Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Advanced characterization”

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 847 records · Page 47
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Defect characterization and charge transport measurements in high-resolution Ni/n-4H-SiC Schottky barrier radiation detectors fabricated on 250 μ m epitaxial layers

Advances in the growth processes of 4H-SiC epitaxial layers have led to the continued expansion of epilayer thickness, allowing for the detection of more penetrative radioactive particles. In this work, we report the fabrication and characterization of high-resolution Schottky barrier radiation detectors on 250 μm thick n-type 4H-SiC epitaxial layers, the highest reported thickness to date. Several 8 × 8 mm 2 detectors were fabricated from a diced 100 mm diameter 4H-SiC epitaxial wafer grown on a conductive 4H-SiC substrate with a mean micropipe density of 0.11 cm -2 From the Mott–Schottky plots, the effective doping concentration was found to be in the range (0.95–1.85) × 10 14 cm -3 , implying that full depletion could be achieved at ~5.7 kV (0.5 MV/cm at the interface). The current-voltage characteristics demonstrated consistently low leakage current densities of 1–3 nA/cm 2 at a reverse bias of -800 V. This resulted in the pulse-height spectra generated using a 241Am alpha source (5486 keV) manifesting an energy resolution of less than 0.5% full width at half maximum (FWHM) for all the detectors at -200 V. The charge collection efficiencies (CCEs) were measured to be 98–99% with no discernable correlation to the energy resolution. A drift-diffusion model fit to the variation of CCE as a function of bias voltage, revealed a minority carrier diffusion length of ~10 μm. Deep level transient spectroscopy measurements on the best resolution detector revealed that the excellent performance was the result of having ultralow concentrations of the order of 1011 cm-3 lifetime limiting defects—Z 1/2 and EH 6/7 .

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2025

Evolution of dislocations during the rapid solidification in additive manufacturing

Materials processed by fusion-based additive manufacturing (AM) typically exhibit relatively high dislocation densities, along with cellular structures and elemental segregation. This representative structural feature significantly influences material performance; however, post-mortem microstructure characterizations of AM materials cannot capture the dynamic evolution of dislocations during the manufacturing process, thereby offering limited mechanism-based guidance for further advancing AM techniques and facilitating the qualification and certification of AM products. In this study, we conduct operando high-energy synchrotron X-ray diffraction experiments on wire-laser directed energy deposition of 316 L stainless steel. Through a unique configuration, our operando synchrotron experiments semi-quantitatively probe the dislocation density in solid phases and their dynamic changes during solidification and subsequent cooling. By integrating this advanced synchrotron technique with multi-physics simulation, in-situ neutron diffraction, and multi-scale electron microscopy characterization, our mechanistic study aims to elucidate the effects of rapid cooling and subsequent thermal cycling on the dislocation generation and evolution.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Status on Genetic Resistance to Rice Blast Disease in the Post-Genomic Era

Rice blast, caused by Magnaporthe oryzae, is a major threat to global rice production, necessitating the development of resistant cultivars through genetic improvement. Breakthroughs in rice genomics, including the complete genome sequencing of japonica and indica subspecies and the availability of various sequence-based molecular markers, have greatly advanced the genetic analysis of blast resistance. To date, approximately 122 blast-resistance genes have been identified, with 39 of these genes cloned and molecularly characterized. The application of these findings in marker-assisted selection (MAS) has significantly improved rice breeding, allowing for the efficient integration of multiple resistance genes into elite cultivars, enhancing both the durability and spectrum of resistance. Pangenomic studies, along with AI-driven tools like AlphaFold2, RoseTTAFold, and AlphaFold3, have further accelerated the identification and functional characterization of resistance genes, expediting the breeding process. Future rice blast disease management will depend on leveraging these advanced genomic and computational technologies. Emphasis should be placed on enhancing computational tools for the large-scale screening of resistance genes and utilizing gene editing technologies such as CRISPR-Cas9 for functional validation and targeted resistance enhancement and deployment. These approaches will be crucial for advancing rice blast resistance, ensuring food security, and promoting agricultural sustainability.

Pedrozo, Rodrigo↗
DOE OSTI2021

In situ synchrotron X-ray imaging and mechanical properties characterization of additively manufactured high-entropy alloy composites

Laser beam directed energy deposition has become an increasingly popular advanced manufacturing technique for materials discovery as a result of the in situ alloying capability. In this study, we leverage an additive manufacturing enabled high throughput materials discovery approach to explore the composition space of a graded W x (CoCrFeMnNi) 100–x sample spanning 0 ≤ x ≤ 21 at%. In addition to microstructural and mechanical characterization, synchrotron high speed x-ray computer aided tomography was conducted on a W 20 (CoCrFeMnNi) 80 composition to visualize melting dynamics, powder-laser interactions, and remelting effects of previously consolidated material. Results reveal the formation of the Fe 7 W 6 intermetallic phase at W concentrations> 6 at%, despite the high configurational entropy. Unincorporated W particles also occurred at W concentrations> 10 at% accompanied by a dissolution band of Fe 7 W 6 at the W/matrix interface and hardness values greater than 400 HV. In this work, the primary strengthening mechanism is attributed to the reinforcement of the Fe 7 W 6 and W phases as a metal matrix composite. The in situ high speed x-ray imaging during remelting showed that an additional laser pass did not promote further mixing of the Fe 7 W 6 or W phases suggesting that, despite the dissolution of the W into the Fe 7 W 6 phase being thermodynamically favored, it is kinetically limited by the thickness/diffusivity of the intermetallic phase, and the rapid solidification of the laser-based process.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Computational Tools and Workflows for Quantitative Risk Assessment and Decision Support for Geologic Carbon Storage Sites: Progress and Insights from the U.S. DOE’s National Risk Assessment Partnership

The 2005 Intergovernmental Panel on Climate Change (IPCC) Special Report on CCS raised the profile of CO2 capture and storage (CCS) as an important technology for reducing greenhouse gas (GHG) emissions. CCS is now recognized as a key component of most climate change mitigation scenarios. Since publication of that report the international research, development, and deployment (RD&D) community has advanced key technical aspects, clarified regulatory requirements, explored value chain and infrastructure solutions, and developed incentive paradigms to enable and promote large-scale deployment of CCS. These efforts have included research to better characterize geologic storage resources, to improve injection performance and storage efficiency, to assess and manage subsurface environmental risks, and to advance monitoring technologies to assure system conformance. These efforts have helped to build confidence in the viability of geologic carbon storage (GCS), but stakeholder concerns about long-term risks and liability associated with GCS remain a hurdle to broad acceptance and large-scale deployment of CCS. Since 2010, the U.S. DOE’s National Risk Assessment Partnership (NRAP) – a research collaboration between five contributing national laboratories – has worked to establish and demonstrate methods and tools to quantify and manage the subsurface environmental risks associated with GCS, amidst uncertainty. This work supports the Office of Fossil Energy and Carbon Management Carbon Transport and Storage Program’s goal of advancing safe and secure commercial-scale GCS deployment. To address the technical challenge of simulating the physical response of the GCS site to large-scale CO2 injection, NRAP has adopted an approach that relies on coupling computationally efficient reduced-order and/or data-driven proxy models of important system components (i.e., storage reservoir, sealing caprock, leakage pathways, intermediate formations, overlying groundwater aquifers, and the atmosphere) in integrated assessment framework. That integrated model of the physical system is complemented with fit-for purpose functionality to support site characterization and risk-related decisions. The recently released NRAP Phase II toolset includes the Open-Source Integrated Assessment Model (NRAP-Open-IAM) for evaluation of trends in leakage risk and potential impact, tools to support monitoring design optimization (Designs for Risk Evaluation and Management – DREAM v3.0 and Passive Seismic Monitoring Tool - PSMT), and tools for state of stress evaluation (State-of-Stress Analysis Tool - SOSAT) and forecasting induced seismicity risk. The NRAP team has also released a pair of reports describing conceptual workflows to incorporate physics-based, quantitative risk assessment into many of the design, planning, operation, and closure decisions for GCS projects. An online catalogue highlights published studies where these tools and methods are demonstrated. In this presentation, the utility of these products to assess risks and address key stakeholder questions will be highlighted through examples, and related insights about the safety and security of geologic carbon storage in qualified storage sites will be discussed. The prospect of rapid, large-scale deployment of GCS technology to aggressively reduce anthropogenic CO2 emissions requires careful consideration of interference between multiple commercial-scale storage projects within a basin. Going forward, NRAP is expanding and adapting site-scale risk quantification tools and methods to enable assessment of risks and inform management decisions for basin-scale deployment. Increasingly, this work will leverage next-generation approaches for surrogate modelling, fast prediction, and advanced visualization enabled by machine learning and artificial intelligence to promote virtual learning, scenario evaluation, and augment risk-based decision making.

quantitative risk assessment, geologic carbon stor↗
DOE OSTI2024

Weak shock compaction on granular salt

This study conducted integrated experiments and computational modeling to investigate the speeds of a developing shock within granular salt and analyzed the effect of various impact velocities up to 245 m/s. Experiments were conducted on table salt utilizing a novel setup with a considerable bore length for the sample, enabling visualization of a moving shock wave. Experimental analysis using particle image velocimetry enabled the characterization of shock velocity and particle velocity histories. Mesoscale simulations further enabled advanced analysis of the shock wave’s substructure. In simulations, the shock front’s precursor was shown to have a heterogeneous nature, which is usually modeled as uniform in continuum analyses. The presence of force chains results in a spread out of the shock precursor over a greater ramp distance. With increasing impact velocity, the shock front thickness reduces, and the precursor of the shock front becomes less heterogeneous. Furthermore, mesoscale modeling suggests the formation of force chains behind the shock front, even under the conditions of weak shock. This study presents novel mesoscale simulation results on salt corroborated with data from experiments, thereby characterizing the compaction front speeds in the weak shock regime.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2026

Quantifying the Impact of Metal Population Distribution in MFI-Supported Mo Catalysts for Methane Dehydroaromatization

Precise evaluation of intrinsic kinetic behavior in Mo/MFI catalysts for methane dehydroaromatization (MDA) is confounded by variations in Mo dispersion and speciation, which are influenced by metal loading and zeolite acidity. This work advances the utility of H 2 -temperature programmed reduction (H 2 -TPR) for characterizing Mo/MFI catalysts by enabling quantitative comparison of MoO x populations distinguished by reduction behavior and linked to initial catalytic performance. A comprehensive H 2 reduction pathway is established through systematic H 2 - TPR studies varying catalyst composition (1–10 wt % Mo, Si/Al = 15, 40, ∞), supplemented by UV-Raman spectroscopy, X-ray powder diffraction, N 2 physisorption, NH 3 -TPD, and advanced spectroscopic analysis (in situ XAS with principal component analysis/multivariate curve resolution—alternating least squares). Two low-temperature H 2 -TPR regions capture distinct Mo populations undergoing initial Mo(VI)→Mo(IV) reduction: a lower-temperature population (Mo-RI) associated primarily with highly dispersed, anchored MoO x species expected to predominantly reside within MFI channels, and a higher-temperature population (Mo-RII) corresponding to a broader set of MoO x species that becomes increasingly bulk-like/extrazeolitic at higher Mo loading. Quantification of these populations provides practical, kinetically relevant descriptors for comparing initial MDA rates across catalysts with varying Mo loading, Si/Al ratio, and MoO x heterogeneity. Application of lower-temperature Mo-RI estimates to kinetic measurements reveals a minimum threshold of ∼0.12 × 10 –3 mol Mo-RI/g cat , above which initial forward benzene rates normalized to this population converge despite differences in metal loading and zeolite Brønsted acidity. This threshold coincides with a transition toward a common C 2 -mediated benzene-forming regime, as indicated by approach-to-equilibrium analysis of methane-to-ethane, ethane-to-ethylene, and ethylene-to-benzene reaction steps. Above this threshold, normalized initial benzene rates are nearly invariant with increasing Mo-RII/Mo-RI population ratio, indicating that excess Mo-RII populations, including bulk-like/extrazeolitic MoO x domains present at higher loading, do not measurably suppress benzene formation associated with anchored and mostly channel-confined Mo population under the initial-rate conditions examined.

Mo/MFI↗
DOE OSTI2025

Initial Characterization of the NREL Large-Amplitude Motion Platform

The Large Amplitude Motion Platform (LAMP) at NREL represents a significant advancement in the controlled testing of Wave Energy Converters (WECs) under laboratory conditions. Originally designed by E2M as a six-degree-of-freedom (DOF) Stewart platform for flight simulation, LAMP has been adapted by NREL to facilitate the mounting and evaluation of WECs. This adaptation enables dry testing of WECs using motion profiles similar to the ocean, facilitating the iterative design, testing, and validation of WEC performance prior to ocean deployments. This report presents the initial work completed to characterize LAMP, with particular emphasis on its stability and operational capabilities across various single and multi-degree-of-freedom (DOF) motion profiles. The report includes planned comparisons at three distinct mass payloads, aimed at assessing the platform's positional accuracy, frequency response, and endurance over extended runtime periods. These experimental tests are critical for establishing the platform's limitations and ensuring that the data generated during WEC validation is both accurate and reproducible. The outcomes of this study not only contribute to a deeper understanding of LAMP's capabilities but also lay the groundwork for future advancements in WEC testing methodologies. By providing robust and reliable performance data within a controlled laboratory setting, the findings are expected to significantly enhance the development and commercialization of marine energy technologies. This report presents the initial findings of the LAMP Characterization work and proposed steps to further understand and characterize LAMP. Data collected during this work can be found on MHKDR at: https://mhkdr.openei.org/submissions/602 Note that this report shares the measured/found instantaneous maximum operating range of LAMP. For most applications the maximum operating range cannot be used for system health and longevity. The operating range and capabilities of LAMP will be evaluated on a case-by-case basis, single-DOF position, velocity, and acceleration values presented in Table 4a-c, and Table 5 should be taken as instantaneous absolute maximum values. Future use of LAMP will likely be limited to smaller values.

16 TIDAL AND WAVE POWER↗
DOE OSTI2022

Characterization of beam ion loss in high poloidal beta regime on EAST

A critical issue for achieving the integrated operation of steady-state long-pulse high-confinement (H-mode) plasmas on experimental advanced superconducting tokamak (EAST) is to improve beam ion population confinement during neutral beam injection (NBI). To study the characterization of beam ion loss and improve beam ion confinement, the steady-state long pulse scenario discharges were conducted on EAST (β p $\geqslant$ 2.0, β N $\geqslant$ 1.7, q 95 $\geqslant$ 6.7 and H 98y2 $\geqslant$ 1.1) with NBI heating. Based on neutron yield, the beam voltage and line-averaged electron density were adjusted from 50 kV to 60 kV and 4.4 × 10 19 m –3 to 5.0 × 10 19 m –3 , respectively. The results show that the dominant mechanisms of beam ion loss are shine-through loss, prompt loss, and stochastic ripple loss. The shine-through loss fraction is determined by initial velocity, flight time and entire beam path. The change in prompt loss fraction is caused by the change in the deposition of beam ions. The change in stochastic ripple loss fraction is caused by the change in the initial fraction of trapped-confined ions. Detailed physics shows that the prompt loss fraction during counter-Ip injections (~45%) is far larger than during co-Ip injections (~5%) due to the finite orbit width. The lost ions are mainly deposited on the lower divertor or below the midplane since the direction of magnetic drift is vertical down. The orbit types of prompt loss during counter-Ip injections are mainly trapped-lost and ctr-passing lost. To minimize the prompt loss fraction during counter-Ip injections, a reversed Ip configuration (rev-Ip) discharge #94758 was conducted. The result suggests that the beam ion wall load fraction during counter-Ip tangential injection (~3%) is far lower than that in normal Ip configuration (nor-Ip) discharge #94820. It is also found that the confinement of beam ion population in the counter-Ip injection #94758 was greatly improved when compared to #94820. Finally, this study can provide unique support for the improvement of beam ion population confinement and for the performance evaluation of the NBI system on EAST and future tokamaks.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2021

Quantifying and modeling of stress-driven short-circuits in lithium-ion batteries in electrified vehicles

Despite the huge expansion of electric vehicle sales in the market, customers are discouraged by the possible catastrophic consequences brought by the safety issues of lithium-ion batteries, such as internal short circuits, especially in crash scenarios. Herein, we reveal the quantitative relationship between the deformations of the battery and the internal short circuit. By in situ and ex situ observations and post mortem characterization of the component materials, we quantify the stress-driven internal short circuit and failure behavior of the component material. With the aid of a validated numerical computational model as well as the in situ characterization of the global-field temperature, we successfully identify the minor and major short circuits of the cells upon various mechanical abusive loadings. Finally, we establish the internal short circuit criteria for typical formats of batteries. Furthermore, this discovery also provides a fundamental understanding of both internal and external stress-driven short circuits in a much broader context.

25 ENERGY STORAGE↗
DOE OSTI2023

Characterization of wind turbine flow through nacelle-mounted lidars: a review

This article provides a comprehensive review of the most recent advances in the planning, execution, and analysis of inflow and wake measurements from nacelle-mounted wind Doppler lidars. Lidars installed on top of wind turbines provide a holistic view of the inflow and wake characteristics required to characterize and optimize wind turbine performance, carry out model validation and calibration, and aid in real-time control. The need to balance the enhanced capabilities and limitations of lidars compared to traditional anemometers inspired a broad variety of approaches for scan design and wind reconstruction, which we discuss in this review. We give particular emphasis to identifying common guidelines and gaps in the available literature with the aim of providing an exhaustive picture of the state-of-the-art techniques for reconstructing wind plant flow using nacelle-mounted lidars.

17 WIND ENERGY↗
DOE OSTI2021

Deep learning approaches to semantic segmentation of fatigue cracking within cyclically loaded nickel superalloy

Improvements to synchrotron-based micro-computed tomography scanning capabilities have gifted researchers the ability to characterize 4D material thermomechanical responses more thoroughly than ever before. These advancements, however, have brought about new challenges in analyzing the resulting deluge of data. We report on a nickel-based superalloy specimen imaged 26 times in-situ during cyclic loading at Argonne National Laboratory Advanced Photon Source beamline 1ID, in order to monitor crack growth within the microstructure. Therefore, several deep learning approaches which utilize convolutional neural networks are implemented to segment crack features from reconstructed tomography scans. U-Net architecture implementations are found to be especially effective, achieving IoU = 0.995 +/- 0.004 and Matthews correlation coefficient scores of Φ = 0.826 +/- 0.085. These advancements broaden possibilities for scientists seeking to automate segmentation analyses of similar large datasets.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Heterogeneous M-N-C Catalysts for Aerobic Oxidation Reactions: Lessons from Oxygen Reduction Electrocatalysts

Nonprecious metal heterogeneous catalysts composed of first-row transition metals incorporated into nitrogen-doped carbon matrices (M-N-Cs) have been studied for decades as leading alternatives to Pt for the electrocatalytic O 2 reduction reaction (ORR). More recently, similar M-N-C catalysts have been shown to catalyze the aerobic oxidation of organic molecules. Here, this Focus Review highlights mechanistic similarities and distinctions between these two reaction classes and then surveys the aerobic oxidation reactions catalyzed by M-N-Cs. As the active-site structures and kinetic properties of M-N-C aerobic oxidation catalysts have not been extensively studied, the array of tools and methods used to characterize ORR catalysts are presented with the goal of supporting further advances in the field of aerobic oxidation.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Transcription factor binding divergence drives transcriptional and phenotypic variation in maize

Regulatory elements are essential components of plant genomes that have shaped the domestication and improvement of modern crops. However, their identity, function and diversity remain poorly characterized, limiting our ability to harness their full power for agricultural advances using induced or natural variation. Here, in this study, we mapped transcription factor (TF) binding for 200 TFs from 30 families in two distinct maize inbred lines historically used in maize breeding. TF binding comparison revealed widespread differences between inbreds, driven largely by structural variation, that correlated with gene expression changes and explained complex quantitative trait loci such as Vgt1, an important determinant of flowering time, and DICE, an herbivore resistance enhancer. CRISPR–Cas9 editing of TF binding regions validated the function and structure of regulatory regions at various loci controlling plant architecture and biotic resistance. Our maize TF binding catalogue identifies functional regulatory regions and enables collective and comparative analysis, highlighting its value for agricultural improvement.

Galli, Mary [Rutgers Univ., Piscataway, NJ (United↗
DOE OSTI2020

SUGAR: An improved empirical model of Type Ia supernovae based on spectral features

Type Ia supernovae (SNe Ia) are widely used to measure the expansion of the Universe. Improving distance measurements of SNe Ia is one technique to better constrain the acceleration of expansion and determine its physical nature. This document develops a new SNe Ia spectral energy distribution (SED) model, called the SUpernova Generator And Reconstructor (SUGAR), which improves the spectral description of SNe Ia, and consequently could improve the distance measurements. This model was constructed from SNe Ia spectral properties and spectrophotometric data from the Nearby Supernova Factory collaboration. In a first step, a principal component analysis-like method was used on spectral features measured at maximum light, which allowed us to extract the intrinsic properties of SNe Ia. Next, the intrinsic properties were used to extract the average extinction curve. Third, an interpolation using Gaussian processes facilitated using data taken at different epochs during the lifetime of an SN Ia and then projecting the data on a fixed time grid. Finally, the three steps were combined to build the SED model as a function of time and wavelength. This is the SUGAR model. Results. The main advancement in SUGAR is the addition of two additional parameters to characterize SNe Ia variability. The first is tied to the properties of SNe Ia ejecta velocity and the second correlates with their calcium lines. The addition of these parameters, as well as the high quality of the Nearby Supernova Factory data, makes SUGAR an accurate and efficient model for describing the spectra of normal SNe Ia as they brighten and fade. The performance of this model makes it an excellent SED model for experiments like the Zwicky Transient Facility, the Large Synoptic Survey Telescope, or the Wide Field Infrared Survey Telescope.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2024

Archive of AGR-5/6/7 Particle Radiographs for Identification of Particles with Defective IPyC

As a part of fuel quality control characterization, 2D radiographs of large numbers of particles produced by the Advanced Gas Reactor Fuel Development and Qualification (AGR) Program’s AGR-5/6/7 irradiation were acquired. These radiographs were used for the identification of particles with excessive uranium dispersion from the kernel into the surrounding buffer layer caused by chlorine infiltration through a defective inner pyrolytic carbon (IPyC) layer during silicon carbide (SiC) deposition. Additional features of interest associated with fabrication anomalies were also catalogued. Raw radiography images, along with the noted defective IPyC defects found by analysis at Oak Ridge National Laboratory are reported herein.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2025

Unveiling the Hidden Evolution of Crystal Defects and Disorder in Energy Materials

Control of point defects and disorder in functional thin films and 2D materials is critical to realizing their full potential in applications ranging from energy storage to advanced electronics. However, these phenomena are often poorly understood, difficult to characterize, and challenging to direct with precision. This presentation explores emerging multi-modal computer vision to decipher and predict order in materials across multiple length scales in the electron microscope, from the atomic to the nanoscale. By fusing data from diverse sources, these powerful models provide unprecedented insights into materials' lifecycles, enabling the control of defects and their associated properties at a fundamental level. This capability promises to transform materials design and accelerate the development of next-generation technologies.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Technical note: Using long short-term memory models to fill data gaps in hydrological monitoring networks

Abstract. Quantifying the spatiotemporal dynamics in subsurface hydrological flows over a long time window usually employs a network of monitoring wells. However, such observations are often spatially sparse with potential temporal gaps due to poor quality or instrument failure. In this study, we explore the ability of recurrent neural networks to fill gaps in a spatially distributed time-series dataset. We use a well network that monitors the dynamic and heterogeneous hydrologic exchanges between the Columbia River and its adjacent groundwater aquifer at the U.S. Department of Energy's Hanford site. This 10-year-long dataset contains hourly temperature, specific conductance, and groundwater table elevation measurements from 42 wells with gaps of various lengths. We employ a long short-term memory (LSTM) model to capture the temporal variations in the observed system behaviors needed for gap filling. The performance of the LSTM-based gap-filling method was evaluated against a traditional autoregressive integrated moving average (ARIMA) method in terms of error statistics and accuracy in capturing the temporal patterns of river corridor wells with various dynamics signatures. Our study demonstrates that the ARIMA models yield better average error statistics, although they tend to have larger errors during time windows with abrupt changes or high-frequency (daily and subdaily) variations. The LSTM-based models excel in capturing both high-frequency and low-frequency (monthly and seasonal) dynamics. However, the inclusion of high-frequency fluctuations may also lead to overly dynamic predictions in time windows that lack such fluctuations. The LSTM can take advantage of the spatial information from neighboring wells to improve the gap-filling accuracy, especially for long gaps in system states that vary at subdaily scales. While LSTM models require substantial training data and have limited extrapolation power beyond the conditions represented in the training data, they afford great flexibility to account for the spatial correlations, temporal correlations, and nonlinearity in data without a priori assumptions. Thus, LSTMs provide effective alternatives to fill in data gaps in spatially distributed time-series observations characterized by multiple dominant frequencies of variability, which are essential for advancing our understanding of dynamic complex systems.

54 ENVIRONMENTAL SCIENCES↗