SEARCH · Engineering Papers

Results for “time series 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 433 records · Page 24

All sources NASA NTRS DOE OSTI

DOE OSTI2025

Estimating soybean yields from high-temporal-resolution multi-source data using deep learning

Accurate and timely crop yield prediction is crucial for ensuring food security and maintaining stable agricultural markets. In recent years, there has been a surge in interest in leveraging high-temporal-resolution, multi-source data for effective crop growth monitoring and yield estimation. A notable challenge arises from the difficulty in capturing the intricate interactions between variables across different time steps within these high-temporal-resolution time series datasets. This complexity hinders the reliable extraction of yield information from voluminous and often noisy datasets, especially during periods of extreme weather events. Here, in this study, we propose an Attention and Graph Isomorphism Network-enhanced Bi-directional Long Short-Term Memory network (AGB-LSTM) for estimating county-level soybean yield in the United States. This model integrates a diverse set of remote sensing data, including Near-Infrared Reflectance of Vegetation (NIRv), Sun-Induced chlorophyll Fluorescence (SIF), and Gross Primary Productivity (GPP), along with environmental covariates. The AGB-LSTM effectively leverages information related to crop yield from high-temporal-resolution time series data (5-days), achieving an accuracy of R²= 0.67 and rRMSE = 14.46%. This approach significantly outperforms traditional machine learning methods such as Random Forest (RF) (R²= 0.52, rRMSE = 17.36%) and Bi-LSTM (R²= 0.58, rRMSE = 16.17%). Sensitivity experiments with different time steps and ranges demonstrated that our model could accurately and stably predict yields 1 to 2 months before harvest. Moreover, data with a finer temporal resolution consistently improved prediction performance, resulting in an approximately 20% increase in and an approximately 20% decrease in rRMSE compared to using monthly composites. We also evaluated the robustness of the model under extreme climate events and observed strong performance (R²= 0.50, rRMSE = 21.32%). Finally, yield mapping for major soybean-producing regions in North America in 2023 revealed spatial patterns that closely matched USDA yield reports. Our findings suggest that the AGB-LSTM model is a promising and effective method for estimating yield and has notable potential for global crop yield forecasting.

Deep learning↗

NASA NTRS2011

Modeling Humans as Reinforcement Learners: How to Predict Human Behavior in Multi-Stage Games

This paper introduces a novel framework for modeling interacting humans in a multi-stage game environment by combining concepts from game theory and reinforcement learning. The proposed model has the following desirable characteristics: (1) Bounded rational players, (2) strategic (i.e., players account for one anothers reward functions), and (3) is computationally feasible even on moderately large real-world systems. To do this we extend level-K reasoning to policy space to, for the first time, be able to handle multiple time steps. This allows us to decompose the problem into a series of smaller ones where we can apply standard reinforcement learning algorithms. We investigate these ideas in a cyber-battle scenario over a smart power grid and discuss the relationship between the behavior predicted by our model and what one might expect of real human defenders and attackers.

Lee, Ritchie↗

NASA NTRS2010

Extracting Independent Local Oscillatory Geophysical Signals by Geodetic Tropospheric Delay

Zenith Tropospheric Delay (ZTD) due to water vapor derived from space geodetic techniques and numerical weather prediction simulated-reanalysis data exhibits non-linear and non-stationary properties akin to those in the crucial geophysical signals of interest to the research community. These time series, once decomposed into additive (and stochastic) components, have information about the long term global change (the trend) and other interpretable (quasi-) periodic components such as seasonal cycles and noise. Such stochastic component(s) could be a function that exhibits at most one extremum within a data span or a monotonic function within a certain temporal span. In this contribution, we examine the use of the combined Ensemble Empirical Mode Decomposition (EEMD) and Independent Component Analysis (ICA): the EEMD-ICA algorithm to extract the independent local oscillatory stochastic components in the tropospheric delay derived from the European Centre for Medium-Range Weather Forecasts (ECMWF) over six geodetic sites (HartRAO, Hobart26, Wettzell, Gilcreek, Westford, and Tsukub32). The proposed methodology allows independent geophysical processes to be extracted and assessed. Analysis of the quality index of the Independent Components (ICs) derived for each cluster of local oscillatory components (also called the Intrinsic Mode Functions (IMFs)) for all the geodetic stations considered in the study demonstrate that they are strongly site dependent. Such strong dependency seems to suggest that the localized geophysical signals embedded in the ZTD over the geodetic sites are not correlated. Further, from the viewpoint of non-linear dynamical systems, four geophysical signals the Quasi-Biennial Oscillation (QBO) index derived from the NCEP/NCAR reanalysis, the Southern Oscillation Index (SOI) anomaly from NCEP, the SIDC monthly Sun Spot Number (SSN), and the Length of Day (LoD) are linked to the extracted signal components from ZTD. Results from the synchronization analysis show that ZTD and the geophysical signals exhibit (albeit subtle) site dependent phase synchronization index.

Botai, O. J.↗

NASA NTRS2018

On the Limitations of Variational Bias Correction

Satellite radiances are the largest dataset assimilated into Numerical Weather Prediction (NWP) models, however the data are subject to errors and uncertainties that need to be accounted for before assimilating into the NWP models. Variational bias correction uses the time series of observation minus background to estimate the observations bias. This technique does not distinguish between the background error, forward operator error, and observations error so that all these errors are summed up together and counted as observation error. We identify some sources of observations errors (e.g., antenna emissivity, non-linearity in the calibration, and antenna pattern) and show the limitations of variational bias corrections on estimating these errors.

Moradi, Isaac↗

NASA NTRS2017

Methodology for Time-Domain Estimation of Storm Time Geoelectric Fields Using the 3-D Magnetotelluric Response Tensors

Geoelectric fields at the Earth's surface caused by magnetic storms constitute a hazard to the operation of electric power grids and related infrastructure. The ability to estimate these geoelectric fields in close to real time and provide local predictions would better equip the industry to mitigate negative impacts on their operations. Here we report progress toward this goal: development of robust algorithms that convolve a magnetic storm time series with a frequency domain impedance for a realistic three-dimensional (3-D) Earth, to estimate the local, storm time geoelectric field. Both frequency domain and time domain approaches are presented and validated against storm time geoelectric field data measured in Japan. The methods are then compared in the context of a real-time application.

Kelbert, Anna↗

DOE OSTI2022

Predictive Direct DC-Link Control for 7.2 kV Three-Port Low-Inertia Solid-State Transformer With Active Power Decoupling

Promising for applications including renewable energy and electric vehicle fast charging, a medium-voltage (MV) solid-state transformer (SST) typically has multiple modules series stacked, which requires the modules to be single phase. One critical issue of the single-phase SST is the double-line-frequency power ripple. Traditionally, large passives are used to buffer the ripple, resulting in significantly increased volume and cost. This article for the first time proposes active power decoupling (APD) for the SST and a predictive direct DC-link control method, using a current-source single-stage SST as an example. An electrolytic capacitor-less buffer port is used to absorb the ripple, which tolerates up to a 30% voltage ripple for small capacitance. The APD SST control is challenging. First, the current-source SST realizes isolation, DC link, and multiport in a single stage. Second, different from conventional SSTs or APD converters, the low-inertia DC link in this SST has a 40% switching ripple, which is difficult to stabilize. To address the control challenge, a direct DC-link control architecture and a predictive control method are proposed. Simulation and experimental results verify the proposed control method on an MV SiC modular soft-switching solid-state transformer (M-S4T) prototype. Here, the proposed APD reduces the SST volume by 53.8%. Importantly, the proposed concept is not limited to the M-S4T but is generic to other SSTs or APD converters.

14 SOLAR ENERGY↗

NASA NTRS1995

Prediction of coronal and heliospheric magnetic fields: The promise of SOI-MDI on SOHO

Models of the coronal magnetic field have been developed over the years that reproduce the static characteristics of coronal and heliospheric structures fairly well. Limitations of spatial and temporal resolution and nonuniform quality of the input data have made it particularly difficult to investigate the response of the corona to rapidly changing photospheric conditions. The Solar Oscillations Investigation (SOI) experiment on SOHO, scheduled for launch late in 1995, will produce a series of full-disk photospheric magnetic field observations with 4" resolution about every 2 hours for the next several years. Higher resolution observations of the center of the disk will be available several times per day. These data should provide a basis for predicting the coronal and heliospheric field and their changes with unprecedented accuracy during the rising phase of Solar Cycle 23.

Hoeksema, J. T.↗

NASA NTRS2023

NASA GRC ICME Schema for Materials Data Management: An Executive Summary

Integrated Computational Materials Engineering (ICME) has received a growing emphasis in attention due its potential impact on rapid material design, reduction in cost and time to market for new applications, and the promise of ‘fit-for-purpose’ materials coupled with recent advances in high performance computing and material characterization tools. However, for an organization to implement ICME practices for material discovery and design, a series of both technical and cultural challenges must be overcome to foster an environment that enables efficient, traceable, and predictive multiscale simulations of material behavior to enable virtual design of materials. In 2016, NASA sponsored a 2040 Vision study to define the potential 25-year future state required for integrated multiscale modeling of materials and systems to improve both the associated time and cost for aerospace and aeronautical innovation. The study envisions a cyber-physical-social ecosystem of experimentally validated computational models, tools, and techniques, along with the associated digital tapestry, that can enable rapid, optimized, ‘fit-for-purpose’ design of materials, components, and systems. A key requirement for such an ecosystem is the development of a robust information management system for materials across their full lifecycle, including material pedigree, experimental (real) and virtual (simulation) data, developed material models, and the implementation of models in engineering applications, such that process-structure-property-performance relationships can be established, thereby enabling the virtual design and optimization of materials. Such an information management system must be able to effectively capture: i) material information at each length scale; ii) test data and analysis; iii) associated material models; and iv) material and model deployment in engineering applications. These systems must also provide traceability between experimental and virtual representations of the material to ensure, when appropriate, the material digital twin is maintained. Additionally, this robust material information management system must be able to seamlessly connect with both commercial and an organization’s in-house software tools, be they analysis tools, other material databases, product lifecycle management (PLM) or simulation data management (SDM) tools, etc., such that automation of the design and analysis of a material across multiple length scales is possible. In this paper, an executive summary of the NASA GRC ICME Schema for materials information management is presented. The database best practices and schema design philosophy specifically for ICME materials data management and an overview description of each element in the schema is given, along with its associated role in an ICME workflow. Additionally, auxiliary tools that interact with the database and provide judicious automation with regards to importing, exporting, and analyzing materials data are presented. Such tools are critical to an ICME ecosystem, not only for their role in enabling optimization, but also in relieving users of tedious manual tasks, thus helping to promote adoption and combat the cultural challenges organizations face in enabling ICME.

Materials↗

NASA NTRS1987

Sunspot umbral oscillations in Mg II k

Time series observations of the profile of the Mg II k line 2795.52 A have been obtained in five sunspots with the Ultraviolet Spectrometer and Polarimeter on the Solar Maximum Mission. The three sunspots with umbrae larger than the 3 x 3 arcsec pixel size show significant oscillations in integrated line intensity and line centroid, with frequencies in the range 5.29-7.55 mHz (periods of 132-190 s). The frequencies of significant peaks in average umbral power spectra agree well with the frequencies of the three lowest-frequency transmission peaks predicted by a model of resonant transmission of acoustic waves. If radiative delays are unimportant, and the line centroid can be interpreted straightforwardly as a Doppler shift, the measured velocity-intensity phase differences indicate the superposition of upward-propagating and downward-propagating waves in the umbral chromosphere; this is further evidence for the resonant transmission model. A single, quiet sun time series of k core profiles yields power spectra and a phase difference consistent with the existence of a chromospheric p-mode.

Gurman, Joseph B.↗

NASA NTRS2015

Estimation of Aerosol Optical Depth at Different Wavelengths by Multiple Regression Method

This study aims to investigate and establish a suitable model that can help to estimate aerosol optical depth (AOD) in order to monitor aerosol variations especially during non-retrieval time. The relationship between actual ground measurements (such as air pollution index, visibility, relative humidity, temperature, and pressure) and AOD obtained with a CIMEL sun photometer was determined through a series of statistical procedures to produce an AOD prediction model with reasonable accuracy. The AOD prediction model calibrated for each wavelength has a set of coefficients. The model was validated using a set of statistical tests. The validated model was then employed to calculate AOD at different wavelengths. The results show that the proposed model successfully predicted AOD at each studied wavelength ranging from 340 nm to 1020 nm. To illustrate the application of the model, the aerosol size determined using measure AOD data for Penang was compared with that determined using the model. This was done by examining the curvature in the ln [AOD]-ln [wavelength] plot. Consistency was obtained when it was concluded that Penang was dominated by fine mode aerosol in 2012 and 2013 using both measured and predicted AOD data. These results indicate that the proposed AOD prediction model using routine measurements as input is a promising tool for the regular monitoring of aerosol variation during non-retrieval time.

Tan, Fuyi↗

DOE OSTI2020

Cryogenic Stress-Driven Grain Growth Observed via Microcompression with in situ Electron Backscatter Diffraction

The deformation of materials at cryogenic temperature is of interest for space, arctic, and fundamental science applications. In this work, a custom-built cooling system attached to a commercial picoindenter was used for in situ cryogenic microcompression testing of equal-channel angular-pressed copper with real-time electron backscatter diffraction. Stress-driven grain growth at cryogenic temperatures was observed during a series of elastic and plastic deformations. Finally, these results provide direct evidence for the previously predicted phenomenon, whereas previous ex situ examinations demonstrated coarsening after cryogenic loading when samples were not maintained at cryogenic temperatures between deformation and characterization.

36 MATERIALS SCIENCE↗

NASA NTRS1987

Performance prediction of concurrent systems

Concurrent systems are computers that use multiple processors to solve a single problem. A means to predict the application performance on these systems is a useful tool in many areas of concurrent system research. A computationally efficient and accurate method to predict performance for a class of parallel computations on concurrent systems is described. A parallel computation is modeled as a task system with precedence relationships expressed as a series parallel directed acyclic graph. Resources in concurrent systems are modeled as service centers in queueing network models. Using these two models as inputs, the method outputs predictions of both the time to complete the computation and the concurrent system utilization. The algorithm used is based on the approximate Mean Value Analysis in queueing network modeling with extensions to model concurrency in the computation. The new algorithm was validated against both detailed simulation and actual execution on a commercial multiprocessor.

Mak, Victor W. K.↗

NASA NTRS1992

Modeling orbital changes on tectonic time scales

Geologic time series indicate significant 100 ka and 400 ka pre-Pleistocene climate fluctuations, prior to the time of such fluctuations in Pleistocene ice sheets. The origin of these fluctuations must therefore depend on phenomena other than the ice sheets. In a previous set of experiments, we tested the sensitivity of an energy balance model to orbital insolation forcing, specifically focusing on the filtering effect of the Earth's geography. We found that in equatorial areas, the twice-yearly passage of the sun across the equator interacts with the precession index to generate 100 ka and 400 ka power in our modeled time series. The effect is proportional to the magnitude of land in equatorial regions. We suggest that such changes may reflect monsoonal variations in the real climate system, and the subsequent wind and weathering changes may transfer some of this signal to the marine record. A comparison with observed fluctuations of Triassic lake levels is quite favorable. A number of problems remain to be studied or clarified: (1) the EBM experiments need to be followed up by a limited number of GCM experiments; (2) the sensitivity to secular changes in orbital forcing needs to be examined; (3) the possible modifying role of sedimentary processes on geologic time series warrants considerably more study; (4) the effect of tectonic changes on Earth's rotation rate needs to be studied; and (5) astronomers need to make explicit which of their predictions are robust and geologists and astronomers have to agree on which of the predictions are most testable in the geologic record.

Crowley, Thomas J.↗

NASA NTRS1999

Fluid Physics in a Fluctuating Acceleration Environment

Our program of research aims at developing a stochastic description of the residual acceleration field onboard spacecraft (g-jitter) to describe in quantitative detail its effect on fluid motion. Our main premise is that such a statistical description is necessary in those cases in which the characteristic time scales of the process under investigation are long compared with the correlation time of g-jitter. Although a clear separation between time scales makes this approach feasible, there remain several difficulties of practical nature: (i), g-jitter time series are not statistically stationary but rather show definite dependences on factors such as active or rest crew periods; (ii), it is very difficult to extract reliably the low frequency range of the power spectrum of the acceleration field. This range controls the magnitude of diffusive processes; and (iii), models used to date are Gaussian, but there is evidence that large amplitude disturbances occur much more frequently than a Gaussian distribution would predict. The lack of stationarity does not constitute a severe limitation in practice, since the intensity of the stochastic components changes very slowly during space missions (perhaps over times of the order of hours). A separate analysis of large amplitude disturbances has not been undertaken yet, but it does not seem difficult a priori to devise models that may describe this range better than a Gaussian distribution. The effect of low frequency components, on the other hand, is more difficult to ascertain, partly due to the difficulty associated with measuring them, and partly because they may be indistinguishable from slowly changing averages. This latter effect is further complicated by the lack of statistical stationarity of the time series. Recent work has focused on the effect of stochastic modulation on the onset of oscillatory instabilities as an example of resonant interaction between the driving acceleration and normal modes of the system, and on cavity flow as an example of how an oscillatory response under periodic driving becomes diffusive if the forcing is random instead. This paper describes three different topics that illustrate behavior that is peculiar to a stochastic acceleration field. In the first case, we show that g-jitter can induce effective attractive or repulsive forces between a pair of spherical particles that are suspended in an incompressible fluid of different density provided that the momentum diffusion length is larger than the interparticle separation (as in the case in most colloidal suspensions). Second, a stochastic modulation of the control parameter in the vicinity of a pitchfork or supercritical bifurcation is known not to affect the location of the threshold. We show, however, that resonance between the modulation and linearly stable modes close to onset can lead to a shift in threshold. Finally, we discuss the classical problem of vorticity diffusion away from a plane boundary that is being vibrated along its own plane. Periodic motion with zero average vorticity production results in an exponential decay of the vorticity away from the boundary. Random vibration, on the other hand, results in power law decay away from the boundary even if vorticity production averages to zero.

Drolet, Francois↗

NASA NTRS1992

Measurements of nitrogen productivity in the equatorial Pacific

During the R/V Wecoma WEC88 cruise that sampled a meridional transect along 150 deg W from 15 deg N to 15 deg S, uptake of nitrate and ammonium by phytoplankton was measured using the stable isotope N-15 with simulated in-situ bottle incubations and shipboard mass spectrometry. A set of 25 daily productivity stations showed the influence of equatorial upwelling on nitrate distribution and N-15 uptake in a band from 6 deg N to 7.5 deg S compared with the oligotrophic waters to the north and south, with the highest values of nitrate uptake occurring at the equator. During a 5-day time series at the equator, there was an increase in nitrate accompanied by increased nitrate uptake. Interestingly, nitrate uptake rates (equivalent to new production rates) at the equator were lower than those predicted by previous investigators. Holdover experiments and uptake versus irradiance curves showed that the phytoplankton was in an early stage of metabolic adaptation and that can be a contributing factor.

Wilkerson, Frances P.↗

NASA NTRS2016

Satellite Sea Ice Measurements in the Arctic Ocean

Satellite observations sustained since 1979 have been the primary source of information to reveal Arctic Ocean sea ice extent is diminishing rapidly. Sea ice is also thinning based on historical surface and near-surface records and more recent satellite retrievals. There is a new “normal” environment in the Arctic environment with substantial socio-economic impacts. A key question is: What are the spatial and temporal characteristics of the sea ice thickness distribution throughout the annual cycle, and what is the evolving inter-annual trend? Arctic Ocean sea ice conditions have been determined with passive microwave radiometer, active scatterometer and other satellite instrument measurements recorded by a continuing series of satellites. An important feature of the continuous time series measurements has been the overlap of each new satellite dataset with ongoing measurements so as to provide adequate time intervals for calibration and validation. Satellite instrument diversity has provided both coarse spatial resolution measurements over the entire Arctic Ocean for long-period time series and limited-duration fine spatial resolution data over selected regions for navigation and other applications. Satellite observations of Arctic Ocean sea ice will continue to increase in importance because predictability of sea ice is poor and societal interest is great. Unfortunately, however the sustainability of some critical elements of the current Arctic Ocean satellite measurement suite beyond 2020 remains uncertain, e.g., after the CryoSat-2 and ICESat-2 missions have concluded.

Halpern, David↗

NASA NTRS2021

A Robust Hybrid Deep Learning Model for Spatiotemporal Image Fusion

Dense time-series remote sensing data with detailed spatial information are highly desired for the monitoring of dynamic earth systems. Due to the sensor tradeoff, most remote sensing systems cannot provide images with both high spatial and temporal resolutions. Spatiotemporal image fusion models provide a feasible solution to generate such a type of satellite imagery, yet existing fusion methods are limited in predicting rapid and/or transient phenological changes. Additionally, a systematic approach to assessing and understanding how varying levels of temporal phenological changes affect fusion results is lacking in spatiotemporal fusion research. The objective of this study is to develop an innovative hybrid deep learning model that can effectively and robustly fuse the satellite imagery of various spatial and temporal resolutions. The proposed model integrates two types of network models: super-resolution convolutional neural network (SRCNN) and long short-term memory (LSTM). SRCNN can enhance the coarse images by restoring degraded spatial details, while LSTM can learn and extract the temporal changing patterns from the time-series images. To systematically assess the effects of varying levels of phenological changes, we identify image phenological transition dates and design three temporal phenological change scenarios representing rapid, moderate, and minimal phenological changes. The hybrid deep learning model, alongside three benchmark fusion models, is assessed in different scenarios of phenological changes. Results indicate the hybrid deep learning model yields significantly better results when rapid or moderate phenological changes are present. It holds great potential in generating high-quality time-series datasets of both high spatial and temporal resolutions, which can further benefit terrestrial system dynamic studies. The innovative approach to understanding phenological changes’ effect will help us better comprehend the strengths and weaknesses of current and future fusion models.

spatiotemporal fusion↗

NASA NTRS2021

Diagnosing Seasonal to Multi-decadal Phytoplankton Group Dynamics in a Highly Productive Coastal Ecosystem

The Santa Barbara Channel, CA (SBC) is a biodiverse marine ecosystem fueled largely by phytoplankton productivity, and the composition of the phytoplankton community influences the magnitude and fates of this productivity. Here, we create a 22-year monthly time series of phytoplankton biomarker pigment concentrations in the SBC by combining 12 years of high performance liquid chromatography phytoplankton pigment concentrations with bio-optical models and 10 additional years of bio-optical observations. The bio-optical models skillfully predict biomarker pigment concentrations representative of five distinct phytoplankton groups (PGs; diatoms, dinoflagellates, chlorophytes, prymnesiophytes, and picophytoplankton) and resolve seasonal responses to the annual upwelling-relaxation cycle for all PGs except the dinoflagellates. Our observations indicate that nanophytoplankton groups respond most rapidly to seasonal upwelling, followed by diatoms, and then by picophytoplankton as the water column stratifies in the summer. A Regional Ocean Modeling System (ROMS) solution is used to relate advection of different source waters to the observed PG dynamics. The ROMS simulation results suggest that, on seasonal time scales, pronounced cross-SBC differences in PG seasonality are related to cross-SBC differences in source waters. El Niño Southern Oscillation events drive interannual variability in the upwelling response of most PGs. On decadal time scales, dinoflagellate blooms are associated with the warm phase of the North Pacific Gyre Oscillation and anomalous advection of Southern California Bight source waters into the SBC. Taken together, our results provide a novel view of phytoplankton community succession in response to seasonal upwelling by considering the dynamics of pico- and nano-phytoplankton and suggest that regional surface ocean advection plays a substantial role in driving phytoplankton composition in the SBC.

Phytoplankton↗