Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “ensemble 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.

At least 451 records · Page 25
All sourcesNASA NTRSDOE OSTI
NASA NTRS2004

Ensemble Data Mining Methods

Ensemble Data Mining Methods, also known as Committee Methods or Model Combiners, are machine learning methods that leverage the power of multiple models to achieve better prediction accuracy than any of the individual models could on their own. The basic goal when designing an ensemble is the same as when establishing a committee of people: each member of the committee should be as competent as possible, but the members should be complementary to one another. If the members are not complementary, Le., if they always agree, then the committee is unnecessary---any one member is sufficient. If the members are complementary, then when one or a few members make an error, the probability is high that the remaining members can correct this error. Research in ensemble methods has largely revolved around designing ensembles consisting of competent yet complementary models.

Oza, Nikunj C.↗
DOE OSTI2022

Using neural network ensembles to separate ocean biogeochemical and physical drivers of phytoplankton biogeography in Earth system models

Abstract. Earth system models (ESMs) are useful tools for predicting and understanding past and future aspects of the climate system. However, the biological and physical parameters used in ESMs can have wide variations in their estimates. Even small changes in these parameters can yield unexpected results without a clear explanation of how a particular outcome was reached. The standard method for estimating ESM sensitivity is to compare spatiotemporal distributions of variables from different runs of a single ESM. However, a potential pitfall of this method is that ESM output could match observational patterns because of compensating errors. For example, if a model predicts overly weak upwelling and low nutrient concentrations, it might compensate for this by allowing phytoplankton to have a high sensitivity to nutrients. Recently, we demonstrated that neural network ensembles (NNEs) are capable of extracting relationships between predictor and target variables within ocean biogeochemical models. Being able to view the relationships between variables, along with spatiotemporal distributions, allows for a more mechanistically based examination of ESM outputs. Here, we investigated whether we could apply NNEs to help us determine why different ESMs produce different spatiotemporal distributions of phytoplankton biomass. We tested this using three cases. The first and second case used different runs of the same ESM, except that the physical circulations differed between them in the first case, while the biological equations differed between them in the second. Our results indicated that the NNEs were capable of extracting the relationships between variables for different runs of a single ESM, allowing us to distinguish between differences due to changes in circulation (which do not change relationships) from changes in biogeochemical formulation (which do change relationships). In the third case, we applied NNEs to two different ESMs. The results of the third case highlighted the capability of NNEs to contrast the apparent relationships of different ESMs and some of the challenges it presents. Although applied specifically to the ocean components of an ESM, our study demonstrates that Earth system modelers can use NNEs to separate the contributions of different components of ESMs. Specifically, this allows modelers to compare the apparent relationships across different ESMs and observational datasets.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Skillful Multi‐Month Predictions of Ecosystem Stressors in the Surface and Subsurface Ocean

Abstract Anthropogenic carbon emissions and associated climate change are driving rapid warming, acidification, and deoxygenation in the ocean, which increasingly stress marine ecosystems. On top of long‐term trends, short term variability of marine stressors can have major implications for marine ecosystems and their management. As such, there is a growing need for predictions of marine ecosystem stressors on monthly, seasonal, and multi‐month timescales. Previous studies have demonstrated the ability to make reliable predictions of the surface ocean physical and biogeochemical state months to years in advance, but few studies have investigated forecast skill of multiple stressors simultaneously or assessed the forecast skill below the surface. Here, we use the Community Earth System Model (CESM) Seasonal to Multiyear Large Ensemble (SMYLE) along with novel observation‐based biogeochemical and physical products to quantify the predictive skill of dissolved inorganic carbon (DIC), dissolved oxygen, and temperature in the surface and subsurface ocean. CESM SMYLE demonstrates high physical and biogeochemical predictive skill multiple months in advance in key oceanic regions and frequently outperforms persistence forecasts. We find up to 10 months of skillful forecasts, with particularly high skill in the Northeast Pacific (Gulf of Alaska and California Current Large Marine Ecosystems) for temperature, surface DIC, and subsurface oxygen. Our findings suggest that dynamical marine ecosystem prediction could support actionable advice for decision making.

Environmental Sciences & Ecology↗
DOE OSTI2021

Galaxy Zoo DECaLS: Detailed visual morphology measurements from volunteers and deep learning for 314,000 galaxies

We present Galaxy Zoo DECaLS: detailed visual morphological classifications for Dark Energy Camera Legacy Survey images of galaxies within the SDSS DR8 footprint. Deeper DECaLS images (r = 23.6 versus r = 22.2 from SDSS) reveal spiral arms, weak bars, and tidal features not previously visible in SDSS imaging. To best exploit the greater depth of DECaLS images, volunteers select from a new set of answers designed to improve our sensitivity to mergers and bars. Galaxy Zoo volunteers provide 7.5 million individual classifications over 314 000 galaxies. 140 000 galaxies receive at least 30 classifications, sufficient to accurately measure detailed morphology like bars, and the remainder receive approximately 5. All classifications are used to train an ensemble of Bayesian convolutional neural networks (a state-of-the-art deep learning method) to predict posteriors for the detailed morphology of all 314 000 galaxies. We use active learning to focus our volunteer effort on the galaxies which, if labelled, would be most informative for training our ensemble. When measured against confident volunteer classifications, the trained networks are approximately 99 per cent accurate on every question. Morphology is a fundamental feature of every galaxy; our human and machine classifications are an accurate and detailed resource for understanding how galaxies evolve.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Predicting oxidation damage of ultra high-temperature carbide ceramics in extreme environments using machine learning

Determining the oxidation resistance of UHTC carbides in extreme environments is challenging theoretically and experimentally due to the high dimensional complexity of influencing variables and intricate testing setups. Herein we demonstrate the use of machine learning (ML) models trained with experimental literature data to predict the oxide thickness of UHTC carbides exposed to air based on composition, mean grain size, relative densification, holding time, and temperature. A multi-dimensional database with 76 occurrences is created containing experimental results of Hf, Zr, and Ta carbides plus additives. In this study, the preprocessed database is then used to train ML models to predict their oxidation behavior. The trained model predicts the oxidation damage in the form of an average oxide thickness in UHTC carbides with a Mean Absolute Error (MAE) of ±65.45 μm for samples in the testing set that developed thicknesses up to 1000 μm. The model successfully predicted oxidation damage for a recession rate lower than 60 μm/min. It is noticed that the ensemble method MAE is increased to ±134.34 μm while forecasting the oxidation of samples with a recession rate higher than the threshold. The unprecedented approach is a novel way to predict the damage through the oxidation of carbide compounds before processing for a smarter design with room for improvement.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Data-Efficient Generation of Protein Conformational Ensembles with Backbone-to-Side-Chain Transformers

Excitement at the prospect of using data-driven generative models to sample configurational ensembles of biomolecular systems stems from the extraordinary success of these models on a diverse set of high-dimensional sampling tasks. Unlike image generation or even the closely related problem of protein structure prediction, there are currently no data sources with sufficient breadth to parametrize generative models for conformational ensembles. To enable discovery, a fundamentally different approach to building generative models is required: models should be able to propose rare, albeit physical, conformations that may not arise in even the largest data sets. Here, in this work, we introduce a modular strategy to generate conformations based on “backmapping” from a fixed protein backbone that (1) maintains conformational diversity of the side chains and (2) couples the side-chain fluctuations using global information about the protein conformation. Our model combines simple statistical models of side-chain conformations based on rotamer libraries with the now ubiquitous transformer architecture to sample with atomistic accuracy. Together, these ingredients provide a strategy for rapid data acquisition and hence a crucial ingredient for scalable physical simulation with generative neural networks.

36 MATERIALS SCIENCE↗
DOE OSTI2025

pnnl/SNAP

In this work, we detail two uncertainty quantification (UQ) methods that provide complementary information. Readout ensembling, by finetuning only the readout layers of an ensemble of foundation models, provides information about model uncertainty. Amending the final readout layer to predict upper and lower quantiles replaces point predictions with distributional predictions, which provide information about uncertainty within the underlying training data. We demonstrate our approach with the MACE-MP-0 model, applying UQ to both the foundation model and a series of finetuned models. The uncertainties produced by the ensemble and quantile methods are demonstrated to be distinct measures by which the quality of the NNP output can be judged.

Pope, Jenna (Bilbrey) [Pacific Northwest National ↗
NASA NTRS2019

Forecasting the Ambient Solar Wind with Numerical Models. I. On the Implementation of an Operational Framework

The ambient solar wind conditions in interplanetary space and in the near-Earth environment are determined by activity on the Sun. Steady solar wind streams modulate the propagation behavior of interplanetary coronal mass ejections and are themselves an important driver of recurrent geomagnetic storm activity. The knowledge of the ambient solar wind flows and fields is thus an essential component of successful space weather forecasting. Here, we present an implementation of an operational framework for operating, validating, and optimizing models of the ambient solar wind flow on the example of Carrington Rotation 2077. We reconstruct the global topology of the coronal magnetic field using the potential field source surface model (PFSS) and the Schatten current sheet model (SCS) and discuss three empirical relationships for specifying the solar wind conditions near the Sun, namely the Wang–Sheeley (WS) model, the distance from the coronal hole boundary model (DCHB), and the Wang-Sheeley–Arge (WSA) model. By adding uncertainty in the latitude about the sub-Earth point, we select an ensemble of initial conditions and map the solutions to Earth by the Heliospheric Upwind eXtrapolation (HUX) model. We assess the forecasting performance from a continuous variable validation and find that the WSA model most accurately predicts the solar wind speed time series (RMSE≈83 km/s). We note that the process of ensemble forecasting slightly improves the forecasting performance of all solar wind models investigated. We conclude that the implemented framework is well suited for studying the relationship between coronal magnetic fields and the properties of the ambient solar wind flow in the near-Earth environment.

Reiss, Martin A.↗
DOE OSTI2025

Thermal Weight Determination and Interstate Coupling in State-Averaged ADAPT-VQE

Characterizing electronic thermal states at low temperatures is an important but challenging task in quantum chemistry and condensed matter physics, making it a prime candidate for a useful application in quantum computing. One of the most successful methods for state preparation on quantum computers is the Adaptive, Problem-Tailored (ADAPT) Variational Quantum Eigensolver (VQE), which has recently been generalized to treat excited states within a state-averaged framework as well as Gibbs states. In this work, we introduce Helmholtz-Optimized Thermal (HOT) ADAPT-VQE, an ancilla-free strategy for preparing Gibbs states that directly minimizes the Helmholtz free energy by targeting the dominant eigenstates of the thermal ensemble. We demonstrate the usefulness of HOT-ADAPT-VQE by predicting the free energy of two model systems with strongly correlated ground states: (1) the Fe 2+ cation in a magnetic field and (2) a [Cu 2 O 7 ] 10– fragment of the Mott insulator La 2 CuO 4 . Our results demonstrate that HOT-ADAPT-VQE significantly improves upon Gibbs-state estimates from multistate variants of ADAPT-VQE, often with substantially shallower quantum circuits, making it a promising candidate for thermal-state calculations.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Advancing coupled fire-atmosphere research with HIGRAD/FIRETEC (w20_firetec) (CY 2021 Institutional Computing Annual Progress Report) [Slides]

LANL’s HIGRAD and FIRETEC models were used in a multi-scale simulation framework to investigate impacts of small-scale fuel heterogeneities on large-scale Pyro-Cb formation. Forest structural diversity is found to increase canopy fire spread, and thus risk of tree mortality in FIRETEC simulations. Structural diversity is associated with an increase in ladder fuels, which facilitate fire spread into the canopy. Low-intensity fire burning in marginal conditions is highly sensitive to small perturbation in atmospheric turbulence. These results highlight predictive limitations of individual fire simulations, and emphasize the need for ensemble simulations, especially for low-intensity, prescribed fire.

36 MATERIALS SCIENCE↗
NASA NTRS1990

Moist convection and the vertical structure and water abundance of Jupiter's atmosphere

The cumulative effects of an ensemble of moist convective plumes on a conditionally unstable atmosphere are predicted by a model of moist convection on Jupiter in which the heating/cooling and drying/moistening of the environment occur through (1) compensating subsidence, (2) detrainment of updraft air at cloud tops, and (3) the evaporation and melting of falling condensate. Parahydrogen is transported as a passive tracer. Pure moist convective, mixed moist-dry convective, and primarily dry convective regimes are possible, depending on the assumed deep-water abundance, efficiency of condensate evaporation, and initial temperature profile.

Del Genio, Anthony D.↗
NASA NTRS2004

High Speed Networking and Large-scale Simulation in Geodynamics

Large-scale numerical simulation has been one of the most important approaches for understanding global geodynamical processes. In this approach, peta-scale floating point operations (pflops) are often required to carry out a single physically-meaningful numerical experiment. For example, to model convective flow in the Earth's core and generation of the geomagnetic field (geodynamo), simulation for one magnetic free-decay time (approximately 15000 years) with a modest resolution of 150 in three spatial dimensions would require approximately 0.2 pflops. If such a numerical model is used to predict geomagnetic secular variation over decades and longer, with e.g. an ensemble Kalman filter assimilation approach, approximately 30 (and perhaps more) independent simulations of similar scales would be needed for one data assimilation analysis. Obviously, such a simulation would require an enormous computing resource that exceeds the capacity of a single facility currently available at our disposal. One solution is to utilize a very fast network (e.g. 10Gb optical networks) and available middleware (e.g. Globus Toolkit) to allocate available but often heterogeneous resources for such large-scale computing efforts. At NASA GSFC, we are experimenting with such an approach by networking several clusters for geomagnetic data assimilation research. We shall present our initial testing results in the meeting.

Kuang, Weijia↗
NASA NTRS2000

Predictability of the 1997 and 1998 South Asian Summer Monsoons

The predictability of the 1997 and 1998 south Asian summer monsoon winds is examined from an ensemble of 10 Atmospheric General Circulation Model (AGCM) simulations with prescribed sea surface temperatures (SSTs) and soil moisture, The simulations are started in September 1996 so that they have lost all memory of the atmospheric initial conditions for the periods of interest. The model simulations show that the 1998 monsoon is considerably more predictable than the 1997 monsoon. During May and June of 1998 the predictability of the low-level wind anomalies is largely associated with a local response to anomalously warm Indian Ocean SSTs. Predictability increases late in the season (July and August) as a result of the strengthening of the anomalous Walker circulation and the associated development of easterly low level wind anomalies that extend westward across India and the Arabian Sea. During these months the model is also the most skillful with the observations showing a similar late-season westward extension of the easterly CD wind anomalies. The model shows little predictability or skill in the low level winds over southeast Asia during, 1997. Predictable wind anomalies do occur over the western Indian Ocean and Indonesia, however, over the Indian Ocean they are a response to SST anomalies that were wind driven and they show no skill. The reduced predictability in the low level winds during 1997 appears to be the result of a weaker (compared with 1998) simulated anomalous Walker circulation, while the reduced skill is associated with pronounced intraseasonal activity that is not well captured by the model. Remarkably, the model does produce an ensemble mean Madden-Julian Oscillation (MJO) response that is approximately in phase with (though weaker than) the observed MJ0 anomalies. This is consistent with the idea that SST coupling may play an important role in the MJO.

Schubert, Siegfred D.↗
NASA NTRS2017

Early Estimation of Solar Activity Cycle: Potential Capability and Limits

The variable solar magnetic activity known as the 11-year solar cycle has the longest history of solar observations. These cycles dramatically affect conditions in the heliosphere and the Earth's space environment. Our current understanding of the physical processes that make up global solar dynamics and the dynamo that generates the magnetic fields is sketchy, resulting in unrealistic descriptions in theoretical and numerical models of the solar cycles. The absence of long-term observations of solar interior dynamics and photospheric magnetic fields hinders development of accurate dynamo models and their calibration. In such situations, mathematical data assimilation methods provide an optimal approach for combining the available observational data and their uncertainties with theoretical models in order to estimate the state of the solar dynamo and predict future cycles. In this presentation, we will discuss the implementation and performance of an Ensemble Kalman Filter data assimilation method based on the Parker migratory dynamo model, complemented by the equation of magnetic helicity conservation and longterm sunspot data series. This approach has allowed us to reproduce the general properties of solar cycles and has already demonstrated a good predictive capability for the current cycle, 24. We will discuss further development of this approach, which includes a more sophisticated dynamo model, synoptic magnetogram data, and employs the DART Data Assimilation Research Testbed.

Kitiashvili, Irina N.↗
NASA NTRS1987

Forecasting forecast skill

It is shown that it is possible to predict the skill of numerical weather forecasts - a quantity which is variable from day to day and region to region. This has been accomplished using as predictor the dispersion (measured by the average correlation) between members of an ensemble of forecasts started from five different analyses. The analyses had been previously derived for satellite-data-impact studies and included, in the Northern Hemisphere, moderate perturbations associated with the use of different observing systems. When the Northern Hemisphere was used as a verification region, the prediction of skill was rather poor. This is due to the fact that such a large area usually contains regions with excellent forecasts as well as regions with poor forecasts, and does not allow for discrimination between them. However, when regional verifications were used, the ensemble forecast dispersion provided a very good prediction of the quality of the individual forecasts.

Kalnay, Eugenia↗
DOE OSTI2021

Analytical derivatives of the individual state energies in ensemble density functional theory. II. Implementation on graphical processing units (GPUs)

Conical intersections control excited state reactivity, and thus, elucidating and predicting their geometric and energetic characteristics are crucial for understanding photochemistry. Locating these intersections requires accurate and efficient electronic structure methods. Unfortunately, the most accurate methods (e.g., multireference perturbation theories such as XMS-CASPT2) are computationally challenging for large molecules. The state-interaction state-averaged restricted ensemble referenced Kohn–Sham (SI-SA-REKS) method is a computationally efficient alternative. The application of SI-SA-REKS to photochemistry was previously hampered by a lack of analytical nuclear gradients and nonadiabatic coupling matrix elements. We have recently derived analytical energy derivatives for the SI-SA-REKS method and implemented the method effectively on graphical processing units. We demonstrate that our implementation gives the correct conical intersection topography and energetics for several examples. Furthermore, our implementation of SI-SA-REKS is computationally efficient, with observed sub-quadratic scaling as a function of molecular size. This demonstrates the promise of SI-SA-REKS for excited state dynamics of large molecular systems.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Artificial intelligence “sees” split electrons

Chemical bonds between atoms are stabilized by the exchange-correlation (xc) energy, a quantum-mechanical effect in which “social distancing” by electrons lowers their electrostatic repulsion energy. Kohn-Sham density functional theory (DFT) states that the electron density determines this xc energy, but the density functional must be approximated. Furthermore, this is usually done by satisfying exact constraints of the exact functional (making the approximation predictive), by fitting to data (making it interpolative), or both. Two exact constraints—the ensemble-based piecewise linear variation of the total energy with respect to fractional electron number and fractional electron z-component of spin —require hard-to-control nonlocality. On page 1385 of this issue, Kirkpatrick et al. have taken a big step toward more accurate predictions for chemistry through the machine learning of molecular data plus the fractional charge and spin constraints, expressed as data that a machine can learn.

74 ATOMIC AND MOLECULAR PHYSICS↗
NASA NTRS2017

Using Data Assimilation Methods of Prediction of Solar Activity

The variable solar magnetic activity known as the 11-year solar cycle has the longest history of solar observations. These cycles dramatically affect conditions in the heliosphere and the Earth's space environment. Our current understanding of the physical processes that make up global solar dynamics and the dynamo that generates the magnetic fields is sketchy, resulting in unrealistic descriptions in theoretical and numerical models of the solar cycles. The absence of long-term observations of solar interior dynamics and photospheric magnetic fields hinders development of accurate dynamo models and their calibration. In such situations, mathematical data assimilation methods provide an optimal approach for combining the available observational data and their uncertainties with theoretical models in order to estimate the state of the solar dynamo and predict future cycles. In this presentation, we will discuss the implementation and performance of an Ensemble Kalman Filter data assimilation method based on the Parker migratory dynamo model, complemented by the equation of magnetic helicity conservation and long-term sunspot data series. This approach has allowed us to reproduce the general properties of solar cycles and has already demonstrated a good predictive capability for the current cycle, 24. We will discuss further development of this approach, which includes a more sophisticated dynamo model, synoptic magnetogram data, and employs the DART Data Assimilation Research Testbed.

Prediction↗