SEARCH · Engineering Papers

Results for “latent 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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 73 records · Page 4

All sources NASA NTRS DOE OSTI

NASA NTRS2003

Vertical Profiles of Latent Heat Release over the Global Tropics Using TRMM Rainfall Products from December 1997 to November 2002

NASA Tropical Rainfall Measuring Mission (TRMM) precipitation radar (PR) derived rainfall information will be used to estimate the four-dimensional structure of global monthly latent heating and rainfall profiles over the global tropics from December 1997 to November 2000. Rainfall, latent heating and radar reflectivity structures between El Nino (DJF 1997-98) and La Nina (DJF 1998-99) will be examined and compared. The seasonal variation of heating over various geographic locations (i.e., oceanic vs continental, Indian ocean vs west Pacific, Africa vs S. America) will also be analyzed. In addition, the relationship between rainfall, latent heating (maximum heating level), radar reflectivity and SST is examined and will be presented in the meeting. The impact of random error and bias in straitform percentage estimates from PR on latent heating profiles is studied and will also be presented in the meeting. The Goddard Cumulus Ensemble Model is being used to simulate various mesoscale convective systems that developed in different geographic locations. Specifically, the model estimated rainfall, radar reflectivity and latent heating profiles will be compared to observational data collected from TRMM field campaigns over the South China Sea in 1998 (SCSMXX), Brazil in 1999 (TRMM- LBA), and the central Pacific in 1999 (KWAJEX). Sounding diagnosed heating budgets and radar reflectivity from these experiments can provide the means to validate (heating product) as well as improve the GCE model.

Tao, W.-K.↗

NASA NTRS2003

Vertical Profiles of Latent Heat Release over the Global Tropics using TRMM Rainfall Products from December 1997 to November 2002

NASA Tropical Rainfall Measuring Mission (TRMM) precipitation radar (PR) derived rainfall information will be used to estimate the four-dimensional structure of global monthly latent heating and rainfall profiles over the global tropics from December 1997 to November 2000. Rainfall, latent heating and radar reflectivity structures between El Nino (DJF 1997-98) and La Nina (DJF 1998-99) will be examined and compared. The seasonal variation of heating over various geographic locations (i.e., oceanic vs continental, Indian ocean vs west Pacific, Africa vs. S. America ) will also be analyzed. In addition, the relationship between rainfall, latent heating (maximum heating level), radar reflectivity and SST is examined and will be presented in the meeting. The impact of random error and bias in stratiform percentage estimates from PR on latent heating profiles is studied and will also be presented in the meeting. The Goddard Cumulus Ensemble Model is being used to simulate various mesoscale convective systems that developed in different geographic locations. Specifically, the model estimated rainfall, radar reflectivity and latent heating profiles will be compared to observational data collected from TRMM field campaigns over the South China Sea in 1998 (SCSMEX), Brazil in 1999 (TRMM-LBA), and the central Pacific in 1999 (KWAJEX). Sounding diagnosed heating budgets and radar reflectivity from these experiments can provide the means to validate (heating product) as well as improve the GCE model. Review of other latent heating algorithms will be discussed in the workshop.

Tao, W.-K.↗

DOE OSTI2022

Learning in continuous action space for developing high dimensional potential energy models

Reinforcement learning (RL) approaches that combine a tree search with deep learning have found remarkable success in searching exorbitantly large, albeit discrete action spaces, as in chess, Shogi and Go. Many real-world materials discovery and design applications, however, involve multi-dimensional search problems and learning domains that have continuous action spaces. Exploring high-dimensional potential energy models of materials is an example. Traditionally, these searches are time consuming (often several years for a single bulk system) and driven by human intuition and/or expertise and more recently by global/local optimization searches that have issues with convergence and/or do not scale well with the search dimensionality. Here, in a departure from discrete action and other gradient-based approaches, we introduce a RL strategy based on decision trees that incorporates modified rewards for improved exploration, efficient sampling during playouts and a “window scaling scheme" for enhanced exploitation, to enable efficient and scalable search for continuous action space problems. Using high-dimensional artificial landscapes and control RL problems, we successfully benchmark our approach against popular global optimization schemes and state of the art policy gradient methods, respectively. We demonstrate its efficacy to parameterize potential models (physics based and high-dimensional neural networks) for 54 different elemental systems across the periodic table as well as alloys. We analyze error trends across different elements in the latent space and trace their origin to elemental structural diversity and the smoothness of the element energy surface. Broadly, our RL strategy will be applicable to many other physical science problems involving search over continuous action spaces.

36 MATERIALS SCIENCE↗

DOE OSTI2020

Block Krylov Subspace Methods for Functions of Matrices II: Modified Block FOM

We analyze an expansion of the generalized block Krylov subspace framework of [Electron. Trans. Numer. Anal., 47 (2017), pp. 100--126]. This expansion allows the use of low-rank modifications of the matrix projected onto the block Krylov subspace and contains, as special cases, the block GMRES method and the new block Radau--Arnoldi method. Within this general setting, we present results that extend the interpolation property from the nonblock case to a matrix polynomial interpolation property for the block case, and we relate the eigenvalues of the projected matrix to the latent roots of these matrix polynomials. Some error bounds for these modified block FOM methods for solving linear systems are presented. We then show how cospatial residuals can be preserved in the case of families of shifted linear block systems. This result is used to derive computationally practical restarted algorithms for block Krylov approximations that compute the action of a matrix function on a set of several vectors simultaneously. Finally, we prove some error bounds and present numerical results showing that two modifications of FOM, the block harmonic and the block Radau--Arnoldi methods for matrix functions, can significantly improve the convergence behavior.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

EdgeAI: Machine learning via direct attached accelerator for streaming data processing at high shot rate x-ray free-electron lasers

We present a case for low batch-size inference with the potential for adaptive training of a lean encoder model. We do so in the context of a paradigmatic example of machine learning as applied in data acquisition at high data velocity scientific user facilities such as the Linac Coherent Light Source-II x-ray Free-Electron Laser. We discuss how a low-latency inference model operating at the data acquisition edge can capitalize on the naturally stochastic nature of such sources. We simulate the method of attosecond angular streaking to produce representative results whereby simulated input data reproduce high-resolution ground truth probability distributions. By minimizing the mean-squared error between the decoded output of the latent representation and the ground truth distributions, we ensure that the encoding layers and resulting latent representation maintains full fidelity for any downstream task, be it classification or regression. We present throughput results for data-parallel inference of various batch sizes, some with throughput exceeding 100 k images per second. We also show in situ training below 10 s per epoch for the full encoder–decoder model as would be relevant for streaming and adaptive real-time data production at our nation’s scientific light sources.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1990

Measurement and analysis of workload effects on fault latency in real-time systems

The authors demonstrate the need to address fault latency in highly reliable real-time control computer systems. It is noted that the effectiveness of all known recovery mechanisms is greatly reduced in the presence of multiple latent faults. The presence of multiple latent faults increases the possibility of multiple errors, which could result in coverage failure. The authors present experimental evidence indicating that the duration of fault latency is dependent on workload. A synthetic workload generator is used to vary the workload, and a hardware fault injector is applied to inject transient faults of varying durations. This method makes it possible to derive the distribution of fault latency duration. Experimental results obtained from the fault-tolerant multiprocessor at the NASA Airlab are presented and discussed.

Woodbury, Michael H.↗

DOE OSTI2021

Machine learning of Kondo physics using variational autoencoders and symbolic regression

In this work, we employ variational autoencoders to extract physical insight from a dataset of one-particle Anderson impurity model spectral functions. Autoencoders are trained to find a low-dimensional, latent space representation that faithfully characterizes each element of the training set, as measured by a reconstruction error. Variational autoencoders, a probabilistic generalization of standard autoencoders, further condition the learned latent space to promote highly interpretable features. In our study, we find that the learned latent variables strongly correlate with well known, but nontrivial, parameters that characterize emergent behaviors in the Anderson impurity model. In particular, one latent variable correlates with particle-hole asymmetry, while another is in near one-to-one correspondence with the Kondo temperature, a dynamically generated low-energy scale in the impurity model. Using symbolic regression, we model this variable as a function of the known bare physical input parameters and "rediscover" the nonperturbative formula for the Kondo temperature. The machine learning pipeline we develop suggests a general purpose approach, which opens opportunities to discover new domain knowledge in other physical systems.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

NASA NTRS2014

Modeling Increased Complexity and the Reliance on Automation: FLightdeck Automation Problems (FLAP) Model

This paper highlights the development of a model that is focused on the safety issue of increasing complexity and reliance on automation systems in transport category aircraft. Recent statistics show an increase in mishaps related to manual handling and automation errors due to pilot complacency and over-reliance on automation, loss of situational awareness, automation system failures and/or pilot deficiencies. Consequently, the aircraft can enter a state outside the flight envelope and/or air traffic safety margins which potentially can lead to loss-of-control (LOC), controlled-flight-into-terrain (CFIT), or runway excursion/confusion accidents, etc. The goal of this modeling effort is to provide NASA's Aviation Safety Program (AvSP) with a platform capable of assessing the impacts of AvSP technologies and products towards reducing the relative risk of automation related accidents and incidents. In order to do so, a generic framework, capable of mapping both latent and active causal factors leading to automation errors, is developed. Next, the framework is converted into a Bayesian Belief Network model and populated with data gathered from Subject Matter Experts (SMEs). With the insertion of technologies and products, the model provides individual and collective risk reduction acquired by technologies and methodologies developed within AvSP.

Ancel, Ersin↗

DOE OSTI2018

Impact of numerical choices on water conservation in the E3SM Atmosphere Model version 1 (EAMv1)

Abstract. The conservation of total water is an important numerical feature for global Earth system models. Even small conservation problems in the water budget can lead to systematic errors in century-long simulations. This study quantifies and reduces various sources of water conservation error in the atmosphere component of the Energy Exascale Earth System Model. Several sources of water conservation error have been identified during the development of the version 1 (V1) model. The largest errors result from the numerical coupling between the resolved dynamics and the parameterized sub-grid physics. A hybrid coupling using different methods for fluid dynamics and tracer transport provides a reduction of water conservation error by a factor of 50 at 1∘ horizontal resolution as well as consistent improvements at other resolutions. The second largest error source is the use of an overly simplified relationship between the surface moisture flux and latent heat flux at the interface between the host model and the turbulence parameterization. This error can be prevented by applying the same (correct) relationship throughout the entire model. Two additional types of conservation error that result from correcting the surface moisture flux and clipping negative water concentrations can be avoided by using mass-conserving fixers. With all four error sources addressed, the water conservation error in the V1 model becomes negligible and insensitive to the horizontal resolution. The associated changes in the long-term statistics of the main atmospheric features are small. A sensitivity analysis is carried out to show that the magnitudes of the conservation errors in early V1 versions decrease strongly with temporal resolution but increase with horizontal resolution. The increased vertical resolution in V1 results in a very thin model layer at the Earth's surface, which amplifies the conservation error associated with the surface moisture flux correction. We note that for some of the identified error sources, the proposed fixers are remedies rather than solutions to the problems at their roots. Future improvements in time integration would be beneficial for V1.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2017

Impact of numerical choices on water conservation in the E3SM Atmosphere Model Version 1 (EAM V1)

The conservation of total water is an important numerical feature for global Earth system models. Even small conservation problems in the water budget can lead to systematic errors in century-long simulations for sea level rise projection. This study quantifies and reduces various sources of water conservation error in the atmosphere component of the Energy Exascale Earth System Model. Several sources of water conservation error have been identified during the development of the version 1 (V1) model. The largest errors result from the numerical coupling between the resolved dynamics and the parameterized sub-grid physics. A hybrid coupling using different methods for fluid dynamics and tracer transport provides a reduction of water conservation error by a factor of 50 at 1° horizontal resolution as well as consistent improvements at other resolutions. The second largest error source is the use of an overly simplified relationship between the surface moisture flux and latent heat flux at the interface between the host model and the turbulence parameterization. This error can be prevented by applying the same (correct) relationship throughout the entire model. Two additional types of conservation error that result from correcting the surface moisture flux and clipping negative water concentrations can be avoided by using mass-conserving fixers. With all four error sources addressed, the water conservation error in the V1 model is negligible and insensitive to the horizontal resolution. The associated changes in the long-term statistics of the main atmospheric features are small. A sensitivity analysis is carried out to show that the magnitudes of the conservation errors decrease strongly with temporal resolution but increase with horizontal resolution. The increased vertical resolution in the new model results in a very thin model layer at the Earth’s surface, which amplifies the conservation error associated with the surface moisture flux correction. We note that for some of the identified error sources, the proposed fixers are remedies rather than solutions to the problems at their roots. Future improvements in time integration would be beneficial for this model.

Zhang, Kai↗

DOE OSTI2026

Deep generative learning of magnetic frustration in artificial spin ice from magnetic force microscopy images

Increasingly large datasets of microscopic images with nanoscale resolution facilitate the development of machine learning methods to identify and analyze subtle physical phenomena embedded within the images. In this work, microscopic images of honeycomb lattice spin-ice samples serve as datasets from which we automate the calculation of net magnetic moments and directional orientations of spin-ice configurations. In the first stage of our workflow, machine learning models are trained to accurately predict magnetic moments and directions within spin-ice structures. Variational Autoencoders (VAEs), an emergent unsupervised deep learning technique, are employed to generate high-quality synthetic magnetic force microscopy (MFM) images and extract latent feature representations, thereby reducing experimental and segmentation errors. The second stage of proposed methodology enables precise identification and prediction of frustrated vertices and nanomagnetic segments, effectively correlating structural and functional aspects of microscopic images. This facilitates the design of optimized spin-ice configurations with controlled frustration patterns, enabling potential on-demand synthesis.

36 MATERIALS SCIENCE↗

DOE OSTI2020

Forward sensitivity approach for estimating eddy viscosity closures in nonlinear model reduction

In this paper, we propose a variational approach to estimate eddy viscosity using forward sensitivity method (FSM) for closure modeling in nonlinear reduced order models. FSM is a data assimilation technique that blends model's predictions with noisy observations to correct initial state and/or model parameters. We apply this approach on a projection based reduced order model (ROM) of the one-dimensional viscous Burgers equation with a square wave defining a moving shock, and the two-dimensional vorticity transport equation formulating a decay of Kraichnan turbulence. We investigate the capability of the approach to approximate an optimal value for eddy viscosity with different measurement configurations. Specifically, we show that our approach can sufficiently assimilate information either through full field or sparse noisy measurements to estimate eddy viscosity closure to cure standard Galerkin reduced order model (GROM) predictions. Therefore, our approach provides a modular framework to correct forecasting error from a sparse observational network on a latent space. We highlight that the proposed GROM-FSM framework is promising for emerging digital twin applications, where real-time sensor measurements can be used to update and optimize surrogate model's parameters.

42 ENGINEERING↗

NASA NTRS1978

System for and method of freezing biological tissue

Biological tissue is frozen while a polyethylene bag placed in abutting relationship against opposed walls of a pair of heaters. The bag and tissue are cooled with refrigerating gas at a time programmed rate at least equal to the maximum cooling rate needed at any time during the freezing process. The temperature of the bag, and hence of the tissue, is compared with a time programmed desired value for the tissue temperature to derive an error indication. The heater is activated in response to the error indication so that the temperature of the tissue follows the desired value for the time programmed tissue temperature. The tissue is heated to compensate for excessive cooling of the tissue as a result of the cooling by the refrigerating gas. In response to the error signal, the heater is deactivated while the latent heat of fusion is being removed from the tissue while the tissue is changing phase from liquid to solid.

Williams, T. E.↗

NASA NTRS2004

NASA GEOS-3/TRMM Re-analysis: Capturing Observed Tropical Rainfall Variability in Global Analysis for Climate Research

Understanding climate variability over a wide range of space-time scales requires a comprehensive description of the earth system. Global analyses produced by a fixed assimilation system (i.e., re-analyses) - as their quality continues to improve - have the potential of providing a vital tool for meeting this challenge. But at the present time, the usefulness of re-analyses is limited by uncertainties in such basic fields as clouds, precipitation, and evaporation - especially in the tropics, where observations are relatively sparse. Analyses of the tropics have long been shown to be sensitive to. the treatment of cloud precipitation processes, which remains a major source of uncertainty in current models. Yet, for many climate studies it is crucial that analyses can accurately reproduce the observed rainfall intensity and variability since a small error of 1 mm/d in surface rain translates into an error of approx. 30 W/sq m in energy (latent heat) flux. Currently, discrepancies between the observed and analyzed monthly-mean rain rates averaged to 100 km x 100 km resolution can exceed 4 mm/d (or 120 W/sq m ), compared to uncertainties in surface radiative fluxes of approx. 10-20 W/sq m . Improving precipitation in analyses would reduce a major source of uncertainty in the global energy budget. Uncertainties in tropical precipitation have also been a major impediment in understanding how the tropics interact with other regions, including the remote response to El Nino/Southern Oscillation (ENSO) variability on interannual time scales, the influence of Madden-Julian Oscillation (MJO) and monsoons on intraseasonal time scales. A global analysis that can replicate the observed precipitation variability together with physically consistent estimates of other atmospheric variables provides the key to breaking this roadblock. NASA Goddard Space Flight Center has been exploring the use of satellite-based microwave rainfall measurements in improving global analyses and has recently produced a multi-year, 1 x 1 TRMM re-analysis , which assimilates 6-hourly TMI and SSM/I surface rain rates over tropical oceans using a ID variational continuous assimilation (VCA) procedure in the GEOS-3 global data assimilation system. The analysis period extends from 1 November 1997 through 3 1 December 2002. The goal is to produce a multi-year global analysis that is dynamically consistent with available tropical precipitation observations for the community to assess its utility in climate applications and identify areas for further improvements. A distinct feature of the GEOS-3RRMh4 re-analysis is that its precipitation analysis is not derived from a short-term forecast (as done in most operational systems) but is given by a time- continuous model integration constrained by precipitation observations within a 6-h analysis window, while the wind, temperature, and pressure fields are allowed to directly respond to the improved precipitation and associated latent heating structures within the same analysis window. In this talk, I will assess the impact VCA precipitation assimilation on analyses of climate signals ranging from a few weeks to interannual time scales and compare results against other operational and reanalysis products.

Hou, Arthur Y.↗

NASA NTRS1994

Parameterization of bulk condensation in numerical cloud models

The accuracy of the moist saturation adjustment scheme has been evaluated using a three-dimensional explicit microphysical cloud model. It was found that the error in saturation adjustment depends strongly on the Cloud Condensation Nucleii (CCN) concentration in the ambient atmosphere. The scheme provides rather accurate results in the case where a sufficiently large number of CCN (on the order of several hundred per cubic centimeter) is available. However, under conditions typical of marine stratocumulus cloud layers with low CCN concentration, the error in the amounts of condensed water vapor and released latent heat may be as large as 40%-50%. A revision of the saturation adjustment scheme is devised that employs the CCN concentration, dynamical supersaturation, and cloud water content as additional variables in the calculation of the condensation rate. The revised condensation model reduced the error in maximum updraft and cloud water content in the climatically significant case of marine stratocumulus cloud layers by an order of magnitude.

Kogan, Yefim L.↗

DOE OSTI2022

Towards sharp error analysis of extended Lagrangian molecular dynamics

The extended Lagrangian molecular dynamics (XLMD) method provides a useful framework for reducing the computational cost of a class of molecular dynamics simulations with constrained latent variables. The XLMD method relaxes the constraints by introducing a fictitious mass ε for the latent variables and solving a set of singularly perturbed ordinary differential equations. While favorable numerical performance of XLMD has been demonstrated in several different contexts in the past decade, mathematical analysis of the method remains scarce. Here, we propose the first error analysis of the XLMD method in the context of a classical polarizable force field model. While the dynamics with respect to the atomic degrees of freedom are general and nonlinear, the key mathematical simplification of the polarizable force field model is that the constraints on the latent variables are given by a linear system of equations. We prove that when the initial value of the latent variables is compatible in a sense that we define, XLMD converges as the fictitious mass ε is made small with $\mathscr{O}$(ε) error for the atomic degrees of freedom and with $\mathscr{O}$($\sqrt{ε}$) error for the latent variables, when the dimension of the latent variable d' is 1. Furthermore, when the initial value of the latent variables is improved to be optimally compatible in a certain sense, we prove that the convergence rate can be improved to $\mathscr{O}$(ε) for the latent variables as well. Numerical results verify that both estimates are sharp not only for d'=1, but also for arbitrary d'. In the setting of general d', we do obtain convergence, but with the non-sharp rate of $\mathscr{O}$($\sqrt{ε}$) for both the atomic and latent variables.

74 ATOMIC AND MOLECULAR PHYSICS↗

NASA NTRS2000

The Estimation of Surface Latent Heat Flux Over the Ocean and its Relationship to Marine Atmospheric Boundary Layer (MABL) Structure

A new technique combining active and passive remote sensing instruments for the estimation of surface latent heat flux over the ocean is presented. This synergistic method uses aerosol lidar backscatter data, multi-channel infrared radiometer data and microwave scatterometer data acquired onboard the NASA P-3B research aircraft during an extended field campaign over the Atlantic ocean in support of the Lidar In-space Technology Experiment (LITE) in September of 1994. The 10 meter wind speed derived from the scatterometers and the lidar-radiometer inferred near-surface moisture are used to obtain an estimate of the surface flux of moisture via bulk aerodynamic formulae. The results are compared with the Special Sensor Microwave Imager (SSM/I) daily average latent heat flux and show reasonable agreement with an rms error and bias of about 50 and 25 W per square meters, respectively. In addition, the MABL height, entrainment zone thickness and integrated lidar backscatter intensity are computed from the lidar data and compared with the magnitude of the surface fluxes. The results show that the surface latent heat flux is most strongly correlated with entrainment zone top, bottom and the integrated MABL lidar backscatter, with corresponding correlation coefficients of 0.62, 0.67 and 0.61, respectively.

Palm, Stephen P.↗

NASA NTRS2004

Evaluation and Application of Satellite-Based Latent Heating Profile Estimation Methods

In recent years, methods for estimating atmospheric latent heating vertical structure from both passive and active microwave remote sensing have matured to the point where quantitative evaluation of these methods is the next logical step. Two approaches for heating algorithm evaluation are proposed: First, application of heating algorithms to synthetic data, based upon cloud-resolving model simulations, can be used to test the internal consistency of heating estimates in the absence of systematic errors in physical assumptions. Second, comparisons of satellite-retrieved vertical heating structures to independent ground-based estimates, such as rawinsonde-derived analyses of heating, provide an additional test. The two approaches are complementary, since systematic errors in heating indicated by the second approach may be confirmed by the first. A passive microwave and combined passive/active microwave heating retrieval algorithm are evaluated using the described approaches. In general, the passive microwave algorithm heating profile estimates are subject to biases due to the limited vertical heating structure information contained in the passive microwave observations. These biases may be partly overcome by including more environment-specific a priori information into the algorithm s database of candidate solution profiles. The combined passive/active microwave algorithm utilizes the much higher-resolution vertical structure information provided by spaceborne radar data to produce less biased estimates; however, the global spatio-temporal sampling by spaceborne radar is limited. In the present study, the passive/active microwave algorithm is used to construct a more physically-consistent and environment-specific set of candidate solution profiles for the passive microwave algorithm and to help evaluate errors in the passive algorithm s heating estimates. Although satellite estimates of latent heating are based upon instantaneous, footprint- scale data, suppression of random errors requires averaging to at least half-degree resolution. Analysis of mesoscale and larger space-time scale phenomena based upon passive and passive/active microwave heating estimates from TRMM, SSMI, and AMSR data will be presented at the conference.

Olson, William S.↗