Results for “Generative model” — Page 23

DOE OSTI2024

Plastic Parallel Pathways Platform- 4P Model

The U.S. generates 42 million metric tons of plastic waste each year - the most of any country - of which less than 9% is recycled. This represents an estimated loss of $2.3 billion and 3.4 EJ embodied energy per year. Plastic use reduction, reuse and recycling are thus increasingly important, but making informed policy and research decisions within this space can be challenging given the diverse range of available solutions. Various recycling and waste management options are available for plastics, such as conversion into energy (incineration, pyrolysis, gasification), conversion to lower- or higher-grade plastics (mechanical recycling, chemical or enzymatic depolymerization processes, dissolution) or up-cycling to feedstock chemicals (photo/electro/catalytic methods, pyrolysis, gasification). While many previous publications have investigated a subset of these options, there is currently no modelling platform that can quantitatively assess and compare the economic and environmental impacts of all these different plastic management pathways and their respective products simultaneously. To directly compare the benefits and disadvantages of plastic-to-x technologies, we propose a novel analysis framework: the Plastic Parallel Pathways Platform (4P). 4P will model a systems analysis framework for capturing plastic material flows processed via different waste management scenarios. The total economic cost and greenhouse gas emissions (as well as additional economic, energetic, and environmental metrics indicators) of a given scenario will be calculated enabling comparison to other scenarios, thereby informing decisions on plastic recycling pathways.

circular economy↗

DOE OSTI2026

Operational Parameter Database for Molten Salt Thermal Energy Storage Tank Modeling

The second generation of concentrated solar power (CSP) plants is characterized by the use of a central receiver (either cavity or external), two molten nitrate salt tanks (60 wt.% NaNO 3 and 40 wt.% KNO 3 ), and a steam Rankine power-generation cycle connected through a primary heat exchanger. Molten salt thermal energy storage (TES) tanks have been widely deployed in commercial CSP plants worldwide and have been essential for increasing plant dispatchability and capacity factor, while also reducing the levelized cost of electricity (LCOE). These systems enable energy storage at the gigawatt-hour scale, typically providing 6 to 17 hours of storage duration. Despite being a commercial technology, the multiple failures observed after only a few months or years of operation in plants around the world demonstrate the technology's relative infancy and highlight the need for further research to improve its reliability. The National Laboratory of the Rockies (NLR), in collaboration with industry partners and academic and research institutions, has been leading multiple projects funded by the U.S. Department of Energy (DOE). These projects focus on addressing molten salt tank failures by improving tank design and welding fabrication practices, evaluating new alloys and weld fillers, and providing guidelines for tank commissioning and safe operation. In particular, this report presents modeling results on the effect of key tank operation parameters during 60 minutes of operation, including the mass flow rate and temperature of the salt inflow, tank salt inventory temperature, and inventory level for a representative molten salt tank design. These results form a database of tank operation behaviors that captures the effects of each specific parameter during charging, charging/discharging, and discharging processes.

14 SOLAR ENERGY↗

DOE OSTI2024

Generative network-based approaches to generate stochastic realizations

Generative Adversarial Network (GAN) – based models have been successfully applied in generating different geological models in the literature. However, it is still challenging to use GAN to generate geological realizations with extremely sparse conditioning data (e.g. several well data), which may be regarded as local noise by GAN during the training process. In this work, we propose a novel conditional Generative Adversarial Neural Operator (cGANO) to tackle this challenge. In cGANO, the mapping between conditioning data and output is established through the U-shaped neural operators (UNO), which better preserves local information. Another advantage of using UNO comes from its grid-independent property, which makes the generation of downscaling stochastic geologic realizations possible. We tested the model performance on the IBDP geostatistical dataset with 100 realizations.

58 GEOSCIENCES↗

DOE OSTI2024

Accelerated CO2 Storage Optimization Using Multi-Resolution Fourier Neural Operator at the Illinois Basin Decatur Project (IBDP)

This paper presents a deep learning-based approach for optimizing CO2 injection in carbon capture and storage (CCS) operations. We developed a multi-resolution machine learning model to significantly reduce data generation costs. Utilizing this proxy model, we implemented a multi-objective genetic algorithm to optimize well control during the CO2 injection process. The proposed approach was applied to the Illinois Basin Decatur Project (IBDP), successfully optimizing the CO2 injection schedule based on three key objectives: maximizing the amount of CO2 stored, maximizing sweep efficiency, and minimizing pressure increase. The use of the proxy model accelerated the optimization workflow by two orders of magnitude, while the cost of data generation for the proxy model was reduced by 90% by utilizing a coarse-scale model.

accelerated CO2 storage optimization↗

DOE OSTI2024

Generating synthetic signaling networks for in silico modeling studies

Predictive models of signaling pathways have proven to be difficult to develop. Reasons include the uncertainty in the number of species, the complexity in species’ interactions, and the sparseness and uncertainty in experimental data. Traditional approaches to developing mechanistic models rely on collecting experimental data and fitting a single model to that data. This approach works for simple systems but has proven unreliable for complex systems such as biological signaling networks. For example, uncertainty and sparseness of the data often result in overfitted models that have little predictive value beyond recapitulating the experimental data itself. Thus, there is a need to develop new approaches to create predictive mechanistic models of complex systems. However, to determine the effectiveness of any new algorithm, a baseline model is needed to test its performance. To meet this need, we developed a method for generating artificial synthetic networks that are reasonably realistic and thus can be treated as ground truth models. These synthetic models can then be used to generate synthetic data for developing and testing algorithms designed to recover the underlying network topology and associated parameters. Here, we describe a simple approach for generating synthetic signaling networks that can be used for this purpose.

42 ENGINEERING↗

DOE OSTI2026

ARM Data-Oriented Metrics and Diagnostics Package for Climate Model Evaluation

A Python-based metrics and diagnostics package is currently being developed by the U.S. Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) Infrastructure Team at Lawrence Livermore National Laboratory (LLNL) to facilitate the use of long-term, high-frequency measurements from the ARM Facility in evaluating the regional climate simulation of clouds, radiation, and precipitation. This metrics and diagnostics package computes climatological means of targeted climate model simulation and generates tables and plots for comparing the model simulation with ARM observational data. The Coupled Model Intercomparison Project (CMIP) model data sets are also included in the package to enable model intercomparison as demonstrated in Zhang et al. (2017). The mean of the CMIP model can serve as a reference for individual models. Basic performance metrics are computed to measure the accuracy of mean state and variability of climate models. The evaluated physical quantities include cloud fraction, temperature, relative humidity, cloud liquid water path, total column water vapor, precipitation, sensible and latent heat fluxes, and radiative fluxes, with plan to extend to more fields, such as aerosol and microphysics properties. Process-oriented diagnostics focusing on individual cloud- and precipitation-related phenomena are also being developed for the evaluation and development of specific model physical parameterizations. The version 1.0 package is designed based on data collected at ARM’s Southern Great Plains (SGP) Research Facility, with the plan to extend to other ARM sites. The metrics and diagnostics package is currently built upon standard Python libraries and additional Python packages developed by DOE (such as CDMS and CDAT). The ARM metrics and diagnostic package is available publicly with the hope that it can serve as an easy entry point for climate modelers to compare their models with ARM data. In this report, we first present the input data, which constitutes the core content of the metrics and diagnostics package in section 2, and a user's guide documenting the workflow/structure of the version 1.0 codes, and including step-by-step instruction for running the package in section 3.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2025

High Resolution Siting Suitability of Various Power Plant Technologies

Energy sector planning models determine the aggregate need for new generation, but these models are typically at the state or regional scale and are not equipped to address the wide range of location- and technology-specific issues that are increasingly a factor in power plant siting. These animations demonstrate the aggregate siting suitability of various power plant technology configurations, considering technology-specific factors that can prohibit development. The data presented is from the GRIDCERF (Geospatial Raster Input Data for Capacity Expansion Regional Feasibility) data package. GRIDCERF is a harmonized, open-source geospatial product that can be used to evaluate siting suitability for renewable and non-renewable power plants in the conterminous United States. The animations presented here demonstrate a curated selection of the full suite of technology configurations available. GRIDCERF provides the necessary inputs for models that simulate power plant siting for regional capacity expansion planning such as the Capacity Expansion Regional Feasibility (CERF) model.

Mongird, Kendall [Pacific Northwest National Labor↗

DOE OSTI2025

Hybrid learning techniques for scientific data reduction with performance guarantees

The research initiatives supported by the U.S. Department of Energy (DOE) Grant DE-SC0022265 are fundamentally aimed at pioneering advanced machine learning (ML) techniques for scientific data compression within high-performance computing (HPC) environments. This comprehensive body of work addresses the critical challenge posed by the exponential growth of data generated by scientific simulations in domains such as fusion energy, climate modeling, and computational fluid dynamics (CFD). A core objective is to develop compression algorithms that achieve substantial data reduction—often by orders of magnitude—while rigorously ensuring the fidelity of both the primary data (PD) and scientifically crucial derived quantities of interest (QoI). The methodologies deployed under this grant integrate sophisticated deep learning architectures, prominently featuring autoencoders, advanced generative models like conditional diffusion, and hybrid learning techniques. Key innovations include the development of Guaranteed Autoencoders (GAE) and the Guaranteed Conditional Diffusion with Tensor Correction (GCDTC) framework, which provide explicit, instance-level error bounds on reconstructed data. Furthermore, specialized strategies such as nonlinear constraint satisfaction are employed to preserve the integrity of QoI, a vital requirement for the trustworthiness of downstream scientific analyses. This research also focuses on the design and implementation of scalable, GPU-accelerated software pipelines that seamlessly integrate into existing HPC workflows, ensuring both computational efficiency and practical applicability. The CAESAR framework, for example, unifies foundation and generative models to create an adaptive and efficient compression solution for spatio-temporal scientific data. Collectively, these efforts represent a significant advancement in mitigating the scientific data deluge, enabling more effective data management, accelerated scientific discovery, and optimized utilization of HPC resources.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2025

Final report- UFL - RAPIDS2: A SciDAC Institute for Computer Science, Data, and Artificial Intelligence

The research initiatives supported by the U.S. Department of Energy (DOE) Grant DE-SC0022265 are fundamentally aimed at pioneering advanced machine learning (ML) techniques for scientific data compression within high-performance computing (HPC) environments. This comprehensive body of work addresses the critical challenge posed by the exponential growth of data generated by scientific simulations in domains such as fusion energy, climate modeling, and computational fluid dynamics (CFD). A core objective is to develop compression algorithms that achieve substantial data reduction—often by orders of magnitude—while rigorously ensuring the fidelity of both the primary data (PD) and scientifically crucial derived quantities of interest (QoI). The methodologies deployed under this grant integrate sophisticated deep learning architectures, prominently featuring autoencoders, advanced generative models like conditional diffusion, and hybrid learning techniques. Key innovations include the development of Guaranteed Autoencoders (GAE) and the Guaranteed Conditional Diffusion with Tensor Correction (GCDTC) framework, which provide explicit, instance-level error bounds on reconstructed data. Furthermore, specialized strategies such as nonlinear constraint satisfaction are employed to preserve the integrity of QoI, a vital requirement for the trustworthiness of downstream scientific analyses. This research also focuses on the design and implementation of scalable, GPU-accelerated software pipelines that seamlessly integrate into existing HPC workflows, ensuring both computational efficiency and practical applicability. The CAESAR framework, for example, unifies foundation and generative models to create an adaptive and efficient compression solution for spatio-temporal scientific data. Collectively, these efforts represent a significant advancement in mitigating the scientific data deluge, enabling more effective data management, accelerated scientific discovery, and optimized utilization of HPC resources.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2025

From disorganized data to emergent dynamic models: Questionnaires to partial differential equations

Starting with sets of disorganized observations of spatially varying and temporally evolving systems, obtained at different (also disorganized) sets of parameters, we demonstrate the data-driven derivation of parameter dependent, evolutionary partial differential equation (PDE) models capable of generating the data. This tensor type of data is reminiscent of shuffled (multidimensional) puzzle tiles. The independent variables for the evolution equations (their “space” and “time”) as well as their effective parameters are all emergent , i.e. determined in a data-driven way from our disorganized observations of behavior in them. We use a diffusion map based questionnaire approach to build a smooth parametrization of our emergent space/time/parameter space for the data. This approach iteratively processes the data by successively observing them on the “space,” the “time” and the “parameter” axes of a tensor. Once the data become organized, we use machine learning (here, neural networks) to approximate the operators governing the evolution equations in this emergent space. Our illustrative examples are based (i) on a simple advection–diffusion model; (ii) on a previously developed vertex-plus-signaling model of Drosophila embryonic development; and (iii) on two complex dynamic network models (one neuronal and one coupled oscillator model) for which no obvious smooth embedding geometry is known a priori. This allows us to discuss features of the process like symmetry breaking, translational invariance, and autonomousness of the emergent PDE model, as well as its interpretability.

generative models↗

DOE OSTI2024

A general mechanistic framework for cross-scale understanding of hot spots and hot moments in carbon and water fluxes

Semi-arid ecosystems, like those in the American Southwest, exert a massive impact on the interannual variability of carbon and water cycling. Unfortunately, these carbon and water fluxes are notoriously difficult to predict due to their high spatial and temporal variability, which is poorly captured by the current generation of vegetation models. Indeed, this region is exemplified by the ‘hot spots and hot moments’ concept, which states that small areas in space (‘hot spots’) and transient moments in time (‘hot moments’) exert an outsized influence on biogeochemical cycling. However, the factors that regulate these pulses in biogeochemical activity are unknown, as is their variability across space and time. These uncertainties severely limit efforts to better represent hot spots and hot moments in models. Here, we seek to develop a generalized method for detecting and quantifying the importance of hot spots and hot moments from individual plant to regional scales. Underpinning this method is our recently developed statistical approach for identifying hot spots and hot moments. By applying this method to semi-continuous measurements of plant water status, a depth profile of soil water potential, and ecosystem fluxes via eddy covariance, we will track the fate of water through the soil-plant-atmosphere continuum and identify the mechanistic drivers of these transient pulses in biogeochemical activity. Then, we will expand this approach across a broad network of Ameriflux towers, and apply a machine learning approach that will allow us to upscale measurements of hot spots and hot moments across the American Southwest and quantify their impact on carbon and water cycles. These products will allow us to identify hot spots and hot moments across spatio-temporal scales and will serve as crucial data sources for validating a new generation of models that can better capture highly dynamic carbon and water fluxes. The proposed method will be easily transferable across biomes and will serve as a framework for future research on hot spots and hot moments across the plant ecophysiology, biometeorology, and vegetation modeling communities.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2025

Full event particle-level unfolding with variable-length latent variational diffusion

The measurements performed by particle physics experiments must account for the imperfect response of the detectors used to observe the interactions. One approach, unfolding, statistically adjusts the experimental data for detector effects. Recently, generative machine learning models have shown promise for performing unbinned unfolding in a high number of dimensions. However, all current generative approaches are limited to unfolding a fixed set of observables, making them unable to perform full-event unfolding in the variable dimensional environment of collider data. A novel modification to the variational latent diffusion model (VLD) approach to generative unfolding is presented, which allows for unfolding of high- and variable-dimensional feature spaces. The performance of this method is evaluated in the context of semi-leptonic t\bar{t} t t ‾ production at the Large Hadron Collider.

Shmakov, Alexander↗

DOE OSTI2025

Effect of self-generated magnetic fields on x-ray emission in Kr-filled targets at the National Ignition Facility

We examine the effects of self-generated magnetic fields in a Kr gas pipe x-ray source platform. X-ray emission from Kr plasma is dependent on the plasma conditions, as the ionization state is largely a function of temperature. Magnetic fields are known to limit heat conduction, which increases temperature. We show that the emission in simulations of the gas pipe x-ray source is dependent on how self-generated magnetic fields are modeled. The inclusion of self-generated magnetic fields in simulations more accurately captures the emission of lower energy x-ray emission (L-shell), bringing results closer to experiments. The modeled x-ray emission and self-generated magnetic fields are shown to be particularly sensitive to the inclusion of the Nernst effect in simulations. Severely limiting the Nernst effect leads to a hotter Kr plasma, which can account for the discrepancy seen in earlier studies. By modifying the Nernst effect multiplier, we can achieve better experimental agreement in x-ray emission from gas pipes; the value of the multiplier that leads to the best agreement is dependent on the laser power of the drive. Currently, the suppression factor of the Nernst effect needed for high power drives (PL>200 TW) is more restrictive than what is currently put forward by non-local models.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2025

Hunga Tonga–Hunga Ha′apai Volcano Impact Model Observation Comparison (HTHH-MOC) project: experiment protocol and model descriptions

The 2022 Hunga volcanic eruption injected a significant amount of water vapor and a moderate amount of sulfur dioxide into the stratosphere, causing observable responses in the climate system. We have developed a model–observation comparison project to investigate the evolution of volcanic water and aerosols and their impacts on atmospheric dynamics, chemistry, and climate, using several state-of-the-art chemistry climate models. The project goals are (1) to evaluate the current chemistry–climate models to quantify their performance in comparison to observations and (2) to understand atmospheric responses in the Earth system after this exceptional event and investigate the potential impacts in the projected future. To achieve these goals, we designed specific experiments for direct comparisons to observations, for example from balloons and the Microwave Limb Sounder satellite instrument. Experiment 1 consists of two sets of free-running ensemble experiments from 2022 to 2031: one with fixed sea-surface temperatures and sea ice and one with coupled ocean. These experiments will help to understand the long-term evolution of water vapor and aerosols; quantify HTHH effects on stratospheric and mesospheric temperatures, dynamics, and transport; understand the impact of dynamic changes on ozone chemistry; quantify the net radiative forcings; and evaluate any surface climate impact. Experiment 2 is a nudged-run experiment from 2022 to 2023 using observed meteorology. To allow participation of more climate models with varying complexities of aerosol simulation, we include two sets of simulations in Experiment 2: Experiment 2a is designed for models with internally generated aerosol, while Experiment 2b is designed for models using prescribed aerosol surface area density. This experiment will help to analyze H 2 O and aerosol evolution, quantify the net radiative forcings, understand the impacts on mid-latitude and polar O 3 chemistry, and allow close comparisons with observations.

Zhu, Yunqian [Univ. of Colorado, Boulder, CO (Unit↗

DOE OSTI2024

Exploiting Multi-Domain Features for Detection of Unclassified Electromagnetic Signals

Deep Learning based classification techniques have shown excellent performance in static environments, where the training and testing samples are drawn from the same distribution. However, real world scenarios often present samples that do not belong to the known set of classes chosen during training. This is quite common for electromagnetic signals, where it is impractical to assume that all possible waveforms are known a-priori, specially in scenarios like warfare. To address this problem, we propose a deep learning based adversarial model where the generator learns to generate waveform features that can deceive the discriminator model as true samples. We introduce domain knowledge of wireless signals by decomposing the signal into a lower dimensional unique feature set, which is used for classifying known versus unknown signals. We further introduce multiple domain representations of the signal to extract features and combine them together to accurately classify new waveforms as an unknown class. Our results show that combined features from multiple domains outperform any single domain representation, especially at low SNR regimes with fewer number of samples to classify.

99 - GENERAL AND MISCELLANEOUS↗

DOE OSTI2024

Exploiting Multi-Domain Features for Detection of Unclassified Electromagnetic Signals (Presentation)

Deep Learning based classification techniques have shown excellent performance in static environments, where the training and testing samples are drawn from the same distribution. However, real world scenarios often present samples that do not belong to the known set of classes chosen during training. This is quite common for electromagnetic signals, where it is impractical to assume that all possible waveforms are known a-priori, specially in scenarios like warfare. To address this problem, we propose a deep learning based adversarial model where the generator learns to generate waveform features that can deceive the discriminator model as true samples. We introduce domain knowledge of wireless signals by decomposing the signal into a lower dimensional unique feature set, which is used for classifying known versus unknown signals. We further introduce multiple domain representations of the signal to extract features and combine them together to accurately classify new waveforms as an unknown class. Our results show that combined features from multiple domains outperform any single domain representation, especially at low SNR regimes with fewer number of samples to classify.

99 - GENERAL AND MISCELLANEOUS↗

DOE OSTI2024

Opportunities for Earth Observation to Inform Risk Management for Ocean Tipping Points

Abstract As climate change continues, the likelihood of passing critical thresholds or tipping points increases. Hence, there is a need to advance the science for detecting such thresholds. In this paper, we assess the needs and opportunities for Earth Observation (EO, here understood to refer to satellite observations) to inform society in responding to the risks associated with ten potential large-scale ocean tipping elements: Atlantic Meridional Overturning Circulation; Atlantic Subpolar Gyre; Beaufort Gyre; Arctic halocline; Kuroshio Large Meander; deoxygenation; phytoplankton; zooplankton; higher level ecosystems (including fisheries); and marine biodiversity. We review current scientific understanding and identify specific EO and related modelling needs for each of these tipping elements. We draw out some generic points that apply across several of the elements. These common points include the importance of maintaining long-term, consistent time series; the need to combine EO data consistently with in situ data types (including subsurface), for example through data assimilation; and the need to reduce or work with current mismatches in resolution (in both directions) between climate models and EO datasets. Our analysis shows that developing EO, modelling and prediction systems together, with understanding of the strengths and limitations of each, provides many promising paths towards monitoring and early warning systems for tipping, and towards the development of the next generation of climate models.

Wood, Richard A. (ORCID:0000000239609513)↗

DOE OSTI2026

Model Data Archive for Manuscript Titled "Evaluation of a Coupled Surface–Subsurface Hydrologic Model Using Dense Water‑Level Sensors in a Mixed Urban–Rural Watershed"

This archive provides scripts, input files, and datasets used for the implementation and evaluation of a fully coupled surface–subsurface hydrologic model in the Neches River Basin, southeast Texas. The study uses the Advanced Terrestrial Simulator (ATS) to simulate coupled surface–subsurface hydrologic processes over a mixed urban–rural watershed and evaluates model performance using a dense network of 136 in situ water-level sensors, nine U.S. Geological Survey (USGS) stream gauges, and SSEBop-derived evapotranspiration estimates during the period October 2014–June 2024. The workflow is implemented primarily in Python 3 using the Watershed Workflow package. The Jupyter notebooks can be executed using open-source software such as Anaconda JupyterLab or Visual Studio Code. Other data files include TXT, CSV, XML, SHP, TIF, NetCDF, HDF5, and ExodusII files, which can be processed using the provided Python scripts. ATS input files are provided in XML format and can be edited using any commonly used text editor. This archive contains: *Scripts and input files used to generate the ATS model setup, including watershed discretization, mesh generation, parameter mapping, and model configuration. *Jupyter notebooks used for preprocessing observational data, evaluating streamflow, water levels, and evapotranspiration, computing performance metrics, and generating the figures presented in the manuscript. *ATS simulation outputs and processed observational datasets, including OneRain and DD6 water-level sensors, USGS streamflow observations, GIS data, and supporting spatial datasets used throughout the study.

Dense water-level sensor network↗