Search NASA⌕ Search

SEARCH · Search NASA

Results for “Yield Modeling”

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 433 records · Page 24
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Techno-Economic Case Study: Low-Temperature Conversion Performance Based on Isolated Anatomical Fractions of Corn Stover

This report summarizes analysis conducted to support a case study under the Feedstock-Conversion Interface Consortium (FCIC) focused on techno-economic analysis (TEA) modeling to quantify the yield and cost ramifications for processing isolated anatomical fractions of corn stover through a low-temperature conversion (biochemical pathway) biorefinery. It is hypothesized that different individual anatomical fractions of corn stover vary in composition and recalcitrance, such that processing each fraction on its own e.g. through dedicated campaigns may enable economic benefits relative to processing the whole stover material. To investigate this concept, we worked with FCIC researchers to obtain data on compositional analysis, as well as processing conditions and yields, for conversion of whole corn stover plus three individual anatomical fractions (cobs, husks, and stalks) across key steps of the biorefinery conversion process within FCIC research scope, namely pretreatment and enzymatic hydrolysis. After running the data through the TEA models, fractionated cobs and husks achieved higher fuel yields and lower minimum fuel selling prices than whole corn stover, but stalks fared worse than whole stover on both metrics. The most direct takeaway from this assessment is that a biorefinery would stand to benefit economically from maximizing the use of cob and husk fractions while avoiding or minimizing the use of stalks, though recognizing practical constraints with this approach in maintaining an equivalent processing capacity (fixed at 2,000 dry metric tonne per day in all cases).

09 BIOMASS FUELS↗
DOE OSTI2025

Improving the Transportability of a Deep Learning Denoising Model Using Transfer Learning Techniques

The adoption of machine learning techniques in the seismology community has led to great performance improvements in several areas, including signal processing. Specifically, the development of deep learning–based seismic waveform denoising models has the potential to yield improvements in signal detection capabilities for networks operating in particularly noisy environments. Recent advancements in the design of these deep learning denoising models have included the incorporation of continuous and discrete wavelet transform functions into the network architecture to improve the learning capabilities and efficiency of said models. These wavelet transform–based seismic denoising models have shown improved denoising capabilities in regions where there is good agreement between the data features present in the training and evaluation datasets. However, questions remain about the overall transportability of these models to other monitoring regions. Here, in this study, we will determine the baseline transportability of a newly developed multilevel wavelet‐transform convolutional neural network (MWCNN) seismic denoising model. We accomplish this by taking a version of the MWCNN denoising model trained on data collected from the Utah region and evaluating its denoising performance on datasets collected from the neighboring Nevada region, which differ with regard to monitoring sensor types and event histories. We find that there is a notable variability in denoising performance related to the degree of similarity between the initial and new target datasets. The most notable difference in denoising performance is the ability of the denoising model to preserve accurate amplitude information associated with the signal energy present in the waveform data. Finally, we evaluate the ability of transfer learning techniques to improve the transportability of the MWCNN denoising model. We find that although there is still a performance gap present in the denoising results of the MWCNN model, transfer learning did yield improved results.

Quinones, Louis [Sandia National Laboratories (SNL↗
DOE OSTI2024

Automated Operational Forecasting of Monsoon Low Pressure Systems

Monsoon low pressure systems (LPSs) are the dominant rain-bearing weather system of South Asia, often producing extreme precipitation and hydrological disasters in a region inhabited by nearly two billion people. Despite the importance of these storms, no operational system has automatically identified and tracked LPS in real time in numerical weather prediction model output; many commonly used vortex-tracking algorithms are ill suited for monsoon LPS because of the weak winds and cold cores of these systems. Here, we describe a new system that uses optimized algorithms to identify monsoon LPS in short- to medium-range forecasts from the U.S. Global Ensemble Forecast System (GEFS) and a version of the deterministic Global Forecast System (GFS) adapted and used operationally by the Indian Institute of Tropical Meteorology (IITM). We also assess the historical performance of these models in forecasting South Asian monsoon LPS, comparing this with the performance of the Integrated Forecasting System of the ECMWF. We assess the accuracy of model predictions of LPS genesis, position, intensity, and precipitation rates for forecast lead times of 1–5 days, yielding quantitative information on model biases to guide operational forecasters and disaster managers. The system we introduce here could be extended to other low-latitude regions affected by dynamically weak, heavily precipitating atmospheric vortices that are often not included in tropical cyclone inventories.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

$\mathcal{g}$ modes of neutron stars with hadron-to-quark crossover transitions

Here we perform the first study of the principal core g-mode oscillation in hybrid stars containing quark matter, utilizing a crossover model for the hadron-to-quark transition inspired by lattice QCD. The ensuing results are compared with our recent findings of g-mode frequencies in hybrid stars with a first-order phase transition using Gibbs constructions. We find that models using Gibbs construction yield g-mode amplitudes and the associated gravitational energy radiated that dominate over those of the chosen crossover model owing to the distinct behaviors of the equilibrium and adiabatic sound speeds in the various models. Based on our results, we conclude that were g modes to be detected in upgraded LIGO and Virgo detectors it would indicate a first-order phase transition akin to a Gibbs construction.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2021

Climate impacts on global agriculture emerge earlier in new generation of climate and crop models

Potential climate-related impacts on future crop yield are a major societal concern. Previous projections of the Agricultural Model Intercomparison and Improvement Project’s Global Gridded Crop Model Intercomparison based on the Coupled Model Intercomparison Project Phase 5 identified substantial climate impacts on all major crops, but associated uncertainties were substantial. Here we report new twenty-first-century projections using ensembles of latest-generation crop and climate models. Results suggest markedly more pessimistic yield responses for maize, soybean and rice compared to the original ensemble. Mean end-of-century maize productivity is shifted from +5% to -6% (SSP126) and from +1% to -24% (SSP585)—explained by warmer climate projections and improved crop model sensitivities. In contrast, wheat shows stronger gains (+9% shifted to +18%, SSP585), linked to higher CO2 concentrations and expanded high-latitude gains. The ‘emergence’ of climate impacts consistently occurs earlier in the new projections—before 2040 for several main producing regions. While future yield estimates remain uncertain, these results suggest that major breadbasket regions will face distinct anthropogenic climatic risks sooner than previously anticipated.

Jagermeyr, Jonas↗
DOE OSTI2020

System-Level Optimization to Improve Biofuel Potential via Genetic Engineering and Hydrothermal Liquefaction

The economic success of biofuels and bioproducts depends on system-level optimization including biomass production and conversion. Hydrothermal liquefaction (HTL) can convert wet biomass such as microalgae into a biofuel intermediate (BFI) under elevated temperatures and pressure. An understanding of the impacts of biomass composition on BFI yield and quality can inform genetic engineering strategies in the improvement of biochemical composition for biofuel production. In this work, wild type cyanobacterium Synechocystis sp. PCC 6803 biomass was doped with various common cellular storage compounds in lab-scale HTL experiments. Doping with glycogen or polyhydroxybutyrate (PHB) significantly reduced BFI yields, while doping with triglycerides (TAG) or medium chain-length polyhydroxyalkanoate (mcl-PHA) increased BFI yield and quality. In light of these observations, a genetically engineered Synechocystis strain deficient in glycogen biosynthesis was cultivated to produce biomass for HTL, leading to a 17% increase in BFI yield. In addition, we built a multiphase component additivity (MCA) model that can predict BFI yield and quality with PHAs in the biomass. This work demonstrates an effective strategy to integrate strain development with downstream biomass conversion to maximize biofuel yield, with lessons applicable to microalgae as well as other biomass.

09 BIOMASS FUELS↗
DOE OSTI2022

Comparing Model Representations of Physiological Limits on Transpiration at a Semi-arid Ponderosa Pine Site

Mechanistic representations of biogeochemical processes in ecosystem models are rapidly advancing, requiring advancements in model evaluation approaches. Here we quantify multiple aspects of model functional performance to evaluate improved process representations in ecosystem models. We compare semi-empirical stomatal models with hydraulic constraints against more mechanistic representations of stomatal and hydraulic functioning at a semi-arid pine site using a suite of metrics and analytical tools. We find that models generally perform similarly under unstressed conditions, but performance diverges under atmospheric and soil drought. The more empirical models better capture synergistic information flows between soil water potential and vapor pressure deficit to transpiration, while the more mechanistic models are overly deterministic. Although models can be parameterized to yield similar functional performance, alternate parameterizations could not overcome structural model constraints that underestimate the unique information contained in soil water potential about transpiration. Additionally, both multilayer canopy and big-leaf models were unable to capture the magnitude of canopy temperature divergence from air temperature, and we demonstrate that errors in leaf temperature can propagate to considerable error in simulated transpiration. This study demonstrates the value of merging underutilized observational data streams with emerging analytical tools to characterize ecosystem function and discriminate among model process representations.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Novel Microbial Routes to Synthesize Industrially Significant Precursor Compounds

Ethylene is the most widely employed organic precursor compound in industry. The potential to impact ethylene formation via recently discovered microbial processes is tenable using plentiful CO2 feedstocks. The overall long-term objective of this project was to develop an industrially compatible microbial process to synthesize ethylene in high yields. The key objective of this project was to fully define and initially characterized a recently discovered and genetically regulated anaerobic pathway to produce high levels of ethylene called the Dihydroxyacetone Phosphate - Ethylene Pathway in phototrophic bacteria. This was addressed through the following specific aims: 1. Fully probe the catalytic potential of all enzymes of the DHAP ethylene pathway and determine the regulatory mechanism of DHAP-ethylene pathway gene expression. 2. Discover effective and active ethylene enzymes encoded in cultured and uncultured organisms from anoxic environments. 3.Model the thermodynamics and kinetics of ethylene synthetic pathways to guide engineering efforts in integrating best performing DHAP-ethylene pathway enzymes into model bacteria chassis for enhance ethylene yields. Through this project we discovered the initially missing genetic and enzyme component of the DHAP-ethylene pathway that directly synthesized ethylene and other important industrial compounds like methane and ethane from specific substrates. We uncovered and partially characterized a nitrogenase-like reductase that functions in DHAP-ethylene pathway specifically and in methionine synthesis in general. This nitrogenase-like system is called the Methylthio-Alkane Reductase (MAR) for its ability to cleave volatile organic sulfur compounds into methanethiol (CH3-SH) for methionine synthesis and a hydrocarbon byproduct. Key to the DHAP-ethylene pathway, MAR is the essential enzyme that cleaves 2-methylthioethanol (CH3-S-CH2-CH2-OH) into ethylene. Coordinately, we uncovered that the MAR genes and genes associated with conversion of methanethiol (CH3-SH) to methionine are under genetic control of a LysR Type Transcriptional Regulator called SalR, whose activity is dependent upon the amount of sulfate available to the cell. When sulfate as the preferred sulfur source for cell growth drops below 200 micromolar, SalR become active for expressing the MAR and methionine biosynthesis genes to enable the cell to grow from volatile organic sulfur compounds and make ethylene. Metabolic thermos-kinetic modeling revealed that these MAR reactions for ethylene and other hydrocarbon production are highly thermodynamically favorable and are one of the largest driving forces for ethylene production by the DHAP-ethylene pathway for high ethylene yields. Modeling also indicated that a key aldolase and to a lesser extent an isomerase of the DHAP-ethylene pathway for production of the ethylene precursor, 2-methylthioethanol, also would increase ethylene yields. Through metagenomic mining and gene synthesis by the JGI DNA synthesis program, over 500 aldolase and isomerase homologs were synthesized and screened. From this, variants were uncovered with substantially higher activity that increased ethylene yields 5-fold via the aldolase reaction and 1.5-fold via the isomerase reaction. Each of these elements that increase ethylene production were integrated together via plasmid under appropriate gene promoter elements in the phototrophic bacterium, Rhodospirillum rubrum, resulting in at least 3 orders of magnitude increase in ethylene yield from carbon dioxide feedstock.

10 SYNTHETIC FUELS↗
DOE OSTI2023

Progenitors and explosion properties of supernova remnants hosting central compact objects: II. A global systematic study with a comparison to nucleosynthesis models

ABSTRACT Core-collapse explosions of massive stars leave behind neutron stars, with a known diversity that includes the ‘Central Compact Objects’ (CCOs). Typified by the neutron star discovered near the centre of the Cas A supernova remnant (SNR), CCOs have been observed to shine only in X-rays. To address their supernova progenitors, we perform a systematic study of SNRs that contain a CCO and display X-ray emission from their shock-heated ejecta. We make use of X-ray data primarily using the Chandra X-ray observatory, complemented with XMM–Newton. This study uses a systematic approach to the analysis of each SNR aimed at addressing the supernova progenitor as well as the explosion properties (energy and ambient density). After fitting for the ejecta abundances estimated from a spatially resolved spectroscopic study, we compare the data to six nucleosynthesis models making predictions on supernova ejecta yields in core-collapse explosions. We find that the explosion models commonly used by the astrophysics community do not match the ejecta yields for any of the SNRs, suggesting additional physics, for example multidimensional explosion models or updated progenitor structures, are required. Overall we find low-mass (≤25 solar masses) progenitors among the massive stars population and low-energy explosions (<1051 ergs). We discuss degeneracies in our model fitting, particularly how altering the explosion energy affects the estimate of the progenitor mass. Our systematic study highlights the need for improving on the theoretical models for nucleosynthesis predictions as well as for sensitive, high-resolution spectroscopy observations to be acquired with next generation X-ray missions.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

Stopping Force Analysis of 235 U Elemental Fission Product Yields for $E_n$ = 0.11–92.4 MeV

We report most evaluated elemental fission product yield distributions are not experimentally measured. Instead, the majority of evaluated distributions are based on analytic expressions of the Zp-model for relevant cumulative yields. Here we report independent elemental fission product yield distributions of a 235 U target for incident neutron energies ranging from 0.11 MeV through 92.4 MeV. Atomic numbers are calculated by an approach that combines a 2E analysis with a stopping force analysis method, developed within this paper. These analyses are applied to more than 6.1 × 10 6 fission fragment ionization tracks captured within the NIFFTE (Neutron Induced Fission Fragment Tracking Experiment) collaboration fission time projection chamber (fissionTPC). A 3-Z resolution was obtained with the fissionTPC spatial and energy resolutions. Tabulated results are presented for the atomic yield and experimentally derived Zp values as a function of pre-neutron-emission fragment masses for the complete range of incident neutron energies. The stopping-force-derived Zp values tend to support the unchanged charge distribution theory within uncertainty.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2023

Estimating Field-Level Perennial Bioenergy Grass Biomass Yields Using the Normalized Difference Red-Edge Index and Linear Regression Analysis for Central Virginia, USA

We investigated the indicative power of the normalized difference red-edge index (NDRE) for estimating field-level perennial bioenergy grass biomass yields utilizing Sentinel-2 imagery and a linear regression model as a rapid, cost-effective method for biomass yield estimations for bioenergy. We used 2019 data from three study sites containing mature perennial bioenergy grass stands in central Virginia, USA. Of the simulated daily NDRE values based on the temporally weighted averaging of two temporal neighbors, we found the strongest index–yield correlation on 11 August (R = 0.85). We estimated the perennial bioenergy grass biomass yields for (1) all sites using the data pooled from the three sites (all-site estimation) and (2) each site using the data pooled from the other two sites (cross-site estimation). The estimated field-level perennial bioenergy grass biomass yields strongly correlated with the recorded yields (average R2 = 0.76), with a root mean square error (RMSE) of 1.5 Mg/ha and a mean absolute error (MAE) of 1.2 Mg/ha for the all-site estimation. For the cross-site estimation, the site with diverse perennial grass types had the weakest correlation (R2 = 0.44) of the sites, indicating a difficulty in accounting for heterogeneous index–yield relationships in a single model. In addition to identifying a strong indicative power of the NDRE for estimating the overall perennial bioenergy grass biomass yields at a field level, the findings from this study call for an analysis across multiple perennial grasses and a comparison using multiple sites to understand (1) if the indicative power of the index shifts from the biomass of the specific perennial bioenergy grass type to the overall biomass during the growing season and (2) the level of perennial bioenergy grass heterogeneity that may hinder the remotely sensed biomass yield estimation using a single model.

09 BIOMASS FUELS↗
DOE OSTI2020

In Situ Micro-Pillar Compression to Examine Radiation-Induced Hardening Mechanisms of FeCrAl Alloys

Here, the effects of 5 MeV Fe 2+ ion irradiation at 300°C on the microstructure evolution and deformation behavior of a FeCrAl C26M alloy are presented. It has been found that dislocation loop density increases an order of magnitude from 1 dpa to 16 dpa irradiations, whereas, the dislocation loop size saturates with increasing damage. Micropillars, 600 nm in diameter and 1.3 µm in height, were fabricated and compressed inside grains with <001>, <011> and <111> crystallographic orientations, respectively. {112} <111> has been identified as the primary slip system in both unirradiated and irradiated alloy. The increase in yield stress after irradiation is observed with measurable variation along <001> and <011> vs. along <111>. By applying the Orowan dispersed barrier model, the increase of yield stress is found mainly due to the slip resistance of radiation generated defect loops. Detailed transmission electron microscopy (TEM) studies were performed to quantify the Burgers vector and the distribution of irradiation induced dislocations at elevated strains. It is revealed that localized shear instability is caused by avalanche slip events of ½<111> dislocations gliding out of tested pillars. Simultaneously, a large number of sessile/immobile <100> dislocations formed in the vicinity of slip band, leading to the hardening at elevated strains.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Identification of crystal plasticity model parameters by multi-objective optimization integrating microstructural evolution and mechanical data

Crystal plasticity models evolve a polycrystalline yield surface using meso-scale descriptions of deformation mechanisms. The activation of deformation mechanisms is governed by crystallography and a set of model parameters, which are typically calibrated through the fitting of mechanical data such as stress–strain curves and elastic lattice strains. Microstructural data such as phase fractions and texture evolution are used for verifying crystal plasticity parameters. In this study, we use a multi-objective genetic algorithm to identify hardening parameters from flow stress curves with an option to incorporate texture into the optimization approach. Robust, generalized objective functions are developed and used to identify sets of parameters pertaining to dislocation density-based hardening laws in visco-plastic and elasto-plastic self-consistent (VPSC and EPSC) homogenization models. First, the parameters are identified for pure Nb directly from texture using an objective function based on generalized spherical harmonics. Since texture evolution is driven by the relative contribution of active slip systems, the parameters governing the evolution of slip resistance ratios can be recovered from fitting discrete textures at a series of strains. Next, a comprehensive set of load reversal data for dual phase (DP) 780 steel is used to fit a hardening law and a back-stress law in EPSC. Finally, parameters pertaining to a complex hardening law for the evolution of slip and twinning in pure α-Ti are identified. Remarkably, using texture as an objective in combination with stress–strain objectives constrains the model of Ti to fully reproduce not only stress–strain and texture evolution but also hierarchical twinning measurements as a function of initial grain size and texture. Furthermore, given an appropriate model fit to representative experimental texture evolution, underlying twin volume fractions contributing to texture evolution can be predicted.

42 ENGINEERING↗
DOE OSTI2022

Transfer learning driven design optimization for inertial confinement fusion

Transfer learning is a promising approach to create predictive models that incorporate simulation and experimental data into a common framework. In this technique, a neural network is first trained on a large database of simulations and then partially retrained on sparse sets of experimental data to adjust predictions to be more consistent with reality. Previously, this technique has been used to create predictive models of Omega [Humbird et al., IEEE Trans. Plasma Sci. 48, 61–70 (2019)] and NIF [Humbird et al., Phys. Plasmas 28, 042709 (2021); Kustowski et al., Mach. Learn. 3, 015035 (2022)] inertial confinement fusion (ICF) experiments that are more accurate than simulations alone. Here, in this work, we conduct a transfer learning driven hypothetical ICF campaign in which the goal is to maximize experimental neutron yield via Bayesian optimization. The transfer learning model achieves yields within 5% of the maximum achievable yield in a modest-sized design space in fewer than 20 experiments. Furthermore, we demonstrate that this method is more efficient at optimizing designs than traditional model calibration techniques commonly employed in ICF design. Such an approach to ICF design could enable robust optimization of experimental performance under uncertainty.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2021

Different to the core: The pre-supernova structures of massive single and binary-stripped stars

The majority of massive stars live in binary or multiple systems and will interact with a companion during their lifetimes, which helps to explain the observed diversity of core-collapse supernovae. Donor stars in binary systems can lose most of their hydrogen-rich envelopes through mass transfer. As a result, not only are the surface properties affected, but so is the core structure. However, most calculations of the core-collapse properties of massive stars rely on single-star models. We present a systematic study of the difference between the pre-supernova structures of single stars and stars of the same initial mass (11–21 M ⊙ ) that have been stripped due to stable post-main-sequence mass transfer at solar metallicity. We present the pre-supernova core composition with novel diagrams that give an intuitive representation of the isotope distribution. As shown in previous studies, at the edge of the carbon-oxygen core, the binary-stripped star models contain an extended gradient of carbon, oxygen, and neon. This layer remains until core collapse and is more extended in mass for higher initial stellar masses. It originates from the receding of the convective helium core during core helium burning in binary-stripped stars, which does not occur in single-star models. We find that this same evolutionary phase leads to systematic differences in the final density and nuclear energy generation profiles. Binary-stripped star models have systematically higher total masses of carbon at the moment of core collapse compared to single-star models, which likely results in systematically different supernova yields. In about half of our models, the silicon-burning and oxygen-rich layers merge after core silicon burning. We discuss the implications of our findings for the “explodability”, supernova observations, and nucleosynthesis of these stars. Our models are publicly available and can be readily used as input for detailed supernova simulations.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2020

Bose-Einstein condensation phenomenology in systems with repulsive interactions

The role of repulsive interactions in statistical systems of Bose particles is investigated. Three different phenomenological frameworks are considered: A mean-field model, an excluded volume model, and a model with a medium-dependent effective mass. All three models are tuned to yield similar equations of state, with only minor deviations from the ideal Bose gas at small chemical potentials. Our analysis indicates, however, that these models lead to qualitatively different results for the Bose-Einstein condensation phenomenon. We discuss the different aspects of this phenomenon, namely, an onset of the Bose-Einstein condensation, particle number fluctuations, and a behavior of the Bose condensate. The obtained results can be helpful for interpreting the lattice QCD data at small temperature and large isospin chemical potential and the data on multiple pion production in high-energy nuclear collisions.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2025

Simulating Muon Capture with MARLEY

Core-collapse supernovae are intense sources of tens-of-MeV neutrinos. Each explosion releases about 1058 neutrinos in a burst lasting tens of seconds. As neutrinos are weakly interacting, they escape more easily compared to photons, allowing them to be observed before the light signal. Understanding low-energy neutrino interactions with argon will be critical for future supernova measurements with the Deep Underground Neutrino Experiment (DUNE). Interpreting future DUNE supernova neutrino data will require a high-quality model of O(10 MeV) neutrino interactions, but the dominant charged-current cross section has never been measured at these low energies for an argon target. We exploit the theoretical similarities between muon capture, which has been measured, and antineutrino scattering to test MARLEY (Model of Argon Reaction Low Energy Yields), the simulation code used by DUNE to model supernova neutrino reactions. I implemented C++ code to add a muon capture mode to MARLEY and compared the predictions of the new simulation to measured isotope branching ratios of the residual nuclei. The preliminary branching ratios predicted by MARLEY running in muon capture mode demonstrate promising baseline performance when compared to experimental data and are expected to improve with further model adjustments.

Wong, Baker Louie↗
DOE OSTI2024

Bridging Hydrological Ensemble Simulation and Learning Using Deep Neural Operators

Ensemble-based simulation and learning (ESnL) has long been used in hydrology for parameter inference, but computational demands of process-based ESnL can be quite high. To address this issue, we propose a deep neural operator learning approach. Neural operators are generic machine learning algorithms that can learn functional mappings between infinite-dimensional spaces, providing a highly flexible tool for scientific machine learning. Our approach is built upon DeepONet, a specific deep neural operator, and is designed to address several common problems in hydrology, namely, model parameter estimation, prediction at ungaged locations, and uncertainty quantification. Here we demonstrate the effectiveness of our DeepONet-based workflow using an existing large model ensemble created for an eastern U.S. watershed that is instrumented with 10 streamflow gages. Results suggest DeepONet achieves high efficiency in learning an ML surrogate model from the model ensemble, with the modified Kling-Gupta Efficiency exceeding 0.9 on holdout test sets. Parameter inference, carried out using the trained DeepONet surrogate model and genetic algorithm, also yields robust results. Additionally, we formulate and train a separate DeepONet model for physics-informed, seq-to-seq streamflow forecasting, which further reduces biases in the pre-trained DeepONet surrogate model. While this study focuses primarily on a single watershed, our approach is general and may be extended to enable learning from model ensembles across multiple basins or models. Thus, this research represents a significant contribution to the application of hybrid machine learning in hydrology.

54 ENVIRONMENTAL SCIENCES↗