Search NASA⌕ Search

SEARCH · Search NASA

Results for “Statistical error”

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 1,153 records · Page 64

The muon g-2 experiment at FERMILAB

The muon anomaly, a µ = $\frac{g-2}{2}$, is a low energy observable that can be both measured and computed with high precision. Therefore it provides an important test of the Standard Model (SM) and it is a sensitive probe for new physics. The a µ value has been measured to a precision of 0.54 ppm by the E821 experiment at the Brookhaven National Laboratory (BNL). This result shows a difference greater than 3 σ compared to the SM prediction. In an effort to clarify this discrepancy between experimental measurement and theoretical calculation, the Muon g-2(E989) experiment at Fermilab aims to reduce the experimental error on a µ by a factor of four. E989 collected a dataset with the same statistical power of the BNL experiment during the Run 1 data taking (2018). The analysis of data is ongoing and the first result should become available in early 2020. In this paper, I will discuss the experimental setup and report on the status of the Run1 analysis.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Statistical Approaches to Aerosol Dynamics for Climate Simulation

In this work, we introduce two general non-parametric regression analysis methods for errors-in-variable (EIV) models: the compound regression, and the constrained regression. It is shown that these approaches are equivalent to each other and, to the general parametric structural modeling approach. The advantages of these methods lie in their intuitive geometric representations, their distribution free nature, and their ability to offer a practical solution when the ratio of the error variances is unknown. Each includes the classic non-parametric regression methods of ordinary least squares, geometric mean regression, and orthogonal regression as special cases. Both methods can be readily generalized to multiple linear regression with two or more random regressors.

Compound regression↗

Computationally efficient zero-noise extrapolation for quantum-gate-error mitigation

Zero noise extrapolation (ZNE) is a widely used technique for gate error mitigation on near term quantum computers because it can be implemented in software and does not require knowledge of the quantum computer noise parameters. Traditional ZNE requires a significant resource overhead in terms of quantum operations. A recent proposal using a targeted (or random) instead of fixed identity insertion method (riim versus fiim) requires significantly fewer quantum gates for the same formal precision. We start by showing that riim can allow for ZNE to be deployed on deeper circuits than fiim but requires many more measurements to maintain the same statistical uncertainty. We develop two extensions to fiim and riim. The List Identity Insertion Method (liim) allows to mitigate the error from certain cnot gates, typically those with the largest error. Set Identity Insertion Method (siim) naturally interpolates between the measurement-efficient fiim and the gate-efficient riim allowing to trade off fewer cnot gates for more measurements. Finally, we investigate a way to boost the number of measurements, namely to run ZNE in parallel, utilizing as many quantum devices as are available. We explore the performance of riim in a parallel setting where there is a non-trivial spread in noise across sets of qubits within or across quantum computers.

97 MATHEMATICS AND COMPUTING↗

The velocity distribution function of galaxy clusters as a cosmological probe

We present a new approach for quantifying the abundance of galaxy clusters and constraining cosmological parameters using dynamical measurements. In the standard method, galaxy line-of-sight velocities, v, or velocity dispersions are used to infer cluster masses, M, to quantify the halo mass function (HMF), dn(M)/dlog(M), which is strongly affected by mass measurement errors. In our new method, the probability distributions of velocities for each cluster in the sample are summed to create a new statistic called the velocity distribution function (VDF), dn(v)/dv. The VDF can be measured more directly and precisely than the HMF and can be robustly predicted with cosmological simulations that capture the dynamics of subhalos or galaxies. We apply these two methods to realistic (ideal) mock cluster catalogs with (without) interlopers and forecast the bias and constraints on the matter density parameter Ω{sub m} and the amplitude of matter fluctuations σ{sub 8} in flat ΛCDM cosmologies. For an example observation of 200 massive clusters, the VDF with (without) interloping galaxies constrains the parameter combination σ{sub 8} Ω{sub m}{sup 0.29(0.29)}=0.589±0.014 (0.584±0.011) and shows only minor bias. However, the HMF with interlopers is biased to low Ω{sub m} and high σ{sub 8} and the fiducial model lies well outside of the forecast constraints, prior to accounting for Eddington bias. When the VDF is combined with constraints from the cosmic microwave background, the degeneracy between cosmological parameters can be significantly reduced. Upcoming spectroscopic surveys that probe larger volumes and fainter magnitudes will provide clusters for applying the VDF as a cosmological probe.

79 ASTRONOMY AND ASTROPHYSICS↗

Evaluating soil moisture feedback on convective triggering: Roles of convective and land–model parameterizations

Feedbacks between soil moisture and convective precipitation are not well represented in current Earth system models, and can affect projections of drought and heavy-precipitation extremes. To explore the atmospheric and land-surface influences on the initiation (triggering) of daytime deep convection, single-column model experiments were performed using the NCAR Community Earth System Model (CESM1.2), over the U.S. Southern Great Plains. The results indicate that the positive and negative feedback mechanisms found in earlier studies (using simplified boundary-layer models) are robust to large-scale forcing and interactions between boundary-layer turbulence, cloud dynamics, and radiation. However, systematic responses of convective triggering to soil moisture emerged only after switching from a convective available potential energy-based to a convective inhibition energy-based convective parameterization. This suggests that the choice of convective mass-flux closure largely determines the sensitivity of parameterized convective clouds to land-surface state. Errors in land-model parameterizations of evapotranspiration also affect the probability of deep convection, with vegetation (transpiration) playing an important role in linking soil moisture to surface energy partitioning and clouds. Parameterizations that permit the triggering feedback mechanism better predict the statistics of daytime shallow and deep convection, with respect to observations from the Atmospheric Radiation Measurement site in the Southern Great Plains. Here, the results illustrate how errors in the representation of evapotranspiration in land-surface models can propagate through the chain of coupled land-atmosphere processes to affect convective clouds.

54 ENVIRONMENTAL SCIENCES↗

Error analysis of a hybrid control drum worth model

This paper presents a perturbation-based model for control drum worth prediction which employs both physics-based and statistics-based components. Control drums, or control shims, are cylindrical in shape and span the axial length of the core. A portion of the cylinder is coated in neutron absorbing material and the drum can rotate to introduce the absorbing material to the body of the core to reduce reactivity. The model can be expensive to create due to the requirement for full-core Monte Carlo eigenvalue calculations. Therefore, it is important to analyze how the errors in Monte Carlo calculated k{sub eff} used for model training affect model performance. It was found that the error in predicted criticalities could average to 70 pcm in the most complex form of the model and 215 pcm in the simplest form of the model. Furthermore, it was found that the Monte Carlo uncertainty in quantities calculated with Serpent used to train the models had minimal impact on the error observed from the model. Lastly, one of the forms of the hybrid model could be trained in considerably less computational time if the Monte Carlo calculations were run to higher uncertainty in k{sub eff} with a small penalty to model performance.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Evaluation of NMME temperature and precipitation bias and forecast skill for South Asia

Systematic error and forecast skill for temperature and precipitation in two regions of Southern Asia are investigated using hindcasts initialized May 1 from the North American Multi-Model Ensemble. We focus on two contiguous but geographically and dynamically diverse regions: the Extended Indian Monsoon Rainfall (70–100E, 10–30 N) and the nearby mountainous area of Pakistan and Afghanistan (60–75E, 23–39 N). Forecast skill is assessed using the Sign test framework, a rigorous statistical method that can be applied to non-Gaussian variables such as precipitation and to different ensemble sizes without introducing bias. We find that models show significant systematic error in both precipitation and temperature for both regions. The multi-model ensemble mean (MMEM) consistently yields the lowest systematic error and the highest forecast skill for both regions and variables. However, we also find that the MMEM consistently provides a statistically significant increase in skill over climatology only in the first month of the forecast. While the MMEM tends to provide higher overall skill than climatology later in the forecast, the differences are not significant at the 95% level. We also find that MMEMs constructed with a relatively small number of ensemble members per model can equal or outperform MMEMs constructed with more members in skill. This suggests some ensemble members either provide no contribution to overall skill or even detract from it.

54 ENVIRONMENTAL SCIENCES↗

0.94-2.42 μm ground-based transmission spectra of the hot Jupiter HD-189733b

We present here new transmission spectra of the hot Jupiter HD-189733b using the SpeX instrument on the NASA Infrared Telescope Facility. We obtained two nights of observations where we recorded the primary transit of the planet in the J, H, and K bands simultaneously, covering a spectral range from 0.94 to 2.42 μm. We used Fourier analysis and other detrending techniques validated previously on other data sets to clean the data. We tested the statistical significance of our results by calculating the autocorrelation function, and we found that, after the detrending, autocorrelative noise is diminished at most frequencies. Additionally, we repeated our analysis on the out-of-transit data only, showing that the residual telluric contamination is well within the error bars. While these techniques are very efficient when multiple nights of observations are combined together, our results prove that even one good night of observations is enough to provide statistically meaningful data. Our observed spectra are consistent with space-based data recorded in the same wavelength interval by multiple instruments, indicating that ground-based facilities are becoming a viable and complementary option to spaceborne observatories. The best fit to the features in our data was obtained with water vapor. Our error bars are not small enough to address the presence of additional molecules; however, by combining the information contained in other data sets with our results, it is possible to explain all the available observations with a modeled atmospheric spectrum containing water vapor, methane, carbon monoxide, and hazes/clouds.

79 ASTRONOMY AND ASTROPHYSICS↗

Scalar excursions in large-eddy simulations

Here, the range of values of scalar fields in turbulent flows is bounded by their boundary values, for passive scalars, and by a combination of boundary values, reaction rates, phase changes, etc., for active scalars. The current investigation focuses on the local conservation of passive scalar concentration fields and the ability of the large-eddy simulation (LES) method to observe the boundedness of passive scalar concentrations. In practice, as a result of numerical artifacts, this fundamental constraint is often violated with scalars exhibiting unphysical excursions. The present study characterizes passive-scalar excursions in LES of a shear flow and examines methods for diagnosis and assesment of the problem. The analysis of scalar-excursion statistics provides support of the main hypothesis of the current study that unphysical scalar excursions in LES result from dispersive errors of the convection-term discretization where the subgrid-scale model (SGS) provides insufficient dissipation to produce a sufficiently smooth scalar field. In the LES runs three parameters are varied: the discretization of the convection terms, the SGS model, and grid resolution. Unphysical scalar excursions decrease as the order of accuracy of non-dissipative schemes is increased, but the improvement rate decreases with increasing order of accuracy. Two SGS models are examined, the stretched-vortex and a constant-coefficient Smagorinsky. Scalar excursions strongly depend on the SGS model. The excursions are significantly reduced when the characteristic SGS scale is set to double the grid spacing in runs with the stretched-vortex model. The maximum excursion and volume fraction of excursions outside boundary values show opposite trends with respect to resolution. The maximum unphysical excursions increase as resolution increases, whereas the volume fraction decreases. The reason for the increase in the maximum excursion is statistical and traceable to the number of grid points (sample size) which increases with resolution. In contrast, the volume fraction of unphysical excursions decreases with resolution because the SGS models explored perform better at higher grid resolution.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

The application of Plant Reliability Data Information System (PRINS) to CANDU reactor

As risk-informed applications (RIAs) are actively implanted in the nuclear industry, an issue associated with technical adequacy of Probabilistic Safety Assessment (PSA) arises in its modeling and data sourcing. In Korea, PSA for all Korean NPPs has been completed and KHNP(Korea Hydro and Nuclear Power Plant Company) developed the database called the Plant Reliability Data Information System (PRinS). It has several characteristics that distinguish it from other database system such as NPRDs (INPO,1994), PRIS (IAEA), and SRDF (EdF). This database has the function of systematic data management such as automatic data-gathering, periodic data deposition and updating, statistical analysis including Bayesian method, and trend analysis of failure rate or unavailability. In recent PSA for CANDU reactor, the component failure data of EPRI ALWR URD and Component Reliability Database were preferentially used as generic data set. The error factor for most component failure data was estimated by using the information NUREG/CR-4550 and NUREG/CR-4639. Also, annual trend analysis was performed for the functional losses of components using the statistical analysis and chart module of PRinS. Furthermore, the database has been updated regularly and maintained as a living program to reflect the current status. This paper presents the failure data analysis using PRinS which provides Bayesian analysis on main components in the CANDU reactor. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

A Performance Comparison of Low-Cost Near-Infrared (NIR) Spectrometers to a Conventional Laboratory Spectrometer for Rapid Biomass Compositional Analysis

The performance of a conventional laboratory near-infrared (NIR) spectrometer and two NIR spectrometer prototypes (a Texas Instruments NIRSCAN Nano evaluation model (EVM) and an InnoSpectra NIR-M-R2 spectrometer) are compared by collecting reflectance spectra of 270 well-characterized herbaceous biomass samples, building calibration models using the partial least squares (PLS-2) algorithm to predict five constituents of the samples from the reflectance spectra, and comparing the resulting model statistics. The prediction models developed using spectra from the Foss XDS spectrometer were slightly better than the prediction models developed using spectra from either the TI NIRSCAN Nano EVM and the InnoSpectra NIR-M-R2 as measured by the root mean square error (RMSECV) and the correlation coefficient (R 2 _cv) for “leave-one-out” cross-validation (CV). The models built from the two prototype units were not statistically significantly different from each other (p = 0.05). The Foss spectrometer has a larger wavelength range (400–2500 nm) compared with the two prototypes (900–1700 nm). When the spectra from the Foss XDS spectrometer were truncated so their wavelength range matched the wavelength range of the two prototype units, the resulting model was not statistically significantly different from the models from either prototype.

47 OTHER INSTRUMENTATION↗

Frankenstein’s ROMster: Avoiding pitfalls of reduced-order model development

Reduced-order models (ROMs) are a widely used and powerful approach to reducing the complexity of predictive physics-based numerical simulations for a wide range of applications, including electronics and fluid mechanics such as geologic CO 2 sequestration (GCS). ROMs are critical for optimization, sensitivity analysis, model calibration and uncertainty quantification where full-order models cannot be feasibly executed many times. Traditional approaches generate a single ROM for each simulated response (e.g., CO 2 injection rates, pH changes) based on a set of training simulations. Here, we demonstrate that a single ROM can display excellent overall predictive statistics, but have predictions that dramatically and unacceptably deviate from simulator responses especially when the response variable has a large range (i.e., vary over multiple orders of magnitude). For example, we show that a traditional statistically-high-performing GCS ROM (coefficient of determination R 2 of 0.99) can have average absolute relative errors of over 200%. To address this, we propose a new and novel approach where a set of sub-ROMs are generated to overcome the potential pitfalls in traditional single ROM development. The effectiveness of the proposed approach—the ROMster framework—is demonstrated using a case study of predicted CO 2 injection rates for GCS. We find our approach is a robust and general framework for ROM development, reducing the average “error” from 200% to only 4% in our case study.

54 ENVIRONMENTAL SCIENCES↗

On the burn topology of hot-spot-initiated reactions

We determine the reaction progress function for an ideal hot spot model problem. The considered problem has an exact analytic solution that can derived from a reduction of Nichols statistical hot spot model. We perform numerical calculations to verify the analytic solution and to illustrate the error realized in real, finite systems. We show how the baseline problem, which does not distinguish between the reactant and product densities, can be scaled to handle general cases for which the two densities differ.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

OBJECTS IN KEPLER'S MIRROR MAY BE LARGER THAN THEY APPEAR: BIAS AND SELECTION EFFECTS IN TRANSITING PLANET SURVEYS

Statistical analyses of large surveys for transiting planets such as the Kepler mission must account for systematic errors and biases. Transit detection depends not only on the planet's radius and orbital period, but also on host star properties. Thus, a sample of stars with transiting planets may not accurately represent the target population. Moreover, targets are selected using criteria such as a limiting apparent magnitude. These selection effects, combined with uncertainties in stellar radius, lead to biases in the properties of transiting planets and their host stars. We quantify possible biases in the Kepler survey. First, Eddington bias produced by a steep planet radius distribution and uncertainties in stellar radius results in a 15%-20% overestimate of planet occurrence. Second, the magnitude limit of the Kepler target catalog induces Malmquist bias toward large, more luminous stars and underestimation of the radii of about one-third of candidate planets, especially those larger than Neptune. Third, because metal-poor stars are smaller, stars with detected planets will be very slightly (<0.02 dex) more metal-poor than the target average. Fourth, uncertainties in stellar radii produce correlated errors in planet radius and stellar irradiation. A previous finding, that highly irradiated giants are more likely to have 'inflated' radii, remains significant, even accounting for this effect. In contrast, transit depth is negatively correlated with stellar metallicity even in the absence of any intrinsic correlation, and a previous claim of a negative correlation between giant planet transit depth and stellar metallicity is probably an artifact.

79 ASTRONOMY AND ASTROPHYSICS↗

A TWO-YEAR TIME DELAY FOR THE LENSED QUASAR SDSS J1029+2623

We present 279 epochs of optical monitoring data spanning 5.4 years from 2007 January to 2012 June for the largest image separation (22.''6) gravitationally lensed quasar, SDSS J1029+2623. We find that image A leads the images B and C by {Delta} t {sub AB} = (744 {+-} 10) days (90% confidence); the uncertainty includes both statistical uncertainties and systematic differences due to the choice of models. With only a {approx}1% fractional error, the interpretation of the delay is limited primarily by cosmic variance due to fluctuations in the mean line-of-sight density. We cannot separate the fainter image C from image B, but since image C trails image B by only 2-3 days in all models, the estimate of the time delay between images A and B is little affected by combining the fluxes of images B and C. There is weak evidence for a low level of microlensing, perhaps created by the small galaxy responsible for the flux ratio anomaly in this system. Interpreting the delay depends on better constraining the shape of the gravitational potential using the lensed host galaxy, other lensed arcs, and the structure of the X-ray emission.

79 ASTRONOMY AND ASTROPHYSICS↗

Omega Electroproduction

The differential cross section for p(e, e'ω)p has been studied at Q{sup 2} ~ 5.5 (GeV/c)2. Here Q{sup 2} represents the four momentum squared of the virtual photon in the excitation of baryonic resonances by an electron projectile. In order to extract the ω-meson differential cross section from the JLAB data, the data was compared to a full Monte Carlo simulation of the detector based on events generated for omega production in a way that the production cross section was varied to achieve a match to the data. The bin selected for this procedure takes into account the measure of robustness of the stripping of the ω peak from the multi-pion background as well as the statistics in the measured data and the Monte Carlo simulation of the signal and background physics. An error estimation technique for the cross section was based on determining the dependence of the extracted cross section parameters on the experimental set-up (including parameters for the spectrometer, target beam geometeries and performance). We compare our results with a Regge-based model for hadronic content in the t-channel exchange of a photon in Q{sup 2} region of overlap. There is an extension of this data into a completely new region, which is the highest yet measured.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Tula: Optimizing Time, Cost, and Generalization in Distributed Large-Batch Training

Distributed training increases the number of batches processed per iteration either by scaling-out (adding more nodes) or scaling-up (increasing the batch-size). However, the largest configuration does not necessarily yield the best performance. Horizontal scaling introduces additional communication overhead, while vertical scaling is constrained by computation cost and device memory limits. Thus, simply increasing the batch-size leads to diminishing returns: training time and cost decrease initially but eventually plateaus, creating a knee-point in the time/cost vs. batch-size pareto curve. The optimal batch-size therefore depends on the underlying model, data and available compute resources. Large batches also suffer from worse model quality due to the well-known “generalization gap”. In this paper, we present Tula, an online service that automatically optimizes time, cost, and convergence quality for large-batch training of convolutional models. It combines parallel-systems modeling with statistical performance prediction to identify the optimal batchsize. Tula predicts training time and cost within 7.5−14% error across multiple models, and achieves up to 20× overall speedup and improves test accuracy by ≈9% on average over standard large-batch training on various vision tasks, thus successfully mitigating the generalization gap and accelerating training at the same time.

Tyagi, Sahil [ORNL] (ORCID:0009000783144745)↗

Computer-aided measurement of liver volumes in CT by means of geodesic active contour segmentation coupled with level-set algorithms

Purpose: Computerized liver extraction from hepatic CT images is challenging because the liver often abuts other organs of a similar density. The purpose of this study was to develop a computer-aided measurement of liver volumes in hepatic CT. Methods: The authors developed a computerized liver extraction scheme based on geodesic active contour segmentation coupled with level-set contour evolution. First, an anisotropic diffusion filter was applied to portal-venous-phase CT images for noise reduction while preserving the liver structure, followed by a scale-specific gradient magnitude filter to enhance the liver boundaries. Then, a nonlinear grayscale converter enhanced the contrast of the liver parenchyma. By using the liver-parenchyma-enhanced image as a speed function, a fast-marching level-set algorithm generated an initial contour that roughly estimated the liver shape. A geodesic active contour segmentation algorithm coupled with level-set contour evolution refined the initial contour to define the liver boundaries more precisely. The liver volume was then calculated using these refined boundaries. Hepatic CT scans of 15 prospective liver donors were obtained under a liver transplant protocol with a multidetector CT system. The liver volumes extracted by the computerized scheme were compared to those traced manually by a radiologist, used as ''gold standard.''Results: The mean liver volume obtained with our scheme was 1504 cc, whereas the mean gold standard manual volume was 1457 cc, resulting in a mean absolute difference of 105 cc (7.2%). The computer-estimated liver volumetrics agreed excellently with the gold-standard manual volumetrics (intraclass correlation coefficient was 0.95) with no statistically significant difference (F=0.77; p(F{<=}f)=0.32). The average accuracy, sensitivity, specificity, and percent volume error were 98.4%, 91.1%, 99.1%, and 7.2%, respectively. Computerized CT liver volumetry would require substantially less completion time (compared to an average of 39 min per case by manual segmentation). Conclusions: The computerized liver extraction scheme provides an efficient and accurate way of measuring liver volumes in CT.

62 RADIOLOGY AND NUCLEAR MEDICINE↗