different preconditioners are investigated: the incomplete LU factorization (ILU), block diagonal factorization, and the symmetric successive over-relaxation (SSOR). The preconditioners have been optimized to have good vectorization properties. The various methods are compared over a wide range of problems. Ordering of the unknowns, which affects the convergence of these sparse matrix iterative methods, is also investigated. Results are presented for inviscid and turbulent viscous calculations on single and multielement airfoil configurations using globally and adaptively generated meshes.

Venkatakrishnan, V.↗
NASA NTRS1991

Comparing direct and iterative equation solvers in a large structural analysis software system

Two direct Choleski equation solvers and two iterative preconditioned conjugate gradient (PCG) equation solvers used in a large structural analysis software system are described. The two direct solvers are implementations of the Choleski method for variable-band matrix storage and sparse matrix storage. The two iterative PCG solvers include the Jacobi conjugate gradient method and an incomplete Choleski conjugate gradient method. The performance of the direct and iterative solvers is compared by solving several representative structural analysis problems. Some key factors affecting the performance of the iterative solvers relative to the direct solvers are identified.

Poole, E. L.↗
NASA NTRS2021

Improved Rainfall Data in the Philippines through Concurrent Use of GPM IMERG and Ground-Based Measurements

The availability of accurate and reliable rainfall data that are applicable to various phenomenological, climatological, and modeling studies is important, especially in the Philippines, which is considered to be highly vulnerable to natural hazards and a changing climate. The presented strategy involved constructing a dataset consisting of synoptic data, automatic rain gauge (ARG) measurements, and satellite data that are co-registered, consistent, and formatted in the same manner. Although sparse in number, the synoptic stations provide the most accurate rainfall information and were used as the baseline for creating the dataset. The ARGs that are within a distance of 1 km to the synoptic stations were used to determine the correction factors needed to make the synoptic and ARG data consistent. Subsequently, the corrected ARGs were used to make the satellite IMERG data consistent with both ARG and synoptic data. In case of the latter, only IMERG pixels with at least 10 ARGs within the relatively large footprint of the satellite sensor were used in estimating the required correction parameters derived from a combination of a power transform and linear regression correction techniques. The final results show good agreement of synoptic and corrected ARG data with correlation coefficients of 0.94 and 0.97 for the 10 day and monthly data, respectively, and improvement in the linear regression slope from 0.67 to 0.90 for 10 day data, and 0.70 to 0.94 for monthly data. In addition, the corrected ARG data agree well with the corrected IMERG data, with correlation coefficients of 0.88 and 0.93 for the 10 day and monthly data, respectively, and an improvement in slope from 0.66 to 0.87 for 10 day data, and 0.74 to 0.99 for monthly data. The merit of using a combined dataset is illustrated through comparative analyses of the IMERG data and spatially interpolated synoptic and ARG data. The results show general agreements in spatial patterns of rainfall across the datasets, especially in areas where in situ measurements are recorded. The observed discrepancy when ground data is limited emphasizes the need for satellite IMERG data to obtain the true spatial patterns of rainfall distribution.

rainfall↗
NASA NTRS1987

Solving very large, sparse linear systems on mesh-connected parallel computers

The implementation of Pan and Reif's Parallel Nested Dissection (PND) algorithm on mesh connected parallel computers is described. This is the first known algorithm that allows very large, sparse linear systems of equations to be solved efficiently in polylog time using a small number of processors. How the processor bound of PND can be matched to the number of processors available on a given parallel computer by slowing down the algorithm by constant factors is described. Also, for the important class of problems where G(A) is a grid graph, a unique memory mapping that reduces the inter-processor communication requirements of PND to those that can be executed on mesh connected parallel machines is detailed. A description of an implementation on the Goodyear Massively Parallel Processor (MPP), located at Goddard is given. Also, a detailed discussion of data mappings and performance issues is given.

Opsahl, Torstein↗
NASA NTRS1991

Implicit solvers for unstructured meshes

Implicit methods were developed and tested for unstructured mesh computations. The approximate system which arises from the Newton linearization of the nonlinear evolution operator is solved by using the preconditioned GMRES (Generalized Minimum Residual) technique. Three different preconditioners were studied, namely, the incomplete LU factorization (ILU), block diagonal factorization, and the symmetric successive over relaxation (SSOR). The preconditioners were optimized to have good vectorization properties. SSOR and ILU were also studied as iterative schemes. The various methods are compared over a wide range of problems. Ordering of the unknowns, which affects the convergence of these sparse matrix iterative methods, is also studied. Results are presented for inviscid and turbulent viscous calculations on single and multielement airfoil configurations using globally and adaptively generated meshes.

Venkatakrishnan, V.↗
NASA NTRS2017

Determining Cloud Thermodynamic Phase from Micropulse Lidar Network Data

Determining cloud thermodynamic phase is a critical factor in studies of Earth's radiation budget. Here we use observations from the NASA Micro Pulse Lidar Network (MPLNET) and thermodynamic profiles from the Goddard Earth Observing System, version 5 (GEOS-5) to distinguish liquid water, mixed-phase, and ice water clouds. The MPLNET provides sparse global, autonomous, and continuous measurements of clouds and aerosols which have been used in a number of scientific investigations to date. The use of a standardized instrument and a common suite of data processing algorithms with thorough uncertainty characterization allows for straightforward comparisons between sites. Lidars with polarization capabilities have recently been incorporated into the MPLNET project which allows, for the first time, the ability to infer a cloud thermodynamic phase. This presentation will look specifically at the occurrence of ice and mixed phase clouds in the temperature region of -10 C to -40 C for different climatological regions and seasons. We compare MPLNET occurrences of mixed-phase clouds to an historical climatology based on observations from the Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) instrument aboard the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) spacecraft.

mixing layer height↗
NASA NTRSDate not supplied

Characterization of OCO-2 and ACOS-GOSAT biases and errors for flux estimates

We characterize the magnitude of seasonally and spatially varying biases in the National Aeronautics and Space Administration (NASA) Orbiting Carbon Observatory-2 (OCO-2) Version 8 (v8) and the Atmospheric CO2 Observations from Space (ACOS) Greenhouse Gas Observing SATellite (GOSAT) version 7.3 (v7.3) satellite CO2 retrievals by comparisons to measurements collected by the Total Carbon Column Observing Network (TCCON), Atmospheric Tomography (ATom) experiment, and National Oceanic and Atmospheric Administration (NOAA) Earth System Research Laboratory (ESRL) and U. S. Department of Energy (DOE) aircraft, and surface stations. Although the ACOS-GOSAT estimates of the column averaged carbon dioxide (CO2) dry air mole fraction (XCO2) have larger random errors than the OCO-2 XCO2 estimates, and the space-based estimates over land have larger random errors than those over ocean, the systematic errors are similar across both satellites and surface types, 0.6 ± 0.1 ppm. We find similar estimates of systematic error whether dynamic versus geometric coincidences or ESRL/DOE aircraft versus TCCON are used for validation (over land), once validation and co-location errors are accounted for. We also find that areas with sparse throughput of good quality data (due to quality flags and preprocessor selection) over land have ~double the error of regions of high-throughput of good quality data. We characterize both raw and bias-corrected results, finding that bias correction improves systematic errors by a factor of 2 for land observations and improves errors by ~0.2 ppm for ocean. We validate the lowermost tropospheric (LMT) product for OCO-2 and ACOS-GOSAT by comparison to aircraft and surface sites, finding systematic errors of ~1.1 ppm, while having 2-3 times the variability of XCO2. We characterize the time and distance scales of correlations for OCO-2 XCO2 errors, and find error correlations on scales of 0.3 degrees, 5-10 degrees, and 60 days. We find comparable scale lengths for the bias correction term. Assimilation of the OCO-2 bias correction term is used to estimate flux errors resulting from OCO-2 seasonal biases, finding annual flux errors on the order of 0.3 and 0.4 PgC/yr for Transcom-3 ocean and land regions, respectively.

Susan S Kulawik↗
NASA NTRS2016

The Global Drivers of Photosynthesis and Light Use Efficiency Seasonality: A Granger Frequency Causality Analysis

Photosynthesis and light use efficiency (LUE) are major factors in the evolution of the continental carbon cycle due to their contribution to gross primary production (GPP). However, while the drivers of photosynthesis and LUE on a plant or canopy scale can often be identified, significant uncertainties exist when modeling these on a global scale. This is due to sparse observations in regions such as the tropics and the lack of a direct global observation dataset. Although others have attempted to address this issue using correlations (Beer, 2010) or calculating GPP from vegetation indices (Running, 2004), in this study we take a new approach. We combine the statistical method of Granger frequency causality and partial Granger frequency causality with remote sensing data products (including sun-induced fluorescence used as a proxy for GPP) to determine the main environmental drivers of GPP across the globe.

photosynthesis↗
NASA NTRS2017

Machine Learning Approaches to Increasing Value of Spaceflight Omics Databases

The number of spaceflight bioscience mission opportunities is too small to allow all relevant biological and environmental parameters to be experimentally identified. Simulated spaceflight experiments in ground-based facilities (GBFs), such as clinostats, are each suitable only for particular investigations -- a rotating-wall vessel may be 'simulated microgravity' for cell differentiation (hours), but not DNA repair (seconds) -- and introduce confounding stimuli, such as motor vibration and fluid shear effects. This uncertainty over which biological mechanisms respond to a given form of simulated space radiation or gravity, as well as its side effects, limits our ability to baseline spaceflight data and validate mission science. Machine learning techniques autonomously identify relevant and interdependent factors in a data set given the set of desired metrics to be evaluated: to automatically identify related studies, compare data from related studies, or determine linkages between types of data in the same study. System-of-systems (SoS) machine learning models have the ability to deal with both sparse and heterogeneous data, such as that provided by the small and diverse number of space biosciences flight missions; however, they require appropriate user-defined metrics for any given data set. Although machine learning in bioinformatics is rapidly expanding, the need to combine spaceflight/GBF mission parameters with omics data is unique. This work characterizes the basic requirements for implementing the SoS approach through the System Map (SM) technique, a composite of a dynamic Bayesian network and Gaussian mixture model, in real-world repositories such as the GeneLab Data System and Life Sciences Data Archive. The three primary steps are metadata management for experimental description using open-source ontologies, defining similarity and consistency metrics, and generating testing and validation data sets. Such approaches to spaceflight and GBF omics data may soon enable unique insight into which measured phenomena correlate to biological mechanisms that are truly affected by spaceflight conditions; which are most likely to be confounded by other variables; and which are insufficiently characterized, significantly increasing existing and future science return from ISS and spaceflight missions.

Gentry, Diana↗
NASA NTRSDate not supplied

Compressive Sensing Based Data Acquisition Architecture for Transient Stellar Events in Crowded Star Fields

Compressive sensing is a mathematical technique for simultaneous data acquisition and compression. In this work, we show a CS based architecture for acquiring and reconstructing transient stellar events. This architecture reconstructs a differenced image itself, eliminating the need for any sparse domain transforms, otherwise required for traditional CS reconstruction. The resulting reconstructed differenced image is of critical importance as the information required for generating a time-series photometric light curve is obtained only from a differenced image. Hence, reconstructing a crowded star spatial image, followed by differencing is wasteful. This architecture eliminates the need to 1.) transform an image to a sparse domain, 2.) Reconstruct a dense field, and then apply differencing on the image to obtain the image of critical value. We study the case of microlensing to depict a star source experiencing magnification in time. Our results show that this architecture is able to reconstruct star source magnitudes with magnification factors greater than 1 for clean images with error less than 2% using only 10% of the Nyquist rate samples.

Asmita Korde-Patel↗
NASA NTRSDate not supplied

Compressive Sensing Based Data Acquisition Architecture for Transient Stellar Events in Crowded Star Fields

Compressive sensing is a mathematical technique for simultaneous data acquisition and compression. In this work, we show a CS based architecture for acquiring and reconstructing transient stellar events. This architecture recon-structs a differenced image itself, eliminating the need for any sparse domain transforms, otherwise required for traditional CS reconstruction. The resulting reconstructed differenced image is of critical importance as the information required for generating a time-series photometric light curve is obtained only from a differenced image. Hence, reconstructing a crowded star spatial image, followed by differencing is wasteful. This architecture eliminates the need to 1.) transform an image to a sparse domain, 2.) Reconstruct a dense field, and then apply differencing on the image to obtain the image of critical value. We study the case of microlensing to depict a star source experiencing magnification in time. Our results show that this architecture is able to reconstruct star source magnitudes with magnification factors greater than 1 for clean images with error less than 2% using only 10% of the Nyquist rate samples.

Compressive Sensing, Data acquisition, Image diffe↗
NASA NTRSDate not supplied

Compressive Sensing Based Data Acquisition Architecture for Transient Stellar Events in Crowded Star Fields

Compressive sensing is a mathematical technique for simultaneous data acquisition and compression. In this work, we show a CS based architecture for acquiring and reconstructing transient stellar events. This architecture recon-structs a differenced image itself, eliminating the need for any sparse domain transforms, otherwise required for traditional CS reconstruction. The resulting reconstructed differenced image is of critical importance as the information required for generating a time-series photometric light curve is obtained only from a differenced image. Hence, reconstructing a crowded star spatial image, followed by differencing is wasteful. This architecture eliminates the need to 1.) transform an image to a sparse domain, 2.) Reconstruct a dense field, and then apply differencing on the image to obtain the image of critical value. We study the case of microlensing to depict a star source experiencing magnification in time. Our results show that this architecture is able to reconstruct star source magnitudes with magnification factors greater than 1 for clean images with error less than 2% using only 10% of the Nyquist rate samples.

Asmita Korde-patel↗
NASA NTRS2007

Interferometry from Space

Space-based interferometric observatories will be challenging projects, equal at least to that of building the Great Observatories (the Hubble Space Telescope (HST), Spitzer Space Telescope (SST), Chandra X-ray Observatory, and the Gamma Ray Observatory), if not the Pyramids of Eygpt - but they represent the next logical step in examining our Universe at substantially higher angular resolution. Increasing our resolving power by factors of 100 or more (as is needed to make meaningful improvements in this observational arena) over existing facilities such as HST and SST requires mirror diameters (100's to 1000's of meters) much larger than can be supported by single or segmented mirrors - and thus the design and construction of sparse aperture, inteferometric arrays such as those described herein will be required. But just imagine the rewards of being able to see, for the first time, the surfaces of other stars, the location and type of extrasolar planets and even pictures of those same planets, the inner workings of Active Galactic Nuclei, the close-in details of supernovae explosions, black hole event horizons, and the infrared universe at the same resolution of the UV-optical Hubble Deep Fields. As a slight variation on the "Star Trek: Enterprise" theme song might say, it'll be a "long road, getting from here to there", but it will one well-worth taking.

Carpenter, Kenneth↗