Search NASA⌕ Search

SEARCH · Search NASA

Results for “data clustering”

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 865 records · Page 48

An Offload NIC for NASA, NLR, and Grid Computing

This work addresses distributed data management and access dynamically configurable high-speed access to data distributed and shared over wide-area high-speed network environments. An offload engine NIC (network interface card) is proposed that scales at nX10-Gbps increments through 100-Gbps full duplex. The Globus de facto standard was used in projects requiring secure, robust, high-speed bulk data transport. Novel extension mechanisms were derived that will combine these technologies for use by GridFTP, bandwidth management resources, and host CPU (central processing unit) acceleration. The result will be wire-rate encrypted Globus grid data transactions through offload for splintering, encryption, and compression. As the need for greater network bandwidth increases, there is an inherent need for faster CPUs. The best way to accelerate CPUs is through a network acceleration engine. Grid computing data transfers for the Globus tool set did not have wire-rate encryption or compression. Existing technology cannot keep pace with the greater bandwidths of backplane and network connections. Present offload engines with ports to Ethernet are 32 to 40 Gbps f-d at best. The best of ultra-high-speed offload engines use expensive ASICs (application specific integrated circuits) or NPUs (network processing units). The present state of the art also includes bonding and the use of multiple NICs that are also in the planning stages for future portability to ASICs and software to accommodate data rates at 100 Gbps. The remaining industry solutions are for carrier-grade equipment manufacturers, with costly line cards having multiples of 10-Gbps ports, or 100-Gbps ports such as CFP modules that interface to costly ASICs and related circuitry. All of the existing solutions vary in configuration based on requirements of the host, motherboard, or carriergrade equipment. The purpose of the innovation is to eliminate data bottlenecks within cluster, grid, and cloud computing systems, and to add several more capabilities while reducing space consumption and cost. Provisions were designed for interoperability with systems used in the NASA HEC (High-End Computing) program. The new acceleration engine consists of state-ofthe- art FPGA (field-programmable gate array) core IP, C, and Verilog code; novel communication protocol; and extensions to the Globus structure. The engine provides the functions of network acceleration, encryption, compression, packet-ordering, and security added to Globus grid or for cloud data transfer. This system is scalable in nX10-Gbps increments through 100-Gbps f-d. It can be interfaced to industry-standard system-side or network-side devices or core IP in increments of 10 GigE, scaling to provide IEEE 40/100 GigE compliance.

Awrach, James↗

Multiscale studies of the three-dimensional dayside X-line

We review recent experience from the Cluster, Double Star, and THEMIS missions for lessons that apply to the upcoming Magnetospheric Multiscale Mission (MMS) being developed for launch in 2014. On global scales, simulation and statistical studies lead to mean configurations of dayside reconnection, implying specific relative alignments of the inflow magnetic fields and X-line, with implications for MMS operations designed to maximize the number of close encounters with the diffusion region. At intermediate MHD-to-ion scales, reconstruction of features created by one or two X-lines have developed to the point where data from a cluster of spacecraft can determine their temporal trends and the approximate three-dimensional X-line structure. Recent petascale particle-in-cell (PIC) simulations of reconnection encompass three spatial dimensions with excellent resolution, and make striking predictions of electron scale physics that creates complex interacting flux ropes under component reconnection. High time resolution measurements from MMS will determine the detailed electron scale kinetics embedded within the global and MHD–ion scale contexts. These developments will lead to the refinement of our three-dimensional multiscale picture of reconnection, yielding improved understanding of the global, MHD, and local physics controlling the onset or quenching, variability, and mean rate of reconnection. This in turn will enable improved predictability of the structural features created by transient reconnection, and their space weather consequences.

79 ASTRONOMY AND ASTROPHYSICS↗

Multiscale Studies of the Three-Dimensional Dayside X-Line

We review recent experience from the Cluster, Double Star, and THEMIS missions for lessons that apply to the upcoming Magnetospheric Multiscale Mission (MMS) being developed for launch in 2014. On global scales, simulation and statistical studies lead to mean congurations of dayside reconnection, implying specific relative alignments of the inflow magnetic fields and X-line, with implications for MMS operations designed to maximize the number of close encounters with the diffusion region. At intermediate MHD-to-ion scales, reconstruction of features created by one or two X-lines have developed to the point where data from a cluster of spacecraft can determine their temporal trends and the approximate three-dimensional X-line structure. Recent petascale particle-in-cell (PIC) simulations of reconnection encompass three spatial dimensions with excellent resolution, and make striking predictions of electron scale physics that creates complex interacting flux ropes under component reconnection. High time resolution measurements from MMS will determine the detailed electron scale kinetics embedded within the global and MHDion scale contexts. These developments will lead to the refinement of our three-dimensional multiscale picture of reconnection, yielding improved understanding of the global, MHD, and local physics controlling the onset or quenching, variability, and mean rate of reconnection. This in turn will enable improved predictability of the structural features created by transient reconnection, and their space weather consequences.

Reconnection↗

A magnified view of star formation at z = 0.9 from two lensed galaxies

We present new narrowband Hα imaging from the Hubble Space Telescope of two z = 0.91 galaxies that have been lensed by the foreground galaxy cluster A2390. These data probe spatial scales as small as ∼0.3 kpc, providing a magnified look at the morphology of star formation at an epoch when the global star formation rate (SFR) was high. However, dust attenuates our spatially resolved SFR indicators, the Hα and rest-UV emission, and we lack a direct measurement of extinction. Other studies have found that ionized gas in galaxies tends to be roughly 50% more obscured than stars; however, given an unextincted measurement of the SFR we can quantify the relative stellar to nebular extinction and the extinction in Hα. We infer SFRs from Spitzer and Herschel mid- to far-infrared observations and compare these to integrated Hα and rest-UV SFRs; this yields stellar to nebular extinction ratios consistent with previous studies. We take advantage of high spatial resolution and contextualize these results in terms of the source-plane morphologies, comparing the distribution of Hα to that of the rest-frame UV and optical light. In one galaxy, we measure separate SFRs in visually distinct clumps, but can set only a lower limit on the extinction and thus the star formation. Consequently, the data are also consistent with there being an equal amount of extinction along the lines of sight to the ionized gas as to the stars. Future observations in the far-infrared could settle this by mapping out the dust directly.

79 ASTRONOMY AND ASTROPHYSICS↗

THE DEEP2 GALAXY REDSHIFT SURVEY: THE VORONOI-DELAUNAY METHOD CATALOG OF GALAXY GROUPS

We present a public catalog of galaxy groups constructed from the spectroscopic sample of galaxies in the fourth data release from the Deep Extragalactic Evolutionary Probe 2 (DEEP2) Galaxy Redshift Survey, including the Extended Groth Strip (EGS). The catalog contains 1165 groups with two or more members in the EGS over the redshift range 0 < z < 1.5 and 1295 groups at z > 0.6 in the rest of DEEP2. Twenty-five percent of EGS galaxies and fourteen percent of high-z DEEP2 galaxies are assigned to galaxy groups. The groups were detected using the Voronoi-Delaunay method (VDM) after it has been optimized on mock DEEP2 catalogs following similar methods to those employed in Gerke et al. In the optimization effort, we have taken particular care to ensure that the mock catalogs resemble the data as closely as possible, and we have fine-tuned our methods separately on mocks constructed for the EGS and the rest of DEEP2. We have also probed the effect of the assumed cosmology on our inferred group-finding efficiency by performing our optimization on three different mock catalogs with different background cosmologies, finding large differences in the group-finding success we can achieve for these different mocks. Using the mock catalog whose background cosmology is most consistent with current data, we estimate that the DEEP2 group catalog is 72% complete and 61% pure (74% and 67% for the EGS) and that the group finder correctly classifies 70% of galaxies that truly belong to groups, with an additional 46% of interloper galaxies contaminating the catalog (66% and 43% for the EGS). We also confirm that the VDM catalog reconstructs the abundance of galaxy groups with velocity dispersions above {approx}300 km s{sup -1} to an accuracy better than the sample variance, and this successful reconstruction is not strongly dependent on cosmology. This makes the DEEP2 group catalog a promising probe of the growth of cosmic structure that can potentially be used for cosmological tests.

79 ASTRONOMY AND ASTROPHYSICS↗

Soar adaptive optics observations of the globular cluster NGC 6496

We present high-quality BVRI photometric data in the field of globular cluster NGC 6496 obtained with the SOAR Telescope Adaptive Module (SAM). Our observations were collected as part of the ongoing SAM commissioning. The distance modulus and cluster color excess as found from the red clump are (m – M) {sub V} = 15.71 ± 0.02 mag and E(V – I) = 0.28 ± 0.02 mag. An age of 10.5 ± 0.5 Gyr is determined from the difference in magnitude between the red clump and the subgiant branch. These parameters are in excellent agreement with the values derived from isochrone fitting. From the color-magnitude diagram we find a metallicity of [Fe/H] = –0.65 dex and hence support a disk classification for NGC 6496. The complete BVRI data set for NGC 6469 is made available in the electronic edition of the Journal.

79 ASTRONOMY AND ASTROPHYSICS↗

Automated analysis for detecting beams in laser wakefield simulations

Laser wakefield particle accelerators have shown the potential to generate electric fields thousands of times higher than those of conventional accelerators. The resulting extremely short particle acceleration distance could yield a potential new compact source of energetic electrons and radiation, with wide applications from medicine to physics. Physicists investigate laser-plasma internal dynamics by running particle-in-cell simulations; however, this generates a large dataset that requires time-consuming, manual inspection by experts in order to detect key features such as beam formation. This paper describes a framework to automate the data analysis and classification of simulation data. First, we propose a new method to identify locations with high density of particles in the space-time domain, based on maximum extremum point detection on the particle distribution. We analyze high density electron regions using a lifetime diagram by organizing and pruning the maximum extrema as nodes in a minimum spanning tree. Second, we partition the multivariate data using fuzzy clustering to detect time steps in a experiment that may contain a high quality electron beam. Finally, we combine results from fuzzy clustering and bunch lifetime analysis to estimate spatially confined beams. We demonstrate our algorithms successfully on four different simulation datasets.

43 PARTICLE ACCELERATORS↗

Driving photochemistry by clustering: The ICl-Xe case

We present slice imaging data demonstrating the influence of clustering on the photodissociation dynamics of a diatomic molecule: iodine monochloride (ICl) was dissociated at 235 nm in He and Xe seed gasses, probing both Cl and I photofragment energy and angular distributions. We observe that the kinetic energy releases of both Cl and I fragments change from He to Xe seeding. For Cl fragments, the seeding in Xe increases the kinetic energy release of some Cl fragments with a narrow kinetic energy distribution, and leads to some fragments with rather broad statistical distribution falling off exponentially from near-zero energies up to about 2.5 eV. Iodine fragment distribution changes even more dramatically from He to Xe seeding: sharp features essentially disappear and a broad distribution arises reaching to about 2.5 eV. Both these observations are rationalized by a simple qualitative cluster model assuming ICl dissociation inside larger xenon clusters and 'on surface' of smaller Xe species.

74 ATOMIC AND MOLECULAR PHYSICS↗

Hybrid deep learning architecture for general disruption prediction across tokamaks

In this paper, we present a new deep learning disruption prediction algorithm based on important findings from explorative data analysis which effectively allows knowledge transfer from existing devices to new ones, thereby predicting disruptions using very limited disruptive data from the new devices. Here, the explorative data analysis conducted via unsupervised clustering techniques confirms that time-sequence data are much better separators of disruptive and non-disruptive behavior than the instantaneous plasma state data with further advantageous implications for a sequence-based predictor. Based on such important findings, we have designed a new algorithm for multi-machine disruption prediction that achieves high predictive accuracy on the C-Mod (AUC=0.801), DIII-D (AUC=0.947) and EAST (AUC=0.973) tokamaks with limited hyperparameter tuning. Through numerical experiments, we show that boosted accuracy (AUC=0.959) is achieved on EAST predictions by including in the training only 20 disruptive discharges, thousands of non-disruptive discharges from EAST, and combining this with more than a thousand discharges from DIII-D and C-Mod. The improvement of predictive ability obtained by combining disruptive data from other devices is found to be true for all permutations of the three devices. Furthermore, by comparing the predictive performance of each individual numerical experiment, we find that non-disruptive data are machine-specific while disruptive data from multiple devices contain device-independent knowledge that can be used to inform predictions for disruptions occurring on a new device.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Characterizing Intracluster Light in the Hubble Frontier Fields

We investigate the intracluster light (ICL) in the six Hubble Frontier Field clusters at 0.3<z<0.6. We employ a new method, which is free from any functional form of the ICL profile, and exploit the unprecedented depth of this Hubble Space Telescope imaging to map the ICL’s diffuse light out to clustrocentric radii R∼300 kpc (μ{sub ICL}∼27 mag arcsec{sup −2}). From these maps, we construct radial color and stellar mass profiles via SED fitting and find clear negative color gradients in all systems with increasing distance from the Brightest Cluster Galaxy (BCG). While this implies older/more metal-rich stellar components in the inner part of the ICL, we find that the ICL mostly consists of a ≲2 Gyr population, and plausibly originated with logM{sub ∗}/M{sub ⊙}≲10 cluster galaxies. Furthermore, we find that 10%–15% of the ICL’s mass at large radii (≳150 kpc) lies in a younger/bluer stellar population (∼1 Gyr), a phenomenon not seen in local samples. We attribute this light to the higher fraction of star-forming/(post-)starburst galaxies in clusters at z∼0.5. Ultimately, we find the ICL’s total mass to be logM{sub ∗}{sup ICL}/M{sub ⊙}∼11–12, constituting 5%–20% of the clusters’ total stellar mass, or about half of the value at z∼0. The above implies distinct formation histories for the ICL and BCGs/other massive cluster galaxies; i.e., the ICL at this epoch is still being constructed rapidly (∼40 M{sub ⊙} yr{sup −1}), while the BCGs have mostly completed their evolution. To be consistent with the ICL measurements of local massive clusters, such as Virgo, our data suggest mass acquisition mainly from quiescent cluster galaxies is the principal source of ICL material in the subsequent ∼5 Gyr of cosmic time.

79 ASTRONOMY AND ASTROPHYSICS↗

Hybrid deep learning architecture for general disruption prediction across tokamaks

In this paper, we present a new deep learning disruption prediction algorithm based on important findings from explorative data analysis which effectively allows knowledge transfer from existing devices to new ones, thereby predicting disruptions using very limited disruptive data from the new devices. The explorative data analysis conducted via unsupervised clustering techniques confirms that time-sequence data are much better separators of disruptive and non-disruptive behavior than the instantaneous plasma state data with further advantageous implications for a sequence-based predictor. Based on such important findings, we have designed a new algorithm for multi-machine disruption prediction that achieves high predictive accuracy on the C-Mod (AUC=0.801), DIII-D (AUC=0.947) and EAST (AUC=0.973). tokamaks with limited hyperparameter tuning. Through numerical experiments, we show that boosted accuracy (AUC=0.959) is achieved on EAST predictions by including in the training only 20 disruptive discharges, thousands of non-disruptive discharges from EAST, and combining this with more than a thousand discharges from DIII-D and C-Mod. The improvement of predictive ability obtained by combining disruptive data from other devices is found to be true for all permutations of the three devices. Furthermore, by comparing the predictive performance of each individual numerical experiment, we find that non-disruptive data are machine-specific while disruptive data from multiple devices contain device-independent knowledge that can be used to inform predictions for disruptions occurring on a new device.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

A Control Chart Approach for Representing and Mining Data Streams with Shape Based Similarity

The mining of data streams for online condition monitoring is a challenging task in several domains including (electric) power grid system, intelligent manufacturing, and consumer science. Considering a power grid application in which thousands of sensors, called the phasor measurement units, are deployed on the power grid network to continuously collect streams of digital data for real-time situational awareness and system management. Depending on design, each sensor could stream between ten and sixty data samples per second. The myriad of sensory data captured could convey deeper insights about sequence of events in real-time and before major damages are done. However, the timely processing and analysis of these high-velocity and high-volume data streams is a challenge. Hence, a new data processing and transformation approach, based on the concept of control charts, for representing sequence of data streams from sensors is proposed. In addition, an application of the proposed approach for enhancing data mining tasks such as clustering using real-world power grid data streams is presented. The results indicate that the proposed approach is very efficient for data streams storage and manipulation.

big data↗

DRY MERGER RATE AND POST-MERGER FRACTION IN THE COMA CLUSTER CORE

We evaluate the dry merger activity in the Coma cluster, using a spectroscopically complete sample of 70 red-sequence (RS) galaxies, most of which (∼75%) are located within 0.2R{sub 200} (∼0.5 Mpc) from the cluster center, with data from the Coma Treasury Survey obtained with the Hubble Space Telescope. The fraction of close galaxy pairs in the sample is the proxy employed for the estimation of the merger activity. We identify 5 pairs and 1 triplet, enclosing a total of 13 galaxies, based on limits on projected separation and line-of-sight velocity difference. Of these systems, none show signs of ongoing interaction, and therefore we do not find any true mergers in our sample. This negative result sets a 1σ upper limit of 1.5% per Gyr for the major dry merger rate, consistent with the low rates expected in present-day clusters. Detailed examination of the images of all the RS galaxies in the sample reveals only one with low surface brightness features identifiable as the remnant of a past merger or interaction, implying a post-merger fraction below 2%.

79 ASTRONOMY AND ASTROPHYSICS↗

Semi-Supervised Domain Adaptation for Cross-Survey Galaxy Morphology Classification and Anomaly Detection

In the era of big astronomical surveys, our ability to leverage artificial intelligence algorithms simultaneously for multiple datasets will open new avenues for scientific discovery. Unfortunately, simply training a deep neural network on images from one data domain often leads to very poor performance on any other dataset. Here we develop a Universal Domain Adaptation method DeepAstroUDA, capable of performing semi-supervised domain alignment that can be applied to datasets with different types of class overlap. Extra classes can be present in any of the two datasets, and the method can even be used in the presence of unknown classes. For the first time, we demonstrate the successful use of domain adaptation on two very different observational datasets (from SDSS and DECaLS). We show that our method is capable of bridging the gap between two astronomical surveys, and also performs well for anomaly detection and clustering of unknown data in the unlabeled dataset. We apply our model to two examples of galaxy morphology classification tasks with anomaly detection: 1) classifying spiral and elliptical galaxies with detection of merging galaxies (three classes including one unknown anomaly class); 2) a more granular problem where the classes describe more detailed morphological properties of galaxies, with the detection of gravitational lenses (ten classes including one unknown anomaly class).

79 ASTRONOMY AND ASTROPHYSICS↗

SUBARU WEAK-LENSING STUDY OF A2163: BIMODAL MASS STRUCTURE

We present a weak-lensing analysis of the merging cluster A2163 using Subaru/Suprime-Cam and CFHT/Mega-Cam data and discuss the dynamics of this cluster merger, based on complementary weak-lensing, X-ray, and optical spectroscopic data sets. From two-dimensional multi-component weak-lensing analysis, we reveal that the cluster mass distribution is well described by three main components including the two-component main cluster A2163-A with mass ratio 1:8, and its cluster satellite A2163-B. The bimodal mass distribution in A2163-A is similar to the galaxy density distribution, but appears as spatially segregated from the brightest X-ray emitting gas region. We discuss the possible origins of this gas-dark-matter offset and suggest the gas core of the A2163-A subcluster has been stripped away by ram pressure from its dark matter component. The survival of this gas core from the tidal forces exerted by the main cluster lets us infer a subcluster accretion with a non-zero impact parameter. Dominated by the most massive component of A2163-A, the mass distribution of A2163 is well described by a universal Navarro-Frenk-White profile as shown by a one-dimensional tangential shear analysis, while the singular-isothermal sphere profile is strongly ruled out. Comparing this cluster mass profile with profiles derived assuming intracluster medium hydrostatic equilibrium (H.E.) in two opposite regions of the cluster atmosphere has allowed us to confirm the prediction of a departure from H.E. in the eastern cluster side, presumably due to shock heating. Yielding a cluster mass estimate of M{sub 500} = 11.18{sup +1.64}{sub -1.46} Multiplication-Sign 10{sup 14} h {sup -1} M{sub Sun }, our mass profile confirms the exceptionally high mass of A2163, consistent with previous analyses relying on the cluster dynamical analysis and Y{sub X} mass proxy.

79 ASTRONOMY AND ASTROPHYSICS↗

GLOBULAR CLUSTER ABUNDANCES FROM HIGH-RESOLUTION, INTEGRATED-LIGHT SPECTROSCOPY. II. EXPANDING THE METALLICITY RANGE FOR OLD CLUSTERS AND UPDATED ANALYSIS TECHNIQUES

We present abundances of globular clusters (GCs) in the Milky Way and Fornax from integrated-light (IL) spectra. Our goal is to evaluate the consistency of the IL analysis relative to standard abundance analysis for individual stars in those same clusters. This sample includes an updated analysis of seven clusters from our previous publications and results for five new clusters that expand the metallicity range over which our technique has been tested. We find that the [Fe/H] measured from IL spectra agrees to ∼0.1 dex for GCs with metallicities as high as [Fe/H] = −0.3, but the abundances measured for more metal-rich clusters may be underestimated. In addition we systematically evaluate the accuracy of abundance ratios, [X/Fe], for Na i, Mg i, Al i, Si i, Ca i, Ti i, Ti ii, Sc ii, V i, Cr i, Mn i, Co i, Ni i, Cu i, Y ii, Zr i, Ba ii, La ii, Nd ii, and Eu ii. The elements for which the IL analysis gives results that are most similar to analysis of individual stellar spectra are Fe i, Ca i, Si i, Ni i, and Ba ii. The elements that show the greatest differences include Mg i and Zr i. Some elements show good agreement only over a limited range in metallicity. More stellar abundance data in these clusters would enable more complete evaluation of the IL results for other important elements.

79 ASTRONOMY AND ASTROPHYSICS↗

( Suzaku Suzaku) X-ray observations of the nearest non-cool core cluster, Antlia: dynamically young but with remarkably relaxed outskirts

We present the results of seven Suzaku mosaic observations (>200 ks in total) of the nearest non-cool core cluster, the Antlia Cluster (or Group), beyond its degree-scale virial radius in its eastern direction. The temperature is consistent with the scaled profiles of many other clusters. Its pressure follows the universal profile. The density slope in its outskirts is significantly steeper than that of the nearest cool core cluster (Virgo) with a similar temperature as Antlia, but shallower than those of the massive clusters. The entropy increases all the way out to R {sub 200}, which is consistent in value with the baseline model predicted by a gravity heating-only mechanism in the outskirts. Antlia is quite relaxed in this direction. However, the entropy inside ∼R {sub 500} is significantly higher than the baseline model, which is similar to many other nearby low mass clusters or groups. The enclosed gas-mass fraction does not exceed the cosmic value out to 1.3R{sub 200}. Thus, there is no evidence of significant gas clumping, electron-ion non-equipartition, or departure from the hydrostatic equilibrium approximation that are suggested to explain the entropy and gas fraction anomalies found in the outskirts of some massive clusters. We also present scaling relations for the gas fraction (f{sub gas,200}), entropy (K {sub 200}), and temperature (T {sub 500}) using 22 groups and clusters with published data in the literature. The enclosed baryon fraction at R {sub 200} is broadly consistent with the cosmic value. The power law slope of the K {sub 200}–T {sub 500} relation is 0.638 ± 0.205. The entropy deficit at R {sub 200} cannot be fully accounted for by the bias or deviation in the gas fraction.

79 ASTRONOMY AND ASTROPHYSICS↗

Chemical Profiling of the Plant Cell Wall through Raman Microspectroscopy

This paper presents a computational framework for chemical pro.ling of the plant cell wall through the Raman spectroscopy. The system enables query of known spectral signatures and clustering of spectral data based on intrinsic properties. As a result, presence and relative concentration of speci.c chemical bonds can be quanti.ed. The primary contribution of this paper is in representation of raman pro.le in terms of .uorescence background and multiscale peak detection at each grid point (voxel). Such a representation allows ef.cient spatial segmentation based on the coupling between high-level salient properties and low-level symbolic representation at each voxel. The high-level salient properties refer to preferred peaks and their attributes for the entire image. The low-level symbolic representations are based on .uorescence background, spectral peak locations, and their attributes. We present results on a corn stover tissue section that is imaged through Raman microscopy, and the results are consistent with the literature. In addition, automatic clustering indicates several distinct layers of the cell walls with different spectral signatures.

09 BIOMASS FUELS↗