Search NASA⌕ Search

SEARCH · Search NASA

Results for “Structural Model Decomposition”

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 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Polysulfone shock compressed above the decomposition threshold: Velocimetry and modeling of two-wave structures

Polysulfone was shock loaded to pressures of 14.6–26.2 GPa in a series of gas gun-driven plate-impact experiments measuring material response with embedded electro-magnetic particle velocity gauges and optical velocimetry. The embedded electro-magnetic particle velocity gauges did not show a distinct two-wave structure but did show rounding that suggested a reaction but not a distinct separation of the reactants and product waves. In contrast, the transmission experiments fielded with optical velocimetry, with product pressures ranging from 21.0 to 26.2 GPa, showed well-defined two-wave structures due to shock-driven chemical decomposition of the polymer to products at a higher density. Distinct two-wave structures have not previously been published in literature during polymer compression; here, we observed these two-wave structures at both the polymer/lithium fluoride and polymer/polymethylmethacrylate interfaces. Hydrodynamic simulations of the experiments were performed using a newly constructed SESAME equation of state (EOS) for the polymer, a thermochemical EOS for the decomposition products, and an Arrhenius reaction rate model for the kinetics of the reaction. Finally, simulation results also demonstrated two-wave structures but were unable to quantitatively reproduce either the embedded gauge or optical velocimetry data.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2025

Benders Decomposition Using Graph Modeling and Multi-Parametric Programming

Benders decomposition is a widely used method for solving large and structured optimization problems, but its performance is affected by the repeated solution of subproblems. We propose a flexible and modular algorithmic framework for accelerating Benders decomposition. Specifically, we express the problem structure by using a graph-theoretic modeling abstraction in which nodes represent optimization subproblems and edges represent connectivity between subproblems. A key innovation of our approach is that we embed multiparametric programming (mp) surrogates for node subproblems, which maps the exact analytical map of the subproblem solution space. The use of mp surrogates allows us to replace subproblem solves with fast look-ups and function evaluations for primal and dual variables during the iterative Benders process. We formally show the equivalence between classical Benders cuts and those derived from the mp solution. We implement our framework in the open-source PlasmoBenders.jl software package. To demonstrate the capabilities of the proposed framework, we apply it to a two-stage stochastic programming problem, which aims to make optimal capacity expansion decisions under market uncertainty. We evaluate both single-cut and multicut variants of Benders decomposition and show that the use of mp surrogates achieves substantial speedups in subproblem solve time, while preserving the convergence guarantees of Benders decomposition. We highlight advantages in solution analysis and interpretability that is enabled by mp critical region tracking; specifically, we show that these reveal how decisions evolve geometrically across the Benders search. Our results aim to demonstrate that combining surrogate modeling with graph modeling offers a promising and extensible foundation for structure-exploiting decomposition. In addition, by decomposing the problem into more tractable subproblems, the proposed approach also aims to overcome scalability issues of mp. Finally, the use of mp surrogates provides a unifying and modular optimization framework that enables the representation of heterogeneous node subproblems as modeling objects with a homogeneous structure.

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

Local Structure Analysis and Modelling of Lignin‐Based Carbon Composites through the Hierarchical Decomposition of the Radial Distribution Function

Abstract Carbonized lignin has been proposed as a sustainable and domestic source of activated, amorphous, graphitic, and nanostructured carbon for many industrial applications as the structure can be tuned through processing conditions. However, the inherent variability of lignin and its complex physicochemical structure resulting from feedstock and pulping selection make the Process‐Structure‐Property‐Performance (PSPP) relationships hard to define. In this work, radial distribution functions (RDFs) from synchrotron X‐ray and neutron scattering of lignin‐based carbon composites (LBCCs) are investigated using the Hierarchical Decomposition of the Radial Distribution Function (HDRDF) modelling method to characterize the local atomic environment and develop quantitative PSPP relationships. PSPP relationships for LBCCs defined by this work include crystallite size dependence on lignin feedstock as well as increasing crystalline volume fraction, nanoscale composite density, and crystallite size with increasing reduction temperature.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

Polyimide dynamically compressed to decomposition pressures: Two-wave structures captured by velocimetry and modeling

We performed a series of six plate impact experiments on polyimide and modeled them using new reactant and products equations of state combined with an Arrhenius rate model. The first experiment was diagnosed with embedded electromagnetic velocity gauges through which we directly observed attenuation of the lead shock to an approximately constant state over a propagation distance of roughly 4 mm. Simulated gauge profiles were in excellent qualitative agreement with experiment and suggested a sluggish chemical reaction that did not proceed to completion. The remaining five experiments were conducted in a transmission geometry and diagnosed velocimetrically at the sample/window interface. All five of these yielded profiles with a sharp shock followed by a more gradual approach to maximum interface velocity that was “rounded” to varying degree. These profiles proved difficult to interpret unambiguously due to the convolution of the reactive wave upon first shock with reflection of the lead wave and reshock or release by the window. Comparison with thermochemical calculations strongly suggests that the point of maximum interface velocity corresponds to the equilibrium reshock or release locus. We discuss the implications of this point for the practice of impedance matching based on the reflected Hugoniot of reactive materials such as polymers. The reactant and thermochemical products equations of state are developmental SESAME tables 97710 and 97720, respectively.

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

The Stellar decomposition: A compact representation for simplicial complexes and beyond

Here, we introduce the Stellar decomposition, a model for efficient topological data structures over a broad range of simplicial and cell complexes. A Stellar decomposition of a complex is a collection of regions indexing the complex’s vertices and cells such that each region has sufficient information to locally reconstruct the star of its vertices, i.e., the cells incident in the region’s vertices. Stellar decompositions are general in that they can compactly represent and efficiently traverse arbitrary complexes with a manifold or non-manifold domain. They are scalable to complexes in high dimension and of large size, and they enable users to easily construct tailored application-dependent data structures using a fraction of the memory required by a corresponding global topological data structure on the complex. As a concrete realization of this model for spatially embedded complexes, we introduce the Stellar tree, which combines a nested spatial tree with a simple tuning parameter to control the number of vertices in a region. Stellar trees exploit the complex’s spatial locality by reordering vertex and cell indices according to the spatial decomposition and by compressing sequential ranges of indices. Stellar trees are competitive with state-of-the-art topological data structures for manifold simplicial complexes and offer significant improvements for cell complexes and non-manifold simplicial complexes. We conclude with a high-level description of several mesh processing and analysis applications that utilize Stellar trees to process large datasets.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Pathway for decarbonizing residential building operations in the US and China beyond the mid-century

With global carbon budget targets looming, residential buildings in top economies must become carbon neutral as soon as possible to reserve more emission space for emerging carbon-emitting economies. This study is the first to compare the operational decarbonization process of China’s and the United States (US) residential buildings from 2000 to 2060 by combining the end-use emission model with the decomposing structural decomposition (DSD) method and Monte Carlo simulation. The results show that from 2001 to 2020 China decarbonized 1544 mega-tons of carbon dioxide (MtCO 2 ) and the US decarbonized 1848 MtCO 2 . In the business-as-usual scenario, China will hit its emission peak in 2031 (±3) with 934 (±61) MtCO 2 , while the US will maintain a lock-in level of 736 (±133) MtCO 2 since the 2030s. In the decarbonization scenario, operational carbon neutrality for residential buildings in 2060 is promoted by an increase in clean power generation proportion, building-integrated power generation level, building electrification level, and a reduction in end-use energy intensity, which will contribute 34.4 %, 21.4 %, 14.3 %, and 29.9 % in China and 32.9 %, 33.1 %, 8.2 %, and 25.8 % in the US, respectively. Especially, building-integrated power generation in China only costs about 40 % of what it costs in the US. Besides, high-decarbonization strategies for residential building operations are proposed as references for governments to formulate targeted climate policies. Overall, this study offers data benchmarks for buildings’ carbon neutrality of top economies to further promote synergistic carbon neutrality with the buildings of emerging economies in the age of Post COP27.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2026

Dynamic analysis of fully constrained Cable-Driven Parallel Robots for automated prefabricated component installation

This paper presents a dynamic analysis and validation framework to assess a fully constrained six-anchor Cable-Driven Parallel Robot (CDPR) for automated installation of prefabricated facade components. Compared with conventional eight-anchor systems, the six-anchor configuration simplifies setup and reduces cost, but it also reduces control authority, shrinks the wrench-feasible workspace, and tightens orientation limits. Consequently, it is unclear a priori whether dynamically feasible trajectories exist to move the end effector from pickup to the facade. A constrained trajectory optimization is formulated to enforce the system dynamics, cable-tension bounds, and pose/velocity limits, and the framework is evaluated in simulation at three levels: (i) an idealized reference model, (ii) a lab-scale prototype model incorporating measured anchor misalignments and identified damping, and (iii) a full-scale three-story building model with load decomposition for structural feasibility checks. Across these scenarios, the analysis shows that optimal, constraint-satisfying trajectories exist that move the end effector from pickup to installation while maintaining a near-plumb, level orientation at the final pose. Collectively, this multi-scale dynamic analysis and validation framework supports the deployment readiness of the six-anchor CDPR and provides a prototype-based sensitivity case study of how measured anchor placement deviations affect feasibility.

CDPR↗
DOE OSTI2025

Block-Structured Operator Inference for Coupled Multiphysics Model Reduction

This work presents a block-structured formulation of Operator Inference as a way to learn structured reduced-order models for multiphysics systems. The approach specifies the governing equation structure for each physics component and the structure of the coupling terms. Once the multiphysics structure is specified, the reduced-order model is learned from snapshot data following the nonintrusive Operator Inference methodology. In addition to preserving physical system structure, which in turn permits preservation of system properties such as stability and second-order structure, the block-structured approach has the advantages of reducing the overall dimensionality of the learning problem and admitting tailored regularization for each physics component. The numerical advantages of the block-structured formulation over a monolithic Operator Inference formulation are demonstrated for aeroelastic analysis, which couples aerodynamic and structural models. For the benchmark test case of the AGARD 445.6 wing, block-structured Operator Inference provides an average 20% online prediction speedup over monolithic Operator Inference across subsonic and supersonic flow conditions in both the stable and fluttering parameter regimes while preserving the accuracy achieved with monolithic Operator Inference.

42 ENGINEERING↗
DOE OSTI2020

Parameter Sensitivity Analysis of the SparTen High Performance Sparse Tensor Decomposition Software (Extended Analysis)

Tensor decomposition models play an increasingly important role in modern data science applications. One problem of particular interest is fitting a low-rank Canonical Polyadic (CP) tensor decomposition model when the tensor has sparse structure and the tensor elements are nonnegative count data. SparTen is a high-performance C++ library which computes a low-rank decomposition using different solvers: a first-order quasi-Newton or a second-order damped Newton method, along with the appropriate choice of runtime parameters. Since default parameters in SparTen are tuned to experimental results in prior published work on a single real-world dataset conducted using MATLAB implementations of these methods, it remains unclear if the parameter defaults in SparTen are appropriate for general tensor data. Furthermore, it is unknown how sensitive algorithm convergence is to changes in the input parameter values. This report addresses these unresolved issues with large-scale experimentation on three benchmark tensor data sets. Experiments were conducted on several different CPU architectures and replicated with many initial states to establish generalized profiles of algorithm convergence behavior.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Convergence in simulating global soil organic carbon by structurally different models after data assimilation

Abstract Current biogeochemical models produce carbon–climate feedback projections with large uncertainties, often attributed to their structural differences when simulating soil organic carbon (SOC) dynamics worldwide. However, choices of model parameter values that quantify the strength and represent properties of different soil carbon cycle processes could also contribute to model simulation uncertainties. Here, we demonstrate the critical role of using common observational data in reducing model uncertainty in estimates of global SOC storage. Two structurally different models featuring distinctive carbon pools, decomposition kinetics, and carbon transfer pathways simulate opposite global SOC distributions with their customary parameter values yet converge to similar results after being informed by the same global SOC database using a data assimilation approach. The converged spatial SOC simulations result from similar simulations in key model components such as carbon transfer efficiency, baseline decomposition rate, and environmental effects on carbon fluxes by these two models after data assimilation. Moreover, data assimilation results suggest equally effective simulations of SOC using models following either first‐order or Michaelis–Menten kinetics at the global scale. Nevertheless, a wider range of data with high‐quality control and assurance are needed to further constrain SOC dynamics simulations and reduce unconstrained parameters. New sets of data, such as microbial genomics‐function relationships, may also suggest novel structures to account for in future model development. Overall, our results highlight the importance of observational data in informing model development and constraining model predictions.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Thermodynamic re-modelling of the Cu–Nb–Sn system: Integrating the nausite phase

Currently available Cu–Nb–Sn phase diagrams lack the recently discovered nausite phase (Cu,Nb)Sn 2 , which is an important intermediate in the course of thermal processing of superconducting Nb 3 Sn wires. Processing decisively determines the resulting microstructure of Nb 3 Sn and, thus, its superconducting properties. Lack of suitable and complete phase diagrams, however, obstructs rational design of such thermal processing procedures. To close this gap and to obtain valid knowledge of homogeneity and stability range of nausite, various Cu–Nb–Sn samples, which are heat-treated between 300 °C and 500 °C, are investigated. By means of energy-dispersive X-ray spectroscopy (EDX), a temperature-dependent homogeneity range of nausite is observed, which covers average mole fractions of Cu between 0.09 and 0.15. This is correlated with a change in the mean atomic volume and can be seen in the lattice parameters determined by X-ray diffraction (XRD). Additionally performed first-principles calculations on different CuSn 2 and NbSn 2 model structures confirm this trend. Furthermore, the peritectic decomposition of nausite to NbSn 2 and liquid at 586 °C is determined by means of in situ XRD and differential scanning calorimetry (DSC). By using the CALPHAD (CALculation of PHase Diagrams) approach, all these findings are used to extend a previous thermodynamic description of the Cu–Nb–Sn system by including the nausite as an additional phase. Finally, with this noteworthy integration, the updated modelling of the Cu–Nb–Sn system can be used for optimizing the multistage heat-treatment steps during processing superconducting Nb 3 Sn wires.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Detailed modeling of hydrogen release and particle shrinkage during pyrolysis of inhomogeneous wood

Hydrogen release during pyrolysis of woody biomass is studied considering anisotropicity and inhomogeneity of wood structure. A new anisotropic shrinkage model is proposed based on the decomposition of main wood constituents, i.e., cellulose, hemicellulose, and lignin. The new shrinkage model can predict the temporal evolution of the wood structure, and the differences between axial and radial shrinkage during pyrolysis. The model agrees very well with several experimental data from the literature. Based on particle temperature during conversion, the pyrolysis is partitioned into four stages, and the hydrogen release and H 2 formation from each stage are investigated. Stage (IV) of pyrolysis, from 1000 to 1273 K, is found to be efficient for H 2 production owing to the production of considerable mass of H with a minimal amount of tar species. Furthermore, the char quality is found to be different at the end of stages (II), (III), and (IV), where around 67.7, 80.5, and 93.4% wt. of solid residue is made of carbon, respectively. The model is also used to explain how the heating rate affects the temperature distribution inside the particle and how it shifts the peak of hydrogen release. Finally, the pyrolysis of two inhomogeneous wood samples — a beech twig with bark and a beech dowel with growth rings — are investigated. The bark can affect the pyrolysis rate, products, and flow pattern inside the particle. The growth rings do not have a considerable effect on the pyrolysis rate and products, but they have a significant impact on the flow pattern. This has an important implication for char conversion studies where the internal surface area and porosity field distribution have a significant effect on the gasification and oxidation rates.

09 BIOMASS FUELS↗
DOE OSTI2024

Electromagnetic Energy-Assisted Thermal Conversion of Fossil-Based Hydrocarbons to Low-Cost Hydrogen

The goal of this project was to develop and optimize catalysts for methane decomposition, particularly focusing on regeneration via an electromagnetic energy-assisted mechanism, to produce hydrogen more cost-effectively compared to electrolysis routes. To achieve this goal, the project pursued several key objectives. The project began with the preparation and testing of various catalysts. A nickel-silica based catalyst was identified as the most promising material for the pyrolysis of methane into carbon and hydrogen. Kinetic parameters for methane decomposition were determined, aiding in computational modeling efforts. Structured catalysts were investigated, highlighting the need for frequent cleaning or regeneration to maintain performance, with methane conversion rates exceeding 70% in tube furnace tests. Computational fluid dynamics modeling was employed to optimize reactor designs and electrode angles, leading the project team to propose a multi-compartment thermal conversion system for larger setups. This modeling work was important in understanding reaction characteristics, carbon deposition rates, and temperature profiles under various conditions. A bench-scale reactor system was assembled to evaluate catalyst regeneration using electromagnetic energy-assisted mechanisms. Experiments demonstrated the potential for carbon removal, though further optimization is needed. The carbon produced from the methane conversion process was evaluated for potential use in lithium-ion battery electrodes. The carbon exhibited properties similar to commercially available high-purity multi-walled carbon nanotubes and nanofibers, with a carbon content greater than 95%. Coin cell batteries assembled with this carbon showed that lower replacement levels (10% to 33%) outperformed the control group, improving specific capacity density and stability. However, higher replacement levels (100%) demonstrated poorer performance, suggesting that excessive carbon substitution negatively impacts battery performance. These findings indicate the potential marketability of the produced carbon as a component in lithium-ion batteries, though further testing is necessary to confirm long-term advantages and disadvantages associated with the use of the carbon product. These results however justified further technoeconomic assessments to determine if the process can provide low-cost hydrogen. The economic feasibility and technical performance of methane decomposition for hydrogen production were assessed, focusing on three plant configurations: 100E (electrically heated), 100C (combustion heated using produced hydrogen), and PE-Hybrid (a combination of pyrolysis (indicating decomposition) and electrolysis). The Levelized Cost of Hydrogen (LCOH) for the pyrolysis configurations was found to be approximately 25% lower than that of electrolysis. The 100E configuration had the lowest LCOH at $\$$3.12/kg. Including carbon product sales significantly improved the economics, with the 100C configuration achieving a negative LCOH of -$\$$0.35/kg. The PE-Hybrid configuration was not economically advantageous compared to pure pyrolysis plants due to its complexity and additional equipment requirements. Ultimately, methane pyrolysis presents a viable method for near carbon dioxide-free hydrogen production, with significant economic advantages over electrolysis, especially when considering the sale of carbon byproducts. The 100E and 100C configurations showed the most promise, with the choice between them ultimately depending on the prices of power and natural gas. In conclusion, this technology has the potential to lower hydrogen production costs by leveraging the methane decomposition process with the sale of valuable carbon byproducts. By optimizing catalyst performance and integrating electromagnetic energy-assisted regeneration, the process can achieve higher efficiency and economic viability, making it a competitive alternative to traditional hydrogen production methods.

08 HYDROGEN↗
DOE OSTI2022

Energetically consistent model reduction for metriplectic systems

The metriplectic formalism is useful for describing complete dynamical systems which conserve energy and produce entropy. This creates challenges for model reduction, as the elimination of high-frequency information will generally not preserve the metriplectic structure which governs long-term stability of the system. Based on proper orthogonal decomposition, a provably convergent metriplectic reduced-order model is formulated which is guaranteed to maintain the algebraic structure necessary for energy conservation and entropy formation. Further, numerical results on benchmark problems show that the proposed method is remarkably stable, leading to improved accuracy over long time scales at a moderate increase in cost over naive methods.

42 ENGINEERING↗
DOE OSTI2026

BrickQA: Bridging the Semantic Gap in Building Operations with Dynamic Graph Exploration

While standardized ontologies like the Brick schema address data heterogeneity in Building Automation Systems (BAS), accessing this semantic data remains a challenge as domain experts often lack the expertise to formulate complex SPARQL queries. To bridge this gap, we present BrickQA, a Large Language Model (LLM)-based framework that translates natural language into executable SPARQL queries through structured query decomposition, dynamic schema exploration, and inline validation. BrickQA utilizes an iterative reasoning agent to actively navigate graph topology through dynamic exploration actions without requiring exhaustive context injection or model fine-tuning. This approach effectively mitigates hallucinations, particularly in large-scale building knowledge graphs. Empirical evaluation on BuildingQA, a standardized benchmark, demonstrates that BrickQA significantly outperforms ReAct baselines, delivering a 0.291–0.355 absolute F1 improvement while achieving 3 × –12.7 × higher token cost-efficiency. Beyond these metrics, the framework maintains structural fidelity across heterogeneous buildings and remains resilient to ambiguous queries without requiring site-specific fine-tuning. Furthermore, a case study on operational analytics validates the framework’s capability to handle temporal and aggregation constraints, effectively transforming abstract semantic models into actionable facility management insights.1

Ko, Yun-Dam↗
DOE OSTI2023

Applications of the Dulmage–Mendelsohn decomposition for debugging nonlinear optimization problems

Nonlinear modeling and optimization is a valuable tool for aiding decisions by engineering practitioners, but programming an optimization problem based on a complex electrical, mechanical, or chemical process is a time-consuming and error-prone activity. Therefore, there is a need for model analysis and debugging tools that can detect and diagnose modeling errors. One such tool is the Dulmage–Mendelsohn decomposition, which identifies structurally under- and over-determined subsets in systems of equations and variables by partitioning the bipartite graph of the system. This work provides the necessary background to understand the Dulmage–Mendelsohn decomposition and its application to the analysis of nonlinear optimization problems, demonstrates its use in diagnosing a variety of modeling errors, and introduces software implementations for analyzing nonlinear optimization problems in the Pyomo and JuMP algebraic modeling languages.

42 ENGINEERING↗
DOE OSTI2021

First principles modeling of polymer encapsulant degradation in Si photovoltaic modules

An outstanding issue in the longevity of photovoltaic (PV) modules is the accelerated degradation caused by the presence of moisture. Moisture leads to interfacial instability, de-adhesion, encapsulant decomposition, and contact corrosion. However, experimental characterization of moisture in PV modules is not trivial and its impacts can take years or decades to establish in the field, presenting a major obstacle to designing high-reliability modules. First principles calculations provide an alternative way to study the ingress of water and its detrimental effect on the structure and decomposition of the polymer encapsulant and interfaces between the encapsulant and the semiconductor, the metal contacts, or the dielectric layer. Here, we use density functional theory (DFT) computations to model single chain, crystalline and cross-linked structures, infrared (IR) signatures, and degradation mechanisms of ethylene vinyl acetate (EVA), the most common polymer encapsulant used in Si PV modules. IR-active modes computed for low energy EVA structures and possible decomposition products match well with reported experiments. The EVA decomposition energy barriers computed using the Nudged Elastic Band (NEB) method show a preference for acetic acid formation as compared to acetaldehyde, are lowered in the presence of a water solvent or hydroxyl ion catalyst, and match well with reported experimental activation energies. This systematic study leads to a clear picture of the hydrolysis-driven decomposition of EVA in terms of energetically favorable mechanisms, possible intermediate structures, and IR signatures of reactants and products.

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

Equation‐Free Surrogate Modeling of Geophysical Flows at the Intersection of Machine Learning and Data Assimilation

Abstract There is a growing interest in developing data‐driven reduced‐order models for atmospheric and oceanic flows that are trained on data obtained either from high‐resolution simulations or satellite observations. The data‐driven models are non‐intrusive in nature and offer significant computational savings compared to large‐scale numerical models. These low‐dimensional models can be utilized to reduce the computational burden of generating forecasts and estimating model uncertainty without losing the key information needed for data assimilation (DA) to produce accurate state estimates. This paper aims at exploring an equation‐free surrogate modeling approach at the intersection of machine learning and DA in Earth system modeling. With this objective, we introduce an end‐to‐end non‐intrusive reduced‐order modeling (NIROM) framework equipped with contributions in modal decomposition, time series prediction, optimal sensor placement, and sequential DA. Specifically, we use proper orthogonal decomposition (POD) to identify the dominant structures of the flow, and a long short‐term memory network to model the dynamics of the POD modes. The NIROM is integrated within the deterministic ensemble Kalman filter (DEnKF) to incorporate sparse and noisy observations at optimal sensor locations obtained through QR pivoting. The feasibility and the benefit of the proposed framework are demonstrated for the NOAA Optimum Interpolation Sea Surface Temperature (SST) V2 data set. Our results indicate that the NIROM is stable for long‐term forecasting and can model dynamics of SST with a reasonable level of accuracy. Furthermore, the prediction accuracy of the NIROM gets improved by almost one order of magnitude by the DEnKF algorithm.

Pawar, Suraj↗