A Simple Method for Predicting N H Variability in Active Galactic Nuclei
Explore the source record for details and available documents.
Isaiah S. Cox↗
SEARCH · Engineering Papers
Results for “variability prediction”
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.
A provisional analysis of two-dimensional turbulent mixing with variable density
A predictive method for the titled flows based on the Prandtl energy method was developed and assessed by comparing predicted results with experimental results. For constant-density flows, both gross properties such as spreading rate and maximum turbulent kinetic energy and detailed properties such as mean shear stress distributions are shown to be well predicted. For variable-density flows, considerable attention is devoted to the inclusion in the analysis of the added effect of pressure fluctuations and of the variation in the several extant empirical parameters on the turbulent kinetic energy. It is found that a variation with Mach number of the characteristic Reynolds number for turbulent transport is needed to account for the observed decrease in spreading rate. The predictions which result from these considerations are compared with the limited experimental data presently available for the two crucial cases: compressible adiabatic mixing and low-speed isothermal mixing of two dissimilar gases.
Libby, P. A.↗
A deep learning-based workflow for fast prediction of 3D state variables in geological carbon storage: A dimension reduction approach
Deep learning (DL) models are extensively used as surrogate models for high-fidelity simulations of multiphase fluid flow in porous media at large scales, enabling fast forecasts of the spatial–temporal evolution of three-dimensional (3D) state variables in geological carbon storage (GCS). However, training these models in high-dimensional space remains computationally demanding and prone to overfitting because of limited training data. This paper presents a novel workflow to address these challenges by integrating dimension reduction (DR) methods. Here, the proposed workflow employed pre-trained DR models to extract the latent variables of geological models and state variables and utilized the multi-layer perceptron (MLP) for constructing mapping functions between the input and output variables in latent spaces. Subsequently, the pre-trained reconstruction models converted the MLP-predicted latent state variables to their original high-dimensional form. Furthermore, we proposed a novel strategy for the DR and reconstruction of 3D saturation fields to account for the unique data characteristics of sparsity, nonuniformity, and discontinuity. The proposed strategy applied PCA and inverse PCA for 2D average saturation fields and developed a DL-based 3D reconstruction model, leveraging three 2D average saturation fields as input to produce a 3D saturation field as output. The pre-training of DR and reconstruction models and training of MLP models were conducted on 84 Gulf of Mexico (GoM) simulations and evaluated on 12 testing simulations. Each simulation contained 720 monthly time steps, with the first 360 months as the injection period and the rest as the post-injection period. The proposed workflow, incorporating DR and DL models, accurately predicts the normalized 3D pressure fields, achieving mean square error (MSE) of 2.92 × 10 -7 compared to the ground truth obtained from a full-physics simulator. Furthermore, the proposed strategy outperformed PCA and convolutional autoencoder (CAE) models on 3D saturation fields, resulting in minor workflow prediction errors with an MSE of 2.93 × 10 -5 . The results suggest the proposed workflow provides sufficient predictive fidelity across temporal and spatial scales, and enables a speedup of 160 times compared to the full-physics simulator, facilitating improved decision-making and risk assessment for large-scale GCS management in real-time scenarios.
3D reconstruction model↗
Predicting the heat release variability of Li-ion cells under thermal runaway with few or no calorimetry data
Accurate measurement of the variability of thermal runaway behavior of lithium-ion cells is critical for designing safe battery systems. However, experimentally determining such variability is challenging, expensive, and time-consuming. Here, we utilize a transfer learning approach to accurately estimate the variability of heat output during thermal runaway using only ejected mass measurements and cell metadata, leveraging 139 calorimetry measurements on commercial lithium-ion cells available from the open-access Battery Failure Databank. We show that the distribution of heat output, including outliers, can be predicted accurately and with high confidence for new cell types using just 0 to 5 calorimetry measurements by leveraging behaviors learned from the Battery Failure Databank. Fractional heat ejection from the positive vent, cell body, and negative vent are also accurately predicted. We demonstrate that by using low cost and fast measurements, we can predict the variability in thermal behaviors of cells, thus accelerating critical safety characterization efforts.
25 ENERGY STORAGE↗
Solar weather/climate predictions
Solar variability influences upon terrestrial weather and climate are addressed. Both the positive and negative findings are included and specific predictions, areas of further study, and recommendations listed.
Schatten, K. H.↗
Prediction of natural frequency variability due to uncertainty in material properties
Composite materials are widely used in various types of modern engineering structures. Traditional studies on composite structures have been based on the assumption that the material properties of the composites are characterized by a priori known elastic moduli, and no uncertainties of these moduli have been considered. However, the composite materials are invariably subject to a certain amount of scatter in their measured elastic moduli. To a large extent, the properties of composite materials are dependent on the fabrication process. But even the composite materials manufactured by the same process demonstrate differences in their elastic properties. This paper proposes a new, non-probabilistic method to predict the variability in the natural frequencies of the composite cylindrical shell, resulting from the unavoidable scatter in elastic moduli. The available measurements of elastic moduli are fitted by the four-dimensional uncertainty ellipsoid. The upper and lower bounds of the natural frequency are derived. With these bounds, designers will have a better understanding of the real dynamic behavior of the structure.
Li, Y. W.↗
Chaos in the sunspot cycle - Analysis and prediction
The variability of solar activity over long time scales, given semiquantitatively by measurements of sunspot numbers, is examined as a nonlinear dynamical system. First, a discussion of the data set used and the techniques utilized to reduce the noise and capture the long-term dynamics inherent in the data is presented. Subsequently, an attractor is reconstructed from the data set using the method of time delays. The reconstructed attractor is then used to determine both the dimension of the underlying system and also the largest Lyapunov exponent, which together indicate that the sunspot cycle is indeed chaotic and also low dimensional. In addition, recent techniques of exploiting chaotic dynamics to provide accurate, short-term predictions are utilized in order to improve upon current forecasting methods and also to place theoretical limits on predictability extent. The results are compared to chaotic solar-dynamo models as a possible physically motivated source of this chaotic behavior.
Mundt, Michael D.↗
Implicit learning of convective organization explains precipitation stochasticity
Accurate prediction of precipitation intensity is crucial for both human and natural systems, especially in a warming climate more prone to extreme precipitation. Yet, climate models fail to accurately predict precipitation intensity, particularly extremes. One missing piece of information in traditional climate model parameterizations is subgrid-scale cloud structure and organization, which affects precipitation intensity and stochasticity at coarse resolution. Here, using global storm-resolving simulations and machine learning, we show that, by implicitly learning subgrid organization, we can accurately predict precipitation variability and stochasticity with a low-dimensional set of latent variables. Using a neural network to parameterize coarse-grained precipitation, we find that the overall behavior of precipitation is reasonably predictable using large-scale quantities only; however, the neural network cannot predict the variability of precipitation (R 2 ~ 0.45) and underestimates precipitation extremes. The performance is significantly improved when the network is informed by our organization metric, correctly predicting precipitation extremes and spatial variability (R 2 ~ 0.9). The organization metric is implicitly learned by training the algorithm on a high-resolution precipitable water field, encoding the degree of subgrid organization. The organization metric shows large hysteresis, emphasizing the role of memory created by subgrid-scale structures. We demonstrate that this organization metric can be predicted as a simple memory process from information available at the previous time steps. These findings stress the role of organization and memory in accurate prediction of precipitation intensity and extremes and the necessity of parameterizing subgrid-scale convective organization in climate models to better project future changes of water cycle and extremes.
54 ENVIRONMENTAL SCIENCES↗
Key Environmental and Ecological Variables of Wetland CH 4 and CO 2 Fluxes Change With Warming
Wetlands are important ecosystems for the global carbon cycle, impacting regional and global methane (CH 4 ) and carbon dioxide (CO 2 ) budgets. This study examines how environmental and ecological variables impact wetland CH 4 flux and net ecosystem exchange of CO 2 (NEE) across 17 sites globally. We also quantified the importance of variables for each wetland type and site at monthly scale under normal and warm temperatures using dominance analysis. We identified soil and air temperature (TS, TA, respectively) as key variables influencing wetland CH4, and latent heat (LE) and shortwave radiation (SW) for NEE under normal and warm conditions. However, the importance of some variables shifted with warming. For predicting the variability of wetland CH4 flux under warming, gross primary productivity (GPP) and LE, replacing wind direction (WD), were dominant variables for tropical swamps, while NEE was important for high-latitude fens and bogs under warm temperatures. For wetland NEE, the role of TA and TS decreased across all wetland types with warming, while vapor pressure deficit (VPD) became more important for mid and high-latitude wetlands. Our results reveal the complex responses of wetland carbon flux to environmental and ecological variables with warming and provide new insights into improving wetland models by incorporating additional variables and accounting for the changing roles of variables in carbon flux under warming.
54 ENVIRONMENTAL SCIENCES↗
Anomaly Detection and Identification Using a Leave-One-Variable-Out Method
At nuclear power plants (NPPs), anomaly detection and identification (i.e., determining the causes of anomalies) are important tasks for ensuring the safe and efficient operation of NPPs. These tasks are currently labor-intensive and costly, and are made more difficult by the size and complexity of NPP systems. An alternative approach to conducting these tasks is to automate them, such as via the reconstruction-based contribution method, which is a well-researched unsupervised machine learning method that uses a data-driven model of anomaly-free behavior to detect events and then identify each variable’s contributions to those events. The present effort developed a novel contribution approach that utilized a leave-one-variable-out (LOVO) model, with which each variable is predicted using all the other variables. The novelty lay in transforming this model into a reconstruction model and modifying the identification algorithm to work with the new reconstruction model. To evaluate this method in a controlled environment, a synthetic dataset based on spring-mass-damper (SMD) systems (commonly found in mechanical engineering references) was used, with known anomalies introduced into the system. The proposed method successfully detected the anomalies and afforded insights into their causes, thus enabling the appropriate identifications to be made.
46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
Sources of Uncertainty in the Prediction of LAI / fPAR from MODIS
To explicate the sources of uncertainty in the prediction of biophysical variables over space, consider the general equation: where z is a variable with values on some nominal, ordinal, interval or ratio scale; y is a vector of input variables; u is the spatial support of y and z ; x and u are the spatial locations of y and z , respectively; f is a model and B is the vector of the parameters of this model. Any y or z has a value and a spatial extent which is called its support. Viewed in this way, categories of uncertainty are from variable (e.g. measurement), parameter, positional. support and model (e.g. structural) sources. The prediction of Leaf Area Index (LAI) and the fraction of absorbed photosynthetically active radiation (fPAR) are examples of z variables predicted using model(s) as a function of y variables and spatially constant parameters. The MOD15 algorithm is an example of f, called f(sub 1), with parameters including those defined by one of six biome types and solar and view angles. The Leaf Canopy Model (LCM)2, a nested model that combines leaf radiative transfer with a full canopy reflectance model through the phase function, is a simpler though similar radiative transfer approach to f(sub 1). In a previous study, MOD15 and LCM2 gave similar results for the broadleaf forest biome. Differences between these two models can be used to consider the structural uncertainty in prediction results. In an effort to quantify each of the five sources of uncertainty and rank their relative importance for the LAI/fPAR prediction problem, we used recent data for an EOS Core Validation Site in the broadleaf biome with coincident surface reflectance, vegetation index, fPAR and LAI products from the Moderate Resolution Imaging Spectrometer (MODIS). Uncertainty due to support on the input reflectance variable was characterized using Landsat ETM+ data. Input uncertainties were propagated through the LCM2 model and compared with published uncertainties from the MOD15 algorithm.
Dungan, Jennifer L.↗
A Deep Learning-Based Workflow for Fast Prediction of 3D State Variables in Geological Carbon Storage: A Dimension Reduction Approach
In this study, we used deep learning techniques, which are a form of artificial intelligence, to create fast and effective models for predicting how fluids flow in underground geological formations. This is important for managing geological carbon storage, a method used to fight climate change by storing carbon dioxide underground. The challenge lies in the complex nature of these underground spaces and the large amount of data needed to accurately simulate them. To overcome these issues, we developed a new workflow that reduces the data’s complexity before training the deep learning model and then reconstructs the predicted results in their original form. We also proposed a unique approach to handle the specific complexities found in 3D saturation fields, a crucial aspect of fluid flow prediction. We tested our method using real-world data from the Gulf of Mexico. Our results show that our approach not only accurately predicts fluid behavior but also significantly reduces computation time. This will greatly improve real-time decision-making and risk assessment in large-scale geological carbon storage operations.
Wang, Hongsheng↗
Differentiable, Learnable, Regionalized Process-Based Models With Multiphysical Outputs can Approach State-Of-The-Art Hydrologic Prediction Accuracy
Predictions of hydrologic variables across the entire water cycle have significant value for water resources management as well as downstream applications such as ecosystem and water quality modeling. Recently, purely data-driven deep learning models like long short-term memory (LSTM) showed seemingly insurmountable performance in modeling rainfall runoff and other geoscientific variables, yet they cannot predict untrained physical variables and remain challenging to interpret. Here, we show that differentiable, learnable, process-based models (called δ models here) can approach the performance level of LSTM for the intensively observed variable (streamflow) with regionalized parameterization. We use a simple hydrologic model HBV as the backbone and use embedded neural networks, which can only be trained in a differentiable programming framework, to parameterize, enhance, or replace the process-based model's modules. Without using an ensemble or post-processor, δ models can obtain a median Nash-Sutcliffe efficiency of 0.732 for 671 basins across the USA for the Daymet forcing data set, compared to 0.748 from a state-of-the-art LSTM model with the same setup. For another forcing data set, the difference is even smaller: 0.715 versus 0.722. Meanwhile, the resulting learnable process-based models can output a full set of untrained variables, for example, soil and groundwater storage, snowpack, evapotranspiration, and baseflow, and can later be constrained by their observations. Both simulated evapotranspiration and fraction of discharge from baseflow agreed decently with alternative estimates. The general framework can work with models with various process complexity and opens up the path for learning physics from big data.
54 ENVIRONMENTAL SCIENCES↗
Recent developments in analysis of crack propagation and fracture of practical materials
The limitations of linear elastic fracture mechanics in aircraft design and in the study of fatigue crack propagation in aircraft structures are discussed. NASA-Langley research to extend the capabilities of fracture mechanics to predict the maximum load that can be carried by a cracked part and to deal with aircraft design problems are reported. Achievements include: (1) improved stress intensity solutions for laboratory specimens; (2) fracture criterion for practical materials; (3) crack propagation predictions that account for mean stress and high maximum stress effects; (4) crack propagation predictions for variable amplitude loading; and (5) the prediction of crack growth and residual stress in built-up structural assemblies. These capabilities are incorporated into a first generation computerized analysis that allows for damage tolerance and tradeoffs with other disciplines to produce efficient designs that meet current airworthiness requirements.
Hardrath, H. F.↗
GPS Ocean Surface Reflection Technology Requirements for Space Applications
Two key NASA OES research objectives are the Seasonal to Interannual Climate Variability and Prediction and Long-Term Climate/Natural Variability and Change investigations.
GPS↗
Examination of the effect of propulsion system performance variables on the life prediction for the SSME Lox Post
The main objective is to calculate damage evolution at the critical location of selected components of reusable space propulsion systems as the function of their operating environments and to relate the evolution of damage at the critical locations to the performance variables of the propulsion system using advanced constitutive and damage models. The models so developed will be oriented toward use in an advanced diagnostic/prognostic health monitoring system for reusable propulsion systems. The components to be modelled include a Lox Post (injector element), a thrust chamber, and a turbine blade. The study conducted on the Lox Post is reported.
Russell, D. A.↗
Interannual Variability, Global Teleconnection, and Potential Predictability Associated with the Asian Summer Monsoon
In this Chapter, aspects of global teleconnections associated with the interannual variability of the Asian summer monsoon (ASM) are discussed. The basic differences in the basic dynamics of the South Asian Monsoon and the East Asian monsoon, and their implications on global linkages are discussed. Two teleconnection modes linking ASM variability to summertime precipitation over the continental North America were identified. These modes link regional circulation and precipitation anomalies over East Asia and continental North America, via coupled atmosphere-ocean variations over the North Pacific. The first mode has a large zonally symmetrical component and appears to be associated with subtropical jetstream variability and the second mode with Rossby wave dispersion. Both modes possess strong sea surface temperature (SST) expressions in the North Pacific. Results show that the two teleconnection modes may have its origin in intrinsic modes of sea surface temperature variability in the extratropical oceans, which are forced in part by atmospheric variability and in part by air-sea interaction. The potential predictability of the ASM associated with SST variability in different ocean basins is explored using a new canonical ensemble correlation prediction scheme. It is found that SST anomalies in tropical Pacific, i.e., El Nino, is the most dominant forcing for the ASM, especially over the maritime continent and eastern Australia. SST anomalies in the India Ocean may trump the influence from El Nino in western Australia and western maritime continent. Both El Nino, and North Pacific SSTs contribute to monsoon precipitation anomalies over Japan, southern Korea, northern and central China. By optimizing SST variability signals from the world ocean basins using CEC, the overall predictability of ASM can be substantially improved.
Lau, K. M.↗
A Phenomenon-Based Decomposition of Model-Based Estimates of Boreal Winter ENSO Variability
Climate models are now routinely being used to simulate and predict climate variability on time scales ranging from sub-seasonal to seasonal and longer. As such, there are now long histories of such simulations and predictions spanning multiple decades and multiple ensemble members, both of which are crucial for separating climate signal from climate noise. A key focus of such runs has been the El Niño-Southern Oscillation (ENSO), spurred by recent improvements in our ability to predict such events, though questions remain as to how well climate models do beyond simply always predicting the “canonical” atmospheric response to an ENSO event—something simple statistical models already do reasonably well. This is a critical issue that needs addressing, given the importance of event-to-event differences for predicting regional impacts of ENSO teleconnections, and the need to justify the expense of running sophisticated climate models. Unfortunately, current diagnostic tools are not well suited for quantifying the different sources of variability associated with specific phenomena such as ENSO. More generally, while much effort has focused on addressing model bias, less has been done to address errors in second moment statistics—an issue whose importance is gaining increased attention particularly as we build climate prediction systems capable of taking advantage of forecasts of opportunity—a capability that requires reliable estimates of forecast uncertainty. In this report, we outline a phenomenon-based statistical decomposition of climate variance(in essence a detailed variance budget)that is specifically tailored to address the above questions by separating the variability (both the signal and noise) into that tied to the long-term average impact of a particular phenomenon(the composite mean) and the event-to-event(E2E) variability about the composite mean. In addition, we provide related decompositions of the correlations that allow us to quantify how much of the agreement with observations (the skill) comes from the composite mean as opposed to from the E2Evariability. As an example, we present the results of such a decomposition for ENSO based on simulations with the GEOS atmospheric general circulation model (AGCM), with a focus on the monthly mean impacts over North America during boreal winter(December –March). Here we take advantage of existing GEOS AGCM simulations that were produced as companion simulations to MERRA-2for the period 1980-2016. Comparisons are made throughout with MERRA-2.
ENSO↗