Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “ensemble forecast”

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 163 records · Page 9
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Subsetted model output on height-above-mean-sea-level levels for cloud-related variables from a LASSO-CACTI WRF simulation

Cloud-related model output interpolated to heights-above-mean-sea-level from a Weather Research and Forecasting (WRF) simulation from the Large-Eddy Simulation (LES) Atmospheric Radiation Measurement (ARM) Symbiotic Simulation and Observation (LASSO) deep-convection scenario for the Cloud, Aerosol, and Complex Terrain Interactions (CACTI) field campaign. The LASSO-CACTI simulations span grid spacings from 7.5 km to 100 m through the use of four nested domains, labeled d1 through d4. The simulations are of convection near the Sierras de Córdoba mountain range, roughly centered on the ARM Mobile Facility. More information can be found at https://www.arm.gov/capabilities/modeling/lasso. This version of the output is a collection of variables subsetted from the raw WRF output to reduce the file overhead for users not needing the full raw dataset. The subset files are in netCDF format with the height-based variables interpolated to heights based on distance above mean sea level.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Subsetted model output on height-above-mean-sea-level levels for meteorology-related variables from a LASSO-CACTI WRF simulation

Meteorology-related model output interpolated to heights-above-mean-sea-level from a Weather Research and Forecasting (WRF) simulation from the Large-Eddy Simulation (LES) Atmospheric Radiation Measurement (ARM) Symbiotic Simulation and Observation (LASSO) deep-convection scenario for the Cloud, Aerosol, and Complex Terrain Interactions (CACTI) field campaign. The LASSO-CACTI simulations span grid spacings from 7.5 km to 100 m through the use of four nested domains, labeled d1 through d4. The simulations are of convection near the Sierras de Córdoba mountain range, roughly centered on the ARM Mobile Facility. More information can be found at https://www.arm.gov/capabilities/modeling/lasso. This version of the output is a collection of variables subsetted from the raw WRF output to reduce the file overhead for users not needing the full raw dataset. The subset files are in netCDF format with the height-based variables interpolated to heights based on distance above mean sea level.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Urban Flood Modeling: Uncertainty Quantification and Physics‐Informed Gaussian Processes Regression Forecasting

Abstract Estimating uncertainty in flood model predictions is important for many applications, including risk assessment and flood forecasting. We focus on uncertainty in physics‐based urban flooding models. We consider the effects of the model's complexity and uncertainty in key input parameters. The effect of rainfall intensity on the uncertainty in water depth predictions is also studied. As a test study, we choose the Interconnected Channel and Pond Routing (ICPR) model of a part of the city of Minneapolis. The uncertainty in the ICPR model's predictions of the floodwater depth is quantified in terms of the ensemble variance using the multilevel Monte Carlo (MC) simulation method. Our results show that uncertainties in the studied domain are highly localized. Model simplifications, such as disregarding the groundwater flow, lead to overly confident predictions, that is, predictions that are both less accurate and uncertain than those of the more complex model. We find that for the same number of uncertain parameters, increasing the model resolution reduces uncertainty in the model predictions (and increases the MC method's computational cost). We employ the multilevel MC method to reduce the cost of estimating uncertainty in a high‐resolution ICPR model. Finally, we use the ensemble estimates of the mean and covariance of the flood depth for real‐time flood depth forecasting using the physics‐informed Gaussian process regression method. We show that even with few measurements, the proposed framework results in a more accurate forecast than that provided by the mean prediction of the ICPR model.

Kohanpur, Amir H.↗
DOE OSTI2023

Skillful Multi‐Month Predictions of Ecosystem Stressors in the Surface and Subsurface Ocean

Abstract Anthropogenic carbon emissions and associated climate change are driving rapid warming, acidification, and deoxygenation in the ocean, which increasingly stress marine ecosystems. On top of long‐term trends, short term variability of marine stressors can have major implications for marine ecosystems and their management. As such, there is a growing need for predictions of marine ecosystem stressors on monthly, seasonal, and multi‐month timescales. Previous studies have demonstrated the ability to make reliable predictions of the surface ocean physical and biogeochemical state months to years in advance, but few studies have investigated forecast skill of multiple stressors simultaneously or assessed the forecast skill below the surface. Here, we use the Community Earth System Model (CESM) Seasonal to Multiyear Large Ensemble (SMYLE) along with novel observation‐based biogeochemical and physical products to quantify the predictive skill of dissolved inorganic carbon (DIC), dissolved oxygen, and temperature in the surface and subsurface ocean. CESM SMYLE demonstrates high physical and biogeochemical predictive skill multiple months in advance in key oceanic regions and frequently outperforms persistence forecasts. We find up to 10 months of skillful forecasts, with particularly high skill in the Northeast Pacific (Gulf of Alaska and California Current Large Marine Ecosystems) for temperature, surface DIC, and subsurface oxygen. Our findings suggest that dynamical marine ecosystem prediction could support actionable advice for decision making.

Environmental Sciences & Ecology↗
DOE OSTI2023

Efficient Probabilistic Prediction and Uncertainty Quantification of Tropical Cyclone–Driven Storm Tides and Inundation

Abstract This study proposes and assesses a methodology to obtain high-quality probabilistic predictions and uncertainty information of near-landfall tropical cyclone–driven (TC-driven) storm tide and inundation with limited time and resources. Forecasts of TC track, intensity, and size are perturbed according to quasi-random Korobov sequences of historical forecast errors with assumed Gaussian and uniform statistical distributions. These perturbations are run in an ensemble of hydrodynamic storm tide model simulations. The resulting set of maximum water surface elevations are dimensionality reduced using Karhunen–Loève expansions and then used as a training set to develop a polynomial chaos (PC) surrogate model from which global sensitivities and probabilistic predictions can be extracted. The maximum water surface elevation is extrapolated over dry points incorporating energy head loss with distance to properly train the surrogate for predicting inundation. We find that the surrogate constructed with third-order PCs using elastic net penalized regression with leave-one-out cross validation provides the most robust fit across training and test sets. Probabilistic predictions of maximum water surface elevation and inundation area by the surrogate model at 48-h lead time for three past U.S. landfalling hurricanes (Irma in 2017, Florence in 2018, and Laura in 2020) are found to be reliable when compared to best track hindcast simulation results, even when trained with as few as 19 samples. The maximum water surface elevation is most sensitive to perpendicular track-offset errors for all three storms. Laura is also highly sensitive to storm size and has the least reliable prediction. Significance Statement The purpose of this study is to develop and evaluate a methodology that can be used to provide high-quality probabilistic predictions of hurricane-induced storm tide and inundation with limited time and resources. This is important for emergency management purposes during or after the landfall of hurricanes. Our results show that sampling forecast errors using quasi-random sequences combined with machine learning techniques that fit polynomial functions to the data are well suited to this task. The polynomial functions also have the benefit of producing exact sensitivity indices of storm tide and inundation to the forecasted hurricane properties such as path, intensity, and size, which can be used for uncertainty estimation. The code implementing the presented methodology is publicly available on GitHub.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Subsetted model output for static variables from a LASSO-CACTI WRF simulation

Static model output from a Weather Research and Forecasting (WRF) simulation from the Large-Eddy Simulation (LES) Atmospheric Radiation Measurement (ARM) Symbiotic Simulation and Observation (LASSO) deep-convection scenario for the Cloud, Aerosol, and Complex Terrain Interactions (CACTI) field campaign. The LASSO-CACTI simulations span grid spacings from 7.5 km to 100 m through the use of four nested domains, labeled d1 through d4. The simulations are of convection near the Sierras de Córdoba mountain range, roughly centered on the ARM Mobile Facility. More information can be found at https://www.arm.gov/capabilities/modeling/lasso. This version of the output is a collection of variables subsetted from the raw WRF output to reduce the file overhead for users not needing the full raw dataset. The subset files are in netCDF format with the height-based variables on the raw, model eta and soil levels. The variables in this file are ones that do not change during a simulation, such as the terrain height and map-scale factors.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Subsetted model output for aerosol-related variables from a LASSO-CACTI WRF simulation

Aerosol-related model output from a Weather Research and Forecasting (WRF) simulation from the Large-Eddy Simulation (LES) Atmospheric Radiation Measurement (ARM) Symbiotic Simulation and Observation (LASSO) deep-convection scenario for the Cloud, Aerosol, and Complex Terrain Interactions (CACTI) field campaign. The LASSO-CACTI simulations span grid spacings from 7.5 km to 100 m through the use of four nested domains, labeled d1 through d4. The simulations are of convection near the Sierras de Córdoba mountain range, roughly centered on the ARM Mobile Facility. More information can be found at https://www.arm.gov/capabilities/modeling/lasso. This version of the output is a collection of variables subsetted from the raw WRF output to reduce the file overhead for users not needing the full raw dataset. The subset files are in netCDF format with the height-based variables on the raw, model eta levels. The aerosol properties are consistent with the Thompson aerosol-aware cloud microphysics scheme where the simulated aerosol are proxies for representative “water friendly” and “ice friendly” aerosol.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Analyzing and Exploring Training Recipes for Large-Scale Transformer-Based Weather Prediction

Abstract The rapid rise of deep learning (DL) in numerical weather prediction (NWP) has led to a proliferation of models which forecast atmospheric variables with comparable or superior skill than traditional physics-based NWP. However, among these leading DL models, there is a wide variance in both the training settings and architecture used. Further, the lack of thorough ablation studies makes it hard to discern which components are most critical to success. In this work, we show that it is possible to attain high forecast skill even with relatively off-the-shelf architectures, simple training procedures, and moderate compute budgets. Specifically, we train a minimally modified Swin Transformer V2 (SwinV2) on ERA5 data and find that it attains superior skill in terms of mean-square errors of deterministic forecasts when compared against the European Centre for Medium-Range Weather Forecasts’ Integrated Forecasting System (IFS). Almost all DL–NWP systems share a core set of hyperparameters and design decisions. To aid and expedite future DL–NWP research, we present an in-depth, systematic exploration of different loss functions, model sizes and depths, patch sizes, and multistep training objectives. We also examine the model performance with metrics beyond the typical accuracy (ACC) and RMSE and investigate how the performance scales with model size. Through our open-source code, scoring pipelines, and models, we share our findings on key aspects of the training pipeline. These ablations reduce the necessity for expensive hyperparameter tuning and lower the barrier to entry for future DL–NWP research. Significance Statement This study investigates the potential of using large-scale transformer-based models for weather prediction, showing that it is possible to achieve high forecast accuracy with simpler, off-the-shelf architectures. By training a minimally modified SwinV2 transformer on ERA5 data, we show that the model achieves competitive forecast skill in terms of mean-square error for key variables, outperforming the European Centre for Medium-Range Weather Forecasts’ Integrated Forecasting System (IFS) at all lead times. Our findings suggest that effective training strategies, such as multistep fine-tuning and channel-weighted losses, significantly enhance the model’s performance. However, we also highlight that these improvements come with trade-offs in other areas, such as ensemble spread and high-frequency spatial detail. This work highlights the promise of deep learning in improving weather forecasts, which could lead to better preparedness and response to weather events, ultimately benefiting society by providing more reliable weather predictions.

Willard, Jared D. [Lawrence Berkeley National Labo↗
DOE OSTI2024

Reduced‐Order Probabilistic Emulation of Physics‐Based Ring Current Models: Application to RAM‐SCB Particle Flux

Abstract In this work, we address the computational challenge of large‐scale physics‐based simulation models for the ring current. Reduced computational cost allows for significantly faster than real‐time forecasting, enhancing our ability to predict and respond to dynamic changes in the ring current, valuable for space weather monitoring and mitigation efforts. Additionally, it can also be used for a comprehensive investigation of the system. Thus, we aim to create an emulator for the Ring current‐Atmosphere interactions Model with Self‐Consistent magnetic field (RAM‐SCB) particle flux that not only improves efficiency but also facilitates forecasting with reliable estimates of prediction uncertainties. The probabilistic emulator is built upon the methodology developed by Licata and Mehta (2023), https://doi.org/10.1029/2022sw003345 . A novel discrete sampling is used to identify 30 simulation periods over 20 years of solar and geomagnetic activity. Focusing on a subset of particle flux, we use Principal Component Analysis for dimensionality reduction and Long Short‐Term Memory (LSTM) neural networks to perform dynamic modeling. Hyperparameter space was explored extensively resulting in about 5% median symmetric accuracy across all data sets for one‐step dynamic prediction. Using a hierarchical ensemble of LSTMs, we have developed a reduced‐order probabilistic emulator (ROPE) tailored for time‐series forecasting of particle flux in the ring current. This ROPE offers accurate predictions of omnidirectional flux at a single energy with no pitch angle information, providing robust predictions on the test set with an error score below 11% and calibration scores under 8% with bias under 2% providing a significant speed up as compared to the full RAM‐SCB run.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

Convective Updraft and Downdraft Characteristics of Continental Mesoscale Convective Systems in the Model Gray Zone

Here, the "gray zone" of convective modeling is defined as the range of horizontal grid spacings (Δx) at which turbulent transport processes are only partially resolved by the dynamics of the numerical model. This zone typically covers Δx from a few kilometers to several hundred meters, wherein the realistic representation of convective cloud processes can be challenging. This study characterizes the convective draft behaviors at multiple Δx across the gray zone and determines the appropriate Δx that can reliably capture these salient convective properties. We perform an ensemble of idealized simulations of mesoscale convective systems (MCS) using the Weather Research and Forecasting model at various Δx from 4 km to 250 m over the central U.S. An evaluation of key MCS kinematic properties is constrained using unique, long-term vertical velocity estimates obtained by radar wind profilers deployed by the Department of Energy Atmospheric Radiation Measurement user facility. MCS simulations for all Δx tested overestimate (underestimate) the probabilities of convective updrafts (downdrafts) compared to the observations. In terms of the convective draft intensity, finer-Δx models overestimate the updraft intensity, while the opposite is found for downdrafts. Moving from Δx = 4 km to 250 m, downdrafts become stronger and more frequent especially at middle and upper levels, attributed to additional drag from an increasing graupel frequency and compensation of enhanced updraft velocity. Simulated draft characteristics, including core size, intensity, and probability of occurrence, exhibit pronounced changes at Δx greater than 500 m, but suggest less sensitivity when Δx is reduced below 500 m.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

A Data-Agnostic, Continuous Machine Learning Framework for Application in High Energy Physics and Beyond: Phase 1 Final Scientific/Technical Report

This Phase 1 effort has focused on the development of continual learning frameworks for use in machine learning, specifically in the applied context of High Energy Physics (HEP). Machine learning (ML) is a transformative technology by which computers, typically through the use of neural networks, are able to perform tasks with proficiency that rivals or surpasses that of human users. Model Degradation & Catastrophic Forgetting are two undesired phenomena which can occur in ML where the performance of a model degrades when either deployed on novel data streams, or trained on novel data which are sufficiently different than the data the models were initially trained on. A natural example where these sorts of effects can be observed is in the performance of detectors in harsh environments, where the detector signature may change over the lifetime of the detector as it ages and deteriorates — precisely what occurs in the experiments conducted in HEP. Real world HEP data is therefore an excellent test-ground and use-case for Continual Learning paradigms, which are techniques used in ML to counteract these problems. Ensemble learning is one such technique, where multiple smaller models are trained on subsets of the overall data and are ensembled together during inference. The intuition behind this technique is that, although there are shifts in the distributions which govern the incoming data streams, these shifts are not expected to be homogeneous or global. If a sufficient diversity in solutions within the various sub-models has been achieved, then at least one sub-model is expected to retain its performance within the overall ensemble. One further strength of this approach is that the architectures of the various models do not need to be identical, and in fact even different modalities of data can naturally be combined in this way. This work focused on applying ensemble learning techniques to derive results using two main datasets, anomaly detection in HEP data & time-series forecasting in semiconductor manufacturing data. Semiconductor manufacturing involves data with surprising similarity to that of HEP (e.g. wafer maps look very similar to digi-occupancy maps) and Cerium Lab’s prominence within the semiconductor industry makes semiconductor manufacturing a natural opportunity for commercialization of this work. Our efforts have led to two strong results. The first is that we evaluated the proposed ensembling techniques using previously proposed machine learning architectures for use in anomaly detection, namely AutoEncoder based models and their derivatives. We also developed new architectures which have not been evaluated in this context before. In fact, this work marks the first use of Vision Transformers for anomaly detection in HEP. Second, we demonstrated that ensemble learning significantly improves model performance in scenarios prone to degradation, validating its effectiveness across both HEP and semiconductor datasets. These results further support ensemble learning as a powerful strategy for mitigating catastrophic forgetting and maintaining robust performance in evolving data environments.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2023

LES ARM Symbiotic Simulation and Observation Scenario for CACTI

The Large-Eddy Simulation (LES) Atmospheric Radiation Measurement (ARM) Symbiotic Simulation and Observation (LASSO) deep-convection scenario for the Cloud, Aerosol, and Complex Terrain Interactions (CACTI) field campaign. The LASSO-CACTI scenario consists of a library of mesoscale and LES simulations spanning grid spacings from 7.5 km to 100 m for convection near the Sierras de Córdoba mountain range, roughly centered on the ARM Mobile Facility. Supporting skill scores and observation comparisons complement the simulations in addition to subsets of the model output for various categories of model output variables. More information can be found at https://www.arm.gov/capabilities/modeling/lasso.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Assessing decadal variability of subseasonal forecasts of opportunity using explainable AI

Abstract Identifying predictable states of the climate system allows for enhanced prediction skill on the generally low-skill subseasonal timescale via forecasts with higher confidence and accuracy, known as forecasts of opportunity. This study takes a neural network approach to explore decadal variability of subseasonal predictability, particularly during forecasts of opportunity. Specifically, this work quantifies subseasonal prediction skill provided by the tropics within the Community Earth System Model Version 2 (CESM2) Large Ensemble and assesses how this skill evolves on decadal timescales. Utilizing the networks’ confidence and explainable artificial intelligence, physically meaningful sources of predictability associated with periods of enhanced skill are identified. Using these networks, we find that tropically-driven subseasonal predictability varies on decadal timescales during forecasts of opportunity. Further, we investigate the drivers of the low frequency modulation of the tropical-extratropical teleconnection and discuss the implications. Analysis is extended to ECMWF Reanalysis v5 data, revealing that the relationships learned within the CESM2-Large Ensemble holds in modern reanalysis data. These results indicate that the neural networks are capable of identifying predictable decadal states of the climate system within CESM2 that are useful for making confident, accurate subseasonal precipitation predictions in the real world.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Teleconnections of the Quasi–Biennial Oscillation in a multi–model ensemble of QBO–resolving models

The quasi-biennial oscillation (QBO) dominates the interannual variability of the tropical stratosphere and influences other regions of the atmosphere. The high predictability of the QBO implies that its teleconnections could lead to in-creased skill of seasonal and decadal forecasts provided the relevant mechanisms are accurately represented in models. Here modelling and sampling uncertainties of QBO teleconnections are examined using a multi-model ensemble of QBO-resolving atmospheric general circulation models that have carried out a set of coordinated experiments as part of the Stratosphere-troposphere Processes And their Role in Climate (SPARC) QBO initiative (QBOi). During Northern Hemisphere winter the stratospheric polar vortex in most of these models strengthens when the QBO near 50 hPa is westerly and weakens when it is easterly, consistent with, but weaker than, the observed response. These weak responses are likely due to model errors, such as systematically weak QBO amplitudes near 50 hPa, affecting the teleconnection. The teleconnection to the North Atlantic Oscillation is less well captured overall, but of similar strength to the observed signal in the few models that do show it. Furthermore, the models do not show clear evidence of a QBO teleconnection to the Northern Hemisphere Pacific sector subtropical jet.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Predicting peak day and peak hour of electricity demand with ensemble machine learning

Battery energy storage systems can be used for peak demand reduction in power systems, leading to significant economic benefits. Two practical challenges are 1) accurately determining the peak load days and hours and 2) quantifying and reducing uncertainties associated with the forecast in probabilistic risk measures for dispatch decision-making. In this study, we develop a supervised machine learning approach to generate 1) the probability of the next operation day containing the peak hour of the month and 2) the probability of an hour to be the peak hour of the day. Guidance is provided on preparation and augmentation of data as well as selection of machine learning models and decision-making thresholds. The proposed approach is applied to the Duke Energy Progress system and successfully captures 69 peak days out of 72 testing months with a 3% exceedance probability threshold. On 90% of the peak days, the actual peak hour is among the 2 h with the highest probabilities.

25 ENERGY STORAGE↗
DOE OSTI2025

Multi-Agent Hierarchical Deep Reinforcement Learning for HVAC Control With Flexible DERs

As electricity consumption in commercial and residential buildings continues to rise, reducing energy costs presents an increasing challenge. Heating, ventilating, and air-conditioning (HVAC) systems, which typically account for 40%-50% of a building's energy use, are prime targets for energy savings. Intelligent control of HVAC temperature through the exploitation of HVAC load flexibility brings significant potential to reduce energy consumption and electricity expenses. The nonlinear models of HVAC systems challenge traditional control methods, while the uncertainty introduced by HVAC load flexibility complicates distributed energy resource (DER) management using conventional optimal dispatch techniques. In response to these challenges, we propose a hierarchical multi-agent deep reinforcement learning (DRL) approach. The lower-level agents focus on balancing comfort and energy conservation, while the upper-level DRL agents optimize the use of DERs to reduce peak demand based on the control outcomes of the HVAC by the lower-level agents. Here, in the upper-level agents, we incorporate a multi-agent structure based on ensemble learning, which acts based on historical and current data without relying on precise load forecasting to address the delayed rewarding issue in DRL. This allows for the effective reduction of energy costs. The proposed method is tested using a real-world microgrid comprising 413 buildings in Southern California, and the results demonstrate that our approach can significantly reduce overall electricity bills while ensuring the comfort of consumers and residents.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2023

Optimizing Seasonal‐To‐Decadal Analog Forecasts With a Learned Spatially‐Weighted Mask

Abstract Seasonal‐to‐decadal climate prediction is crucial for decision‐making in a number of industries, but forecasts on these timescales have limited skill. Here, we develop a data‐driven method for selecting optimal analogs for seasonal‐to‐decadal analog forecasting. Using an interpretable neural network, we learn a spatially‐weighted mask that quantifies how important each grid point is for determining whether two climate states will evolve similarly. We show that analogs selected using this weighted mask provide more skillful forecasts than analogs that are selected using traditional spatially‐uniform methods. This method is tested on two prediction problems using the Max Planck Institute for Meteorology Grand Ensemble: multi‐year prediction of North Atlantic sea surface temperatures, and seasonal prediction of El Niño Southern Oscillation. This work demonstrates a methodical approach to selecting analogs that may be useful for improving seasonal‐to‐decadal forecasts and understanding their sources of skill.

Geology↗
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↗