SEARCH · Engineering Papers

Results for “decadal 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.

SourceResearch connectionFrom yearThrough year

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

At least 199 records · Page 11

All sources NASA NTRS DOE OSTI

DOE OSTI2020

Advances in Structure Prediction of Lanthanides and Actinides with Genetic Algorithms

Computationally guided materials discovery has advanced significantly in the past two decades. Recent developments of widely available genetic algorithm-based structure prediction software in the past 10+ years has led to increased use of this methodology in the discovery and design of new materials. Owing to the computational challenges of treating f-block elements and computationally expensive methods for structure prediction, the discovery of new phases of lanthanide and actinide materials has been particularly challenging. With advancements in the development of genetic algorithms for structure prediction and the continuing improvement of high-performance computing resources, genetic algorithms provide a tractable method to find new stable phases of materials at high pressure or to search a large phase space for unknown compounds. Genetic algorithms have already been applied to the successful identification of novel lanthanide and actinide materials, as well as the prediction of unusual material properties in new phases that make promising synthetic targets.

Shields, Ashley↗

DOE OSTI2025

Nanometer-thick ultraflat cantilever resonators

Nanomechanical devices made from ultrathin materials are transforming diverse fields, including sensing, signal processing, and quantum technologies. However, as these materials become thinner, their low bending rigidity poses significant fabrication challenges, and achieving nanometer-thick flat cantilevers with consistent and predictable mechanical responses has remained elusive despite decades of research. Here we present nanometer-thick, ultraflat cantilever resonators fabricated using atomic layer deposition. By effectively mitigating the effects of uncontrollable built-in strain and geometric disorder, the ultraflat nanocantilevers exhibit resonance frequencies closely aligned with thin-plate theory predictions and display low sample-to-sample variability. These cantilevers maintain mechanical stability in both vacuum and air environments, even at large length-to-thickness ratios of up to 3000. The ultraflat nanocantilevers are approaching the thickness limit, beyond which thermal fluctuations at room temperature can spontaneously induce random ripples in otherwise flat films.

nanofabrication↗

DOE OSTI2026

Prime Time for Model-Predictive Control? Assessing the Technical and Market Readiness of Advanced Controls in Buildings

Despite three decades of extensive research and field testing that have consistently validated the benefits of Model Predictive Control (MPC) in building applications, the technology has seen limited market adoption. This paper evaluates the readiness of MPC for widespread deployment, showcases recent demonstrations and field tests across diverse building types, including residential, small commercial, large commercial, and campus settings. Our results demonstrate that MPC can optimize system operations to achieve load shifting, minimize curtailment of on-site generation, and reduce energy costs by up to 80 %, while maintaining or improving occupant comfort. We also show that MPC can effectively control large assets, such as MW-sized thermal storage systems, and respond to dynamic pricing signals. However, achieving scale remains difficult due to labor-intensive workflows, reliance on a “PhD-in-the-loop” for MPC design and maintenance, susceptibility to fragile data infrastructure, and persistent workforce education and acceptance barriers. To bridge this gap, we outline a transition from bespoke, labor intensive prototypes toward streamlined, segment-targeted deployment strategies that leverage model templates, semantic tools, and generative AI. By automating control configuration and reducing engineering effort, these recommendations provide a pathway for transforming successful research demonstrations into scalable, market ready solutions for MPC-based controls.

Pritoni, Marco↗

DOE OSTI2022

The Seasonal-to-Multiyear Large Ensemble (SMYLE) prediction system using the Community Earth System Model version 2

Abstract. The potential for multiyear prediction of impactful Earth system change remains relatively underexplored compared to shorter (subseasonal to seasonal) and longer (decadal) timescales. In this study, we introduce a new initialized prediction system using the Community Earth System Model version 2 (CESM2) that is specifically designed to probe potential and actual prediction skill at lead times ranging from 1 month out to 2 years. The Seasonal-to-Multiyear Large Ensemble (SMYLE) consists of a collection of 2-year-long hindcast simulations, with four initializations per year from 1970 to 2019 and an ensemble size of 20. A full suite of output is available for exploring near-term predictability of all Earth system components represented in CESM2. We show that SMYLE skill for El Niño–Southern Oscillation is competitive with other prominent seasonal prediction systems, with correlations exceeding 0.5 beyond a lead time of 12 months. A broad overview of prediction skill reveals varying degrees of potential for useful multiyear predictions of seasonal anomalies in the atmosphere, ocean, land, and sea ice. The SMYLE dataset, experimental design, model, initial conditions, and associated analysis tools are all publicly available, providing a foundation for research on multiyear prediction of environmental change by the wider community.

54 ENVIRONMENTAL SCIENCES↗

NASA NTRS2008

How Deep and Hot was Earth's Magma Ocean? Combined Experimental Datasets for the Metal-silicate Partitioning of 11 Siderophile Elements - Ni, Co, Mo, W, P, Mn, V, Cr, Ga, Cu and Pd

Since approximately 1990 high pressure and temperature (PT) experiments on metal-silicate systems have showed that partition coefficients (D) for siderophile (iron-loving) elements are much different than those measured at low PT conditions. The high PT data have been used to argue for a magma ocean during growth of the early Earth. Initial conclusions were based on experiments and calculations for a small number of elements such as Ni and Co. However, for many elements only a limited number of experimental data were available then, and they only hinted at values of metal-silicate D's at high PT conditions. In the ensuing decades there have been hundreds of new experiments carried out and published on a wide range of siderophile elements. At the same time several different models have been advanced to explain the siderophile elements in the earth's mantle: a) intermediate depth magma ocean; 25-30 GPa, b) deep magma ocean; up to 50 GPa, and c) early reduced and later oxidized magma ocean. Some studies have drawn conclusions based on a small subset of siderophile elements, or a set of elements that provides little leverage on the big picture (like slightly siderophile elements), and no single study has attempted to quantitatively explain more than 5 elements at a time. The purpose of this abstract is to update the predictive expressions outlined by Righter et al. (1997) with new experimental data from the last decade, test the predictive ability of these expressions against independent datasets (there are more data now to do this properly), and to apply the resulting expressions to the siderophile element patterns in Earth's upper mantle. The predictive expressions have the form: lnD = alnfO2 + b/T + cP/T + d(1Xs) + e(1Xc) + SigmafiXi + g These expressions are guided by the thermodynamics of simple metal-oxide equilibria that control each element, include terms that mimic the activity coefficients of each element in the metal and silicate, and quantify the effect of variable oxygen fugacity. Preliminary results confirm that D(Ni) and D(Co) converge at pressures near 25-30 GPa and approximately 2200 K, and show that D(Pd) and D(Cu) become too low at the PT conditions of the deepest models. Furthermore, models which force fit V and Cr mantle concentrations by metal-silicate equilibrium overlook the fact that at early Earth mantle fO2, these elements will be more compatible in Mg-perovskite and (Fe,Mg)O than in metal. Thus an intermediate depth magma ocean, at 25-30 GPa, 2200 K, and at IW-2, can explain more mantle siderophile element concentrations than other models.

Righter, Kevin↗

DOE OSTI2020

Advancing Fusion with Machine Learning Research Needs Workshop Report

Abstract Machine learning and artificial intelligence (ML/AI) methods have been used successfully in recent years to solve problems in many areas, including image recognition, unsupervised and supervised classification, game-playing, system identification and prediction, and autonomous vehicle control. Data-driven machine learning methods have also been applied to fusion energy research for over 2 decades, including significant advances in the areas of disruption prediction, surrogate model generation, and experimental planning. The advent of powerful and dedicated computers specialized for large-scale parallel computation, as well as advances in statistical inference algorithms, have greatly enhanced the capabilities of these computational approaches to extract scientific knowledge and bridge gaps between theoretical models and practical implementations. Large-scale commercial success of various ML/AI applications in recent years, including robotics, industrial processes, online image recognition, financial system prediction, and autonomous vehicles, have further demonstrated the potential for data-driven methods to produce dramatic transformations in many fields. These advances, along with the urgency of need to bridge key gaps in knowledge for design and operation of reactors such as ITER, have driven planned expansion of efforts in ML/AI within the US government and around the world. The Department of Energy (DOE) Office of Science programs in Fusion Energy Sciences (FES) and Advanced Scientific Computing Research (ASCR) have organized several activities to identify best strategies and approaches for applying ML/AI methods to fusion energy research. This paper describes the results of a joint FES/ASCR DOE-sponsored Research Needs Workshop on Advancing Fusion with Machine Learning, held April 30–May 2, 2019, in Gaithersburg, MD (full report available at https://science.osti.gov/-/media/fes/pdf/workshop-reports/FES_ASCR_Machine_Learning_Report.pdf ). The workshop drew on broad representation from both FES and ASCR scientific communities, and identified seven Priority Research Opportunities (PRO’s) with high potential for advancing fusion energy. In addition to the PRO topics themselves, the workshop identified research guidelines to maximize the effectiveness of ML/AI methods in fusion energy science, which include focusing on uncertainty quantification, methods for quantifying regions of validity of models and algorithms, and applying highly integrated teams of ML/AI mathematicians, computer scientists, and fusion energy scientists with domain expertise in the relevant areas.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2024

Quantitative Predictive Theories through Integrating Quantum, Statistical, Equilibrium, and Nonequilibrium Thermodynamics

Today's thermodynamics is largely based on the combined law for equilibrium systems and statistical mechanics derived by Gibbs in 1873 and 1901, respectively, while irreversible thermodynamics for nonequilibrium systems resides essentially on the Onsager Theorem as a separate branch of thermodynamics developed in 1930s. Between them, quantum mechanics was invented and was quantitatively solved in terms of density functional theory (DFT) in 1960s. Furthermore, these three scientific domains operate based on different principles and are very much separated from each other. In analogy to the parable of the blind men and the elephant articulated by Perdew, they individually represent different portions of a complex system and thus are incomplete by themselves alone, resulting in the lack of quantitative agreement between their predictions and experimental observations. Over the last two decades, the author's group has developed a multiscale entropy approach (recently termed as zentropy theory) that integrates DFT-based quantum mechanics and Gibbs statistical mechanics and is capable of accurately predicting entropy and free energy of complex systems. Furthermore, in combination with the combined law for nonequilibrium systems developed by Hillert, the author developed the theory of cross phenomena beyond the phenomenological Onsager Theorem. The zentropy theory and theory of cross phenomena jointly provide quantitative predictive theories for systems from electronic to any observable scales as reviewed in the present work.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

DOE OSTI2022

The economic value of photovoltaic performance loss mitigation in electricity spot markets

Photovoltaic (PV) performance is affected by reversible and irreversible losses. These can typically be mitigated through responsive and proactive operations and maintenance (O&M) activities. However, to generate profit, the cost of O&M must be lower than the value of the recovered electricity. This value depends both on the amount of recovered energy and on the electricity prices, which can vary significantly over time in spot markets. The present work investigates the impact of the electricity price variability on the PV profitability and on the related O&M activities in Italy, Portugal, and Spain. Here, it is found that the PV revenues varied by 1.6 × to 1.8 × within the investigated countries in the last 5 years. Moreover, forecasts predict higher average prices in the current decade compared to the previous one. These will increase the future PV revenues by up to 60% by 2030 compared to their 2015–2020 mean values. These higher revenues will make more funds available for better maintenance and for higher quality components, potentially leading to even higher energy yield and profits. Linearly growing or constant price assumptions cannot fully reproduce these expected price trends. Furthermore, significant price fluctuations can lead to unexpected scenarios and alter the predictions.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

NASA NTRS2016

Remote Sensing Global Surface Air Pressure Using Differential Absorption BArometric Radar (DiBAR)

Tropical storms and severe weathers are listed as one of core events that need improved observations and predictions in World Meteorological Organization and NASA Decadal Survey (DS) documents and have major impacts on public safety and national security. This effort tries to observe surface air pressure, especially over open seas, from space using a Differential-absorption BArometric Radar (DiBAR) operating at the 50-55 gigahertz O2 absorption band. Air pressure is among the most important variables that affect atmospheric dynamics, and currently can only be measured by limited in-situ observations over oceans. Analyses show that with the proposed space radar the errors in instantaneous (averaged) pressure estimates can be as low as approximately 4 millibars (approximately 1 millibar under all weather conditions). With these sea level pressure measurements, the forecasts of severe weathers such as hurricanes will be significantly improved. Since the development of the DiBAR concept about a decade ago, NASA Langley DiBAR research team has made substantial progress in advancing the concept. The feasibility assessment clearly shows the potential of sea surface barometry using existing radar technologies. The team has developed a DiBAR system design, fabricated a Prototype-DiBAR (P-DiBAR) for proof-of-concept, conducted lab, ground and airborne P-DiBAR tests. The flight test results are consistent with the instrumentation goals. Observational system simulation experiments for space DiBAR performance based on the existing DiBAR technology and capability show substantial improvements in tropical storm predictions, not only for the hurricane track and position but also for the hurricane intensity. DiBAR measurements will lead us to an unprecedented level of the prediction and knowledge on global extreme weather and climate conditions.

Lin, Bing↗

NASA NTRS2008

Stratospheric Semi-Decadal Oscillations in NCEP Data

An analysis of the National Centers for Environmental Prediction (NCEP)/National Center for Atmospheric Research (NCAR) data is presented to provide a more complete description of the stratospheric 5-year semi-decadal (SD) oscillation (Mayr et al., 2007). The zonal-mean temperature and zonal wind data from the Atmospheric Research R-1 analysis are employed, covering the years from 1962 to 2002 in the altitude range from 10 to 30km. For diagnostic purposes, the data are separated into the hemispherically symmetric and anti-symmetric components, and spectral analysis is applied to identify the signatures of the SD oscillations. Through the synthesis or filtering of spectral features, the SD modulations of the annual oscillation (AO) and quasi-biennial oscillation (QBO) are delineated. In agreement with the earlier findings, the magnitude of the SD oscillation is more pronounced when the 30-month QBO dominates during the years from 1975 to 1995. This is consistent with results from a numerical model, which shows that such a QBO generates the SD oscillation through interaction with the 12-month AO. In the zonal winds, the SD oscillation in the NCEP data is confined to equatorial latitudes, where it modulates the symmetric AO and QBO by about 5 m/s below 30 km. In the temperature data, the effect is also seen around the equator, but it is much larger at polar latitudes where the SD oscillation produces variations as large as 2 K. Our data analysis indicates that the SD oscillation is mainly hemispherically symmetric, and it appears to originate at equatorial latitudes where most of the energy resides.

Mayr, H. G.↗

NASA NTRS2013

Northern Eurasian Heat Waves and Droughts

This article reviews our understanding of the characteristics and causes of northern Eurasian summertime heat waves and droughts. Additional insights into the nature of temperature and precipitation variability in Eurasia on monthly to decadal time scales and into the causes and predictability of the most extreme events are gained from the latest generation of reanalyses and from supplemental simulations with the NASA GEOS-5 AGCM. Key new results are: 1) the identification of the important role of summertime stationary Rossby waves in the development of the leading patterns of monthly Eurasian surface temperature and precipitation variability (including the development of extreme events such as the 2010 Russian heat wave), 2) an assessment of the mean temperature and precipitation changes that have occurred over northern Eurasia in the last three decades and their connections to decadal variability and global trends in SST, and 3) the quantification (via a case study) of the predictability of the most extreme simulated heat wave/drought events, with some focus on the role of soil moisture in the development and maintenance of such events. A literature survey indicates a general consensus that the future holds an enhanced probability of heat waves across northern Eurasia, while there is less agreement regarding future drought, reflecting a greater uncertainty in soil moisture and precipitation projections. Substantial uncertainties remain in our understanding of heat waves and drought, including the nature of the interactions between the short-term atmospheric variability associated with such extremes and the longer-term variability and trends associated with soil moisture feedbacks, SST anomalies, and an overall warming world.

SST↗

DOE OSTI2022

BASIN-3D: A brokering framework to integrate diverse environmental data

Diverse observational and simulation datasets are needed to understand and predict complex ecosystem behavior over seasonal to decadal and century time-scales. Integration of these datasets poses a major barrier towards advancing environmental science, particularly due to differences in the structure and formats of data provided by various sources. Here, we describe BASIN-3D (Broker for Assimilation, Synthesis and Integration of eNvironmental Diverse, Distributed Datasets), a data integration framework designed to dynamically retrieve and transform heterogeneous data from different sources into a common format to provide an integrated view. BASIN-3D enables users to adopt a standardized approach for data retrieval and avoid customizations for the data type or source. We demonstrate the value of BASIN-3D with two use cases that require integration of data from regional to watershed spatial scales. The first application uses the BASIN-3D Python library to integrate time-series hydrological and meteorological data to provide standardized inputs to analytical and machine learning codes in order to predict the impacts of hydrological disturbances on large river corridors of the United States. The second application uses the BASIN-3D Django framework to integrate diverse time-series data in a mountainous watershed in East River, Colorado, United States to enable scientific researchers to explore and download data through an interactive web portal. Thus, BASIN-3D can be used to support data integration for both web-based tools, as well as data analytics using Python scripting and extensions like Jupyter notebooks. The framework is expected to be transferable to and useful for many other field and modeling studies.

Varadharajan, C↗

NASA NTRS1999

Interannual and Decadal Variability of Summer Rainfall over South America

Using the CPC (Climate Prediction Center) Merged Analysis of Precipitation product along with the Goddard Earth Observing System reanalysis and the Climate Analysis Center sea surface temperature (SST) data, we conduct a diagnostic study of the interannual and decadal scale variability of summer rainfall over South America. Results show three leading modes of rainfall variation identified with interannual, decadal, and long-term trend variability. Together, these modes explain more than half the total variance. The first mode is highly correlated with El Nino/southern oscillation (ENSO), showing severe drought over Northeast Brazil and copious rainfall over the Ecuador coast and the area of Uruguay-Southern Brazil in El Nino years. This pattern is attributed to the large scale zonal shift of the Walker circulation and local Hadley cell anomaly induced by positive (negative) SST anomaly over the eastern (western) equatorial Pacific. In El Nino years, two convective belts indicated by upper tropospheric velocity potential trough and mid-tropospheric rising motion, which are somewhat symmetric about the equator, extend toward the northeast and the southeast into the tropical North and South Atlantic respectively. Sandwiched between the ascent is a region of descending motion over Northeast Brazil. The southern branch of the anomalous Hadley cell is dynamically linked to the increase of rainfall over Uruguay-Southern Brazil. The regional response of anomalous circulation shows a stronger South American summer monsoon and an enhanced (weakened) subtropical high over the South Atlantic (South Pacific) Ocean. The decadal variation displays a meridional shift of the Intertropical Convergence Zone (ITCZ), which is tie to the anomalous cross-equatorial SST gradient over the Atlantic and the eastern Pacific. In conjunction with this mode is a large scale mass swing between the polar regions and midlatitudes in both hemispheres. Over the South Atlantic and the South Pacific, the changes of the strength of the subtropical high and the associated surface wind are dynamically consistent with the distribution of local SST anomalies, suggesting the importance of the atmospheric forcing in the decadal time scale. The decadal mode also presents a weak summer monsoon in its positive phase, which reduces the moisture supply from the equatorial Atlantic and the Amazon Basin and results in negative rainfall anomalies over the central Andes and Gran Chaco. The long-term trend shows decrease of rainfall from the northwest coast to the southeast subtropical region and a southward shift of Atlantic ITCZ that leads to increased rainfall over northern and eastern Brazil. Our result shows a close link of this mode to the observed SST warming trend over the subtropical South Atlantic and a remote connection to the interdecadal SST variation over the extratropical North Atlantic found in previous studies.

Zhou, Jiayu↗

DOE OSTI2025

Convolutional Neural Networks Trained on Internal Variability Predict Forced Response of TOA Radiation by Learning the Pattern Effect

Abstract Predicting forced, long‐term radiative feedbacks from internal climate variability has been a decades‐long quest in climate science. We train a convolutional neural network (CNN) to predict annual‐ and global‐mean top of the atmosphere radiation anomalies from time‐varying maps of near‐surface temperature in climate models. Trained on internal variability alone, the nonlinear CNN can predict radiation under strong climate change, outperforms a regularized linear regression approach, and works within and across different climate models. We show with explainable artificial intelligence methods that the CNN draws predictive skill from physically meaningful regions but at much smaller spatial scales than currently assumed.

Rugenstein, Maria [Colorado State University Fort ↗

DOE OSTI2025

CONFLUX: A standardized framework to calculate reactor antineutrino flux

Nuclear fission reactors are abundant sources of antineutrinos for neutrino physics experiments. The flux and spectrum of antineutrinos emitted by a reactor can indicate its activity and composition, suggesting potential applications of neutrino measurements beyond fundamental scientific studies that may be valuable to society. The utility of reactor antineutrinos for applications and fundamental science is dependent on the availability of precise predictions of these emissions. For example, in the last decade, disagreements between reactor antineutrino measurements and models have inspired revision of reactor antineutrino calculations and standard nuclear databases as well as searches for new fundamental particles not predicted by the Standard Model of particle physics. Past predictions and descriptions of the methods used to generate them are documented to varying degrees in the literature, with different modeling teams incorporating a range of methods, input data, and assumptions. The resulting difficulty in accessing or reproducing past models and reconciling results from differing approaches complicates the future study and application of reactor antineutrinos. The CONFLUX (Calculation Of Neutrino FLUX) software framework is a neutrino prediction tool built with the goal of simplifying, standardizing, and democratizing the process of reactor antineutrino flux calculations. CONFLUX includes three primary methods for calculating the antineutrino emissions of nuclear reactors or individual beta decays that incorporate common nuclear data and beta decay theory. The software is prepackaged with the current nuclear databases, including ENDF.B/VIII, JEFF-3.3, and ENSDF, and it includes the capability to predict time-dependent reactor emissions, adjust nuclear database or beta decay inputs/assumptions, and propagate related sources of uncertainty. Here, this paper describes the CONFLUX software structure, details the methods used for flux and spectrum calculations, and provides examples of potential use cases.

Zhang, Xianyi [Lawrence Livermore National Laborat↗

DOE OSTI2021

Predicting Characteristics of the Water Cycle From Scaling Relationships

Over multi-decadal time scales, assuming that changes in subsurface water storage are negligible, the continental precipitative water flux, P, can be divided into two principal components, Q (run-off, including soil infiltration and groundwater recharge) and ET (evapotranspiration). Taking into account a broadly applied Budyko's phenomenology to describe the relationship of ET/P as a function of PET/P, where PET is the potential evapotranspiration, we propose a theoretical framework for predicting characteristics of the water cycle from scaling relationships. In this framework, the ecosystem net primary productivity is expressed in terms of soil formation and vegetation growth, which is mathematically optimized with respect to the water partitioning, generating directly the value ET/P. Here, the mathematical optimization is based on the general ecological principle that dominant ecosystems tend to be those that, for any given conditions, maximize conversion of atmospheric carbon to biomass. It is shown that application of the results of mathematical optimization to water-limited ecosystems is possible by applying the optimization only to a vegetation covered portion of the surface. For energy-limited ecosystems, the optimization can be applied only to a portion of the precipitation equal to PET, assuming that the remaining P simply runs off. We use theoretical and actual values of plant root fractal dimensionalities, d f , to predict ranges of ET/P as a function of PET/P for 0 ≤ PET/P ≤ 1 and compare with annual and multi-decadal means of ET/P. By comparing the developed approach with a large amount of data collected from the literature, we demonstrate its successful applications to both water- and energy-limited systems.

54 ENVIRONMENTAL SCIENCES↗

NASA NTRSDate not supplied

Perturbations, TTVs & the (Un)reliability of Ephemerides of Kepler Planets

Fractional uncertainties quoted for orbital periods of the vast majority of planet candidates listed in recent Kepler catalogs are $<10^{-5}$, with values $\sim 10^{-6}$ (corresponding to 2 minutes per 4 years) being typical. These small uncertainties suggest ephemeris predictions for most Kepler planets are robust for decades to come. Tabulated periods list formal uncertainties in the mean times between midpoints of successive transits during the time interval in which transits were observed. However, transit timing variations (TTVs) produce errors in estimates of some planets’ orbital periods that need to be accounted for in certain dynamical investigations and ephemeris predictions. Periodic TTVs with timescales short compared to the interval of Kepler observations largely average out and do not produce significant errors in estimates of orbital periods. TTVs with timescales comparable to the interval of Kepler observations have been fit for dozens of Kepler planet candidates to estimate long-term average orbital periods (Holczer et al. 2016, ApJS 225, 9), and more detailed dynamical models have been used to estimate long-term average periods of a small number of well-studied planets. Most Kepler planets that show large TTVs are near mean-motion orbital resonances with other planets. The largest effect for planets moderately close to two-body resonances are due to rotation of the forced eccentricity vector by resonant perturbations; the timescale for this precession is typically short, so the variations tend to average out. Libration of planets locked in resonances typically occurs on timescales longer than the Kepler baseline, but few Kepler planets appear to be resonantly-locked. The more general but smaller (during the era of Kepler observations) effect is caused by secular precession of the planets' free eccentricities usually has a period much longer than the baseline of the Kepler observations, so it is not accounted for in estimates of mean periods or uncertainties. This precession causes a discrepancy between Kepler era mean orbital period and long-term mean orbital period can exist even for planets having TTVs that are too small to be observable during the epoch of Kepler observations.

Kepler↗

DOE OSTI2024

Active and Transfer Learning of High-Dimensional Neural Network Potentials for Transition Metals

Classical molecular dynamics (MD) simulations represent a very popular and powerful tool for materials modeling and design. The predictive power of MD hinges on the ability of the interatomic potential to capture the underlying physics and chemistry. There have been decades of seminal work on developing interatomic potentials, albeit with a focus predominantly on capturing the properties of bulk materials. Such physics-based models, while extensively deployed for predicting the dynamics and properties of nanoscale systems over the past two decades, tend to perform poorly in predicting nanoscale potential energy surfaces (PESs) when compared to high-fidelity first-principles calculations. These limitations stem from the lack of flexibility in such models, which rely on a predefined functional form. Machine learning (ML) models and approaches have emerged as a viable alternative to capture the diverse size-dependent cluster geometries, nanoscale dynamics, and the complex nanoscale PESs, without sacrificing the bulk properties. Here, in this study, we introduce an ML workflow that combines transfer and active learning strategies to develop high-dimensional neural networks (NNs) for capturing the cluster and bulk properties for several different transition metals with applications in catalysis, microelectronics, and energy storage, to name a few. Our NN first learns the bulk PES from the high-quality physics-based models in literature and subsequently augments this learning via retraining with a higher-fidelity first-principles training data set to concurrently capture both the nanoscale and bulk PES. Our workflow departs from status-quo in its ability to learn from a sparsely sampled data set that nonetheless covers a diverse range of cluster configurations from near-equilibrium to highly nonequilibrium as well as learning strategies that iteratively improve the fingerprinting depending on model fidelity. All the developed models are rigorously tested against an extensive first-principles data set of energies and forces of cluster configurations as well as several properties of bulk configurations for 10 different transition metals. Our approach is material agnostic and provides a methodology to transfer and build upon the learnings from decades of seminal work in molecular simulations on to a new generation of ML-trained potentials to accelerate materials discovery and design.

36 MATERIALS SCIENCE↗