Search NASA⌕ Search

SEARCH · Search NASA

Results for “ensemble optimization”

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 55 records · Page 3

Gradient-informed Hamiltonian Monte Carlo for multicomponent CALPHAD model optimization and uncertainty quantification

CALPHAD model parameter optimization is inherently challenging due to non-smooth objective functions, high-dimensional parameter spaces, and the need for uncertainty quantification (UQ). Traditional weighted nonlinear least squares approaches are computationally efficient but local, whereas black-box global optimizers and ensemble Markov Chain Monte Carlo (MCMC) methods provide broader exploration at substantial computational cost. The objective of this work is to combine the global exploration capability of gradient-informed Hamiltonian Monte Carlo – specifically the No-U-Turn Sampler (NUTS) – with local deterministic refinement using BFGS to efficiently optimize multicomponent CALPHAD models with minimal manual intervention. Analytic gradients are computed via the Jansson derivative framework. The methodology is demonstrated on the Cr—Fe binary system and extended to the Cr—Fe—Ni ternary system with 32 degrees of freedom. For Cr—Fe, NUTS achieves comparable or superior optimality relative to ensemble MCMC while requiring over an order-of-magnitude fewer likelihood evaluations. Parameter uncertainties are quantified through NUTS sampling and propagated to thermodynamic observables using local expansion, demonstrating a novel modular approach that combines binary and ternary parameter subsets without requiring global relaxation. These results establish gradient-informed exploration as a scalable strategy for multicomponent CALPHAD optimization and provide a practical route towards efficient higher-order database development with quantified uncertainty.

36 MATERIALS SCIENCE↗

A Predictive Prescription Framework for Stochastic Unit Commitment Using Boosting Ensemble Learning Algorithms

To take unit commitment (UC) decisions under uncertain load, most existing stochastic optimization (SO) frameworks adopt a generic representation of uncertainty. While load levels that materialize on a particular day are influenced by various covariates (such as the day of the week or temperature), SO frameworks typically disregard such side observations, wasting actionable information that could significantly enhance decision quality. Here, this article proposes a contextual SO (CSO) framework for UC under uncertain load, which can effectively exploit covariate observations in conjunction with a class of machine learning (ML) algorithms to improve the out-of-sample performance of UC decisions. It shows how three ML algorithms, adaptive boosting, gradient boosted trees, and extreme gradient boosting, can be used to this end, constituting the first application of these algorithms in any CSO framework. Using real-world data harvested from the New York ISO grid, we measure the out-of-sample performance of the framework in terms of total operation cost, shed load values, locational marginal prices, and total payments by the loads, against several benchmark methods proposed in the literature. The article has an online companion (Yurdakul et al.), wherein we present additional results and lay out further mathematical formulations used in this work.

42 ENGINEERING↗

Enumeration as a Tool for Structure Solution: A Materials Genomic Approach to Solving the Cation-Ordered Structure of Na 3 V 2 (PO 4 ) 2 F 3

While powder diffraction methods are routinely utilized to optimize structural models for compounds whose crystal structures are known, the determination of unknown structures is far more challenging. When the unknown structure is large, structure solution can become a virtually intractable problem using standard structure solution methodologies, especially when the space group cannot be unambiguously resolved. One such system is the promising Na-ion battery cathode material Na 3 V 2 (PO 4 ) 2 F 3 whose high temperature and room temperature structures were previously solved, but whose more complex low-temperature structure could not be determined. Here, a novel materials genomic approach is demonstrated for the solution of the unknown 100 K structure of Na 3 V 2 (PO 4 ) 2 F 3 in which enumeration methods are first used to generate a large number (~3,000) of trial structures based on plausible orderings of Na ions and then automated Rietveld refinements are carried out to optimize each of these trial structures. Based on both the analysis of the ensemble of optimized trial structures and the density functional theory energy minimization of selected trial structures, the 100 K structure of Na 3 V 2 (PO 4 ) 2 F 3 is best described as belonging to the space group A2 1 am with unit cell dimensions of a = 9.01928(4), b = 27.1379(1), c = 10.73307(5). The 100 K unit cell has a large volume of 2627.07(2) Å 3 with Z = 12 and 33 independent crystallographic sites (9 Na, 3 V, 3 P, 12 O, and 6 F) that is 3x and 6x larger than the room- and high-temperature polymorphs of this phase, respectively. Finally, the novel methods described here will be generally applicable for the solution of the complex cation-ordered structures that commonly occur for battery materials.

36 MATERIALS SCIENCE↗

Accelerated iterative beam angle selection in IMRT

Purpose: Iterative methods for beam angle selection (BAS) for intensity-modulated radiation therapy (IMRT) planning sequentially construct a beneficial ensemble of beam directions. In a naïve implementation, the nth beam is selected by adding beam orientations one-by-one from a discrete set of candidates to an existing ensemble of (n − 1) beams. The best beam orientation is identified in a time consuming process by solving the fluence map optimization (FMO) problem for every candidate beam and selecting the beam that yields the largest improvement to the objective function value. This paper evaluates two alternative methods to accelerate iterative BAS based on surrogates for the FMO objective function value. Methods: We suggest to select candidate beams not based on the FMO objective function value after convergence but (1) based on the objective function value after five FMO iterations of a gradient based algorithm and (2) based on a projected gradient of the FMO problem in the first iteration. The performance of the objective function surrogates is evaluated based on the resulting objective function values and dose statistics in a treatment planning study comprising three intracranial, three pancreas, and three prostate cases. Furthermore, iterative BAS is evaluated for an application in which a small number of noncoplanar beams complement a set of coplanar beam orientations. This scenario is of practical interest as noncoplanar setups may require additional attention of the treatment personnel for every couch rotation. Results: Iterative BAS relying on objective function surrogates yields similar results compared to naïve BAS with regard to the objective function values and dose statistics. At the same time, early stopping of the FMO and using the projected gradient during the first iteration enable reductions in computation time by approximately one to two orders of magnitude. With regard to the clinical delivery of noncoplanar IMRT treatments, we could show that optimized beam ensembles using only a few noncoplanar beam orientations often approach the plan quality of fully noncoplanar ensembles. Conclusions: We conclude that iterative BAS in combination with objective function surrogates can be a viable option to implement automated BAS at clinically acceptable computation times.

61 RADIATION PROTECTION AND DOSIMETRY↗

Noncoplanar VMAT for nasopharyngeal tumors: Plan quality versus treatment time

Purpose: The authors investigated the potential of optimized noncoplanar irradiation trajectories for volumetric modulated arc therapy (VMAT) treatments of nasopharyngeal patients and studied the trade-off between treatment plan quality and delivery time in radiation therapy. Methods: For three nasopharyngeal patients, the authors generated treatment plans for nine different delivery scenarios using dedicated optimization methods. They compared these scenarios according to dose characteristics, number of beam directions, and estimated delivery times. In particular, the authors generated the following treatment plans: (1) a 4π plan, which is a not sequenced, fluence optimized plan that uses beam directions from approximately 1400 noncoplanar directions and marks a theoretical upper limit of the treatment plan quality, (2) a coplanar 2π plan with 72 coplanar beam directions as pendant to the noncoplanar 4π plan, (3) a coplanar VMAT plan, (4) a coplanar step and shoot (SnS) plan, (5) a beam angle optimized (BAO) coplanar SnS IMRT plan, (6) a noncoplanar BAO SnS plan, (7) a VMAT plan with rotated treatment couch, (8) a noncoplanar VMAT plan with an optimized great circle around the patient, and (9) a noncoplanar BAO VMAT plan with an arbitrary trajectory around the patient. Results: VMAT using optimized noncoplanar irradiation trajectories reduced the mean and maximum doses in organs at risk compared to coplanar VMAT plans by 19% on average while the target coverage remains constant. A coplanar BAO SnS plan was superior to coplanar SnS or VMAT; however, noncoplanar plans like a noncoplanar BAO SnS plan or noncoplanar VMAT yielded a better plan quality than the best coplanar 2π plan. The treatment plan quality of VMAT plans depended on the length of the trajectory. The delivery times of noncoplanar VMAT plans were estimated to be 6.5 min in average; 1.6 min longer than a coplanar plan but on average 2.8 min faster than a noncoplanar SnS plan with comparable treatment plan quality. Conclusions: The authors’ study reconfirms the dosimetric benefits of noncoplanar irradiation of nasopharyngeal tumors. Both SnS using optimized noncoplanar beam ensembles and VMAT using an optimized, arbitrary, noncoplanar trajectory enabled dose reductions in organs at risk compared to coplanar SnS and VMAT. Using great circles or simple couch rotations to implement noncoplanar VMAT, however, was not sufficient to yield meaningful improvements in treatment plan quality. The authors estimate that noncoplanar VMAT using arbitrary optimized irradiation trajectories comes at an increased delivery time compared to coplanar VMAT yet at a decreased delivery time compared to noncoplanar SnS IMRT.

60 APPLIED LIFE SCIENCES↗

Unraveling Adsorbate-Induced Structural Evolution of Iron Carbide Nanoparticles

Iron carbide (Fe x C y ) nanoparticles (NPs) are promising candidates for replacing platinum group metals in industrial applications, such as high-temperature Fischer–Tropsch synthesis. However, due to their amorphous nature, characterization of the active sites has been challenging experimentally and computationally. Here, using a combined density functional theory (DFT), neural network interatomic potential-assisted global optimization, and ensemble learning study, we evaluate dynamic surface changes associated with syngas (H and CO) interactions. For this purpose, we have developed a general procedure that we use to model an experimentally relevant 270-atom Fe 182 C 88 NP using the neural network-assisted stochastic surface walk global optimization algorithm (SSW-NN). Once generated, the Fe 182 C 88 NP active sites and particle morphology are thoroughly characterized before the effects of syngas adsorbate interactions are explored by using DFT and molecular dynamics simulations. Lastly, we explore correlations between geometric and electronic features of the active sites and the adsorption of H (H ads ), using a regularized random forest machine learning algorithm. In doing so, we identified the Fe–C coordination number and p orbital occupancy as the most important descriptors affecting H ads . Furthermore, using a combined ML and quantum chemistry approach, our work demonstrates a general and efficient procedure for generating and probing complex surface phenomena on binary nanoparticles.

Adsorption↗

The underestimated magnitude and decline trend in near-surface wind over China

This study reports the magnitude, spatial pattern and temporal trend of near-surface wind speed (NWS) by comparing 20th century simulations in Coupled Model Intercomparison Project phase 5 (CMIP5) and 3 (CMIP3) and the climate reanalyses with measurements at 563 weather stations in China over 1961–2005. Both CMIP5 and CMIP3 agree quite well with observations in reproducing the spatial pattern of annual mean NWS. CMIP5 models are superior to CMIP3 models in hindcasting the magnitude and spatial pattern of seasonal mean NWS, the temporal trend in annual and seasonal mean NWS. Although both CMIP3 and CMIP5 reproduced the decline trend in the annual and seasonal mean NWS, the hindcasted decline rate is smaller than observed decline trend by the magnitude of one order. The ensemble of optimal models that are better correlated to observations in both NWS and temporal trend possesses advantages over individual model hindcast and reanalyses. The reanalyzed data are not able to represent the observed either the spatial pattern or the decline trend of NWS. The analyses presented here reveal the uncertainties in the current wind field products including reanalyses and model outputs and highlight the benefits of parameterization development and increased horizontal resolution in the new-generation CMIP models.

54 ENVIRONMENTAL SCIENCES↗

Online LIBS–ML Framework for Dynamic Characterization of Heterogeneous Waste-Derived Gasification Feedstocks

LIBS−ML framework for real time feedstock characterization during continuous conveyor transport Heterogeneous waste derived feedstocks (e.g., waste coal, biomass and blends) introduce rapid variability in heating value and ash chemistry that affect gasifier operation, yet conventional laboratory characterization techniques are too slow to support proactive control. To address this gap, this study reports on an online, in situ, dynamic characterization framework that couple’s laser-induced breakdown spectroscopy (LIBS) with leakage safe machine learning (ML) regression to deliver real time, decision quality predictions of gasifier relevant properties. A controlled sample matrix spanning two different waste coals, two different biomasses, and engineered blends under two particle size conditions were constructed and benchmarked using standardized laboratory analyses for proximate/ultimate properties and ash composition. LIBS spectra were acquired dynamically as material flowed on a conveyor belt, using high energy 1064 nm laser ablation and shot averaging to improve repeatability and precision. Supervised regression models (multi layer perceptron (MLP) /artificial neural network (ANN), random forest (RF), and support vector regression (SVR)) and an optimized weighted ensemble were trained on emission line feature sets using nested cross validation with Bayesian hyperparameter tuning and validated against an independent hold out set. The proposed LIBS−ML workflow achieves near laboratory predictive fidelity across parametric targets (including higher heating value (HHV), ash content, fixed carbon, sulfur, major ash forming oxides, and initial deformation temperature (IDT)), with the weighted ensemble providing a robust default predictor under dynamic measurement conditions. These results demonstrate a practical pathway for real time feedstock characterization that can enable feedforward adjustments and more resilient gasifier operation for variable quality waste derived fuels.

Biomass↗

Adaptive Pore Opening to Form Tailored Adsorption Sites in a Cooperatively Flexible Framework Enables Record Inverse Propane/Propylene Separation

A proposed low-energy alternative to the separation of alkanes from alkenes by energy-intensive cryogenic distillation is separation by porous adsorbents. Unfortunately, most adsorbents preferentially take up the desired, high-value major component alkene, requiring frequent regeneration. Adsorbents with inverse selectivity for the minor component alkane would enable the direct production of purified, reagent-grade alkene, greatly reducing global energy consumption. However, such materials are exceedingly rare, especially for propane/propylene separation. Here, we report that through adaptive and spontaneous pore size and shape adaptation to optimize an ensemble of weak noncovalent interactions, the structurally responsive metal-organic framework CdIF-13 (sod-Cd(benzimidazolate) 2 ) exhibits inverse selectivity for propane over propylene with record-setting separation performance under industrially relevant temperature, pressure, and mixture conditions. Powder synchrotron X-ray diffraction measurements combined with first-principles calculations yield atomic-scale insight and reveal the induced fit mechanism of adsorbate-specific pore adaptation and ensemble interactions between ligands and adsorbates. Dynamic column breakthrough measurements confirm that CdIF-13 displays selectivity under mixed-component conditions of varying ratios, with a record measured selectivity factor of α ≈ 3 at 95:5 propylene:propane at 298 K and 1 bar. When sequenced with a low-cost rigid adsorbent, we demonstrated the direct purification of propylene under ambient conditions. In conclusion, this combined atomic-level structural characterization and performance testing firmly establishes how cooperatively flexible materials can be capable of unprecedented separation factors.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

ETHOS: An automated framework to generate multi-fidelity constitutive data tables and propagate uncertainties to hydrodynamic simulations

Accurate constitutive data, such as equations of state and plasma transport coefficients, are necessary for reliable hydrodynamic simulations of plasma systems such as fusion targets, planets, and stars. Here, we develop a framework for automatically generating transport-coefficient tables using a parameterized model that incorporates data from both high-fidelity sources (e.g., density functional theory calculations and reference experiments) and lower-fidelity sources (e.g., average-atom and analytic models). The framework incorporates uncertainties from these multi-fidelity sources, generating ensembles of optimally diverse tables that are suitable for uncertainty quantification of hydrodynamic simulations. We illustrate the utility of the framework with magnetohydrodynamic simulations of magnetically launched flyer plates, which are used to measure material properties in pulsed-power experiments. We explore how changes in the uncertainties assigned to the multi-fidelity data sources propagate to changes in simulation outputs and find that our simulations are most sensitive to uncertainties near the melting transition. The presented framework enables computationally efficient uncertainty quantification that readily incorporates new high-fidelity measurements or calculations and identifies plasma regimes where additional data will have high impact.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Ensemble-Based Parameter Estimation in a Coupled GCM Using the Adaptive Spatial Average Method

Ensemble-based parameter estimation for a climate model is emerging as an important topic in climate research. And for a complex system such as a coupled ocean–atmosphere general circulation model, the sensitivity and response of a model variable to a model parameter could vary spatially and temporally. An adaptive spatial average (ASA) algorithm is proposed to increase the efficiency of parameter estimation. Refined from a previous spatial average method, the ASA uses the ensemble spread as the criterion for selecting “good” values from the spatially varying posterior estimated parameter values; these good values are then averaged to give the final global uniform posterior parameter. In comparison with existing methods, the ASA parameter estimation has a superior performance: faster convergence and enhanced signal-to-noise ratio.

54 ENVIRONMENTAL SCIENCES↗

Probabilistic Cloud Optimized Day-Ahead Forecasting System Based on WRF-Solar (Final Report)

The most persistent challenge in both intraday and day-ahead solar forecasting is to get numerical weather prediction models to produce the right type of clouds with the right frequency at the right time and place. Another challenge is to understand and communicate the forecast uncertainty. The objective of this project was to develop an optimized ensemble-based solar irradiance forecasting system that will (1) demonstrably improve the current state-of-the-art solar forecasts from the deterministic Weather Research and Forecasting-Solar (WRF-Solar) model and (2) provide probabilistic forecasts for grid operations. This probabilistic solar forecasting system, referred to as the WRF-Solar Ensemble Prediction System (WRF-Solar EPS), aims to significantly enhance both the intraday and the day-ahead solar forecasting capability for grid operations. This technical report summarizes the work performed in the past 3 years through a collaboration between the National Renewable Energy Laboratory and the National Center for Atmospheric Research as part of the U.S. Department of Energy's Solar Forecasting 2 program that aims to improve the accuracy of solar energy forecasts and enable increased deployment of solar energy on the electric grid.

14 SOLAR ENERGY↗

Classical-Quantum Algorithm for Solving Stochastic Programs

Stochastic programming provides a rigorous mathematical framework for making decisions under uncertainty in a risk-aware manner. Two-stage stochastic programming is, perhaps, the simplest form of this framework. Here the first-stage variables represent decisions that must be made "here and now" in the face of uncertainty, while the second-stage variables are decisions made after uncertain events. However, the broad adoption of stochastic programming has been hindered by computational challenges caused by the two-stage stochastic programming formulation which requires solving an ensemble of optimization problems. Using quantum amplitude estimation (QAE), quantum computers have shown the theoretic ability to compute expectations with Monte-Carlo methods with quadratically fewer samples than classical methods. In this work, we present a quantum algorithm for computing the expectation term using QAE for given first-stage decisions. Further, we detail methods of computing gradient information from the quantum calculation enabling the application of classical gradient-based optimization techniques. The result is a classical-quantum hybrid method of solving two-stage stochastic programs. These techniques are demonstrated with computational experiments based an engineering optimization problem.

97 MATHEMATICS AND COMPUTING↗

A Novel Active Optimization Approach for Rapid and Efficient Design Space Exploration Using Ensemble Machine Learning

In this work, a novel design optimization technique based on active learning, which involves dynamic exploration and exploitation of the design space of interest using an ensemble of machine learning algorithms, is presented. In this approach, a hybrid methodology incorporating an explorative weak learner (regularized basis function model) that fits high-level information about the response surface and an exploitative strong learner (based on committee machine) that fits finer details around promising regions identified by the weak learner is employed. For each design iteration, an aristocratic approach is used to select a set of nominees, where points that meet a threshold merit value as predicted by the weak learner are selected for evaluation. In addition to these points, the global optimum as predicted by the strong learner is also evaluated to enable rapid convergence to the actual global optimum once the most promising region has been identified by the optimizer. Additionally, this methodology is first tested by applying it to the optimization of a two-dimensional multi-modal surface and, subsequently, to a complex internal combustion (IC) engine combustion optimization case with nine control parameters related to fuel injection, initial thermodynamic conditions, and in-cylinder flow. It is found that the new approach significantly lowers the number of function evaluations that are needed to reach the optimum design configuration (by up to 80%) when compared to conventional optimization techniques, such as particle swarm and genetic algorithm-based optimization techniques.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Optimizing the Cloud-Aerosol-Radiation Ensemble Modeling System to Improve Future Climate Change Projections at Regional to Local Scales

We have successfully built up, in consistency and completeness, the Cloud-Aerosol-Radiation (CAR) Ensemble Modeling System. Then we have published the CAR general paper to present a general description of the CAR system and illustrate its capabilities for climate modeling applications, especially in the context of estimating climate sensitivity and uncertainty range caused by cloud-aerosol-radiation interactions. For demonstration purpose, the evaluation is based on several CAR standalone experiments driven by CIRC Phase I data or ERA-Interim observational analysis data (ERI) 2004 global data, and coupled climate model experiments by using CWRF, each comparing a limited subset of the full system ensemble with up to 896 members. It is shown that the quantification of radiative forcings and climate impacts strongly depends on the choices of the cloud, aerosol and radiation schemes. The prevailing schemes used in current GCMs are likely insufficient in variety and physically biased in a significant way. There exists large room for improvement by optimally combining radiation transfer with cloud property schemes.

58 GEOSCIENCES↗

Creation of a Robust and Generalizable Machine Learning Classifier for Patient Ventilator Asynchrony

Background: As healthcare increasingly digitizes, streaming waveform data is being made available from an variety of sources, but there still remains a paucity of performant clinical decision support systems. For example, in the intensive care unit (ICU) existing automated alarm systems typically rely on simple thresholding that result in frequent false positives. Recurrent false positive alerts create distrust of alarm mechanisms that can be directly detrimental to patient health. To improve patient care in the ICU, we need alert systems that are both pervasive, and accurate so as to be informative and trusted by providers. Objective: We aimed to develop a machine learning-based classifier to detect abnormal waveform events using the use case of mechanical ventilation waveform analysis, and the detection of harmful forms of ventilation delivery to patients. We specifically focused on detecting injurious subtypes of patient-ventilator asynchrony (PVA).Methods: Using a dataset of breaths recorded from 35 different patients, we used machine learning to create computational models to automatically detect, and classify two types of injurious PVA, double trigger asynchrony (DTA), breath stacking asynchrony (BSA). We examined the use of synthetic minority over-sampling technique (SMOTE) to overcome class imbalance problems, varied methods for feature selection, and use of ensemble methods to optimize the performance of our model. Results: We created an ensemble classifier that is able to accurately detect DTA at a sensitivity/specificity of 0.960/0.975, BSA at sensitivity/specificity of 0.944/0.987, and non-PVA events at sensitivity/specificity of .967/.980. Conclusions: Our results suggest that it is possible to create a high-performing machine learning-based model for detecting PVA in mechanical ventilator waveform data in spite of both intra-patient, and inter-patient variability in waveform patterns, and the presence of clinical artifacts like cough and suction procedures. Our work highlights the importance of addressing class imbalance in clinical data sets, and the combined use of statistical methods and expert knowledge in feature selection.

97 MATHEMATICS AND COMPUTING↗

A novel active optimization approach for rapid and efficient design space exploration using ensemble machine learning

In this work, a novel design optimization technique based on active learning, which involves dynamic exploration and exploitation of the design space of interest using an ensemble of machine learning algorithms, is presented. In this approach, a hybrid methodology incorporating an explorative weak learner (regularized basis function model) which fits high-level information about the response surface, and an exploitative strong learner (based on committee machine) that fits finer details around promising regions identified by the weak learner, is employed. For each design iteration, an aristocratic approach is used to select a set of nominees, where points that meet a threshold merit value as predicted by the weak learner are selected to be evaluated using expensive function evaluation. In addition to these points, the global optimum as predicted by the strong learner is also evaluated to enable rapid convergence to the actual global optimum once the most promising region has been identified by the optimizer. This methodology is first tested by applying it to the optimization of a two-dimensional multi-modal surface. The performance of the new active learning approach is compared with traditional global optimization methods, namely micro-genetic algorithm (pGA) and particle swarm optimization (PSO). It is demonstrated that the new optimizer is able to reach the global optimum much faster with a significantly fewer number of function evaluations. Subsequently, the new optimizer is also applied to a complex internal combustion (IC) engine combustion optimization case with nine control parameters related to fuel injection, initial thermodynamic conditions, and in-cylinder flow. It is again found that the new approach significantly lowers the number of function evaluations that are needed to reach the optimum design configuration (by up to 80%) when compared to particle swarm and genetic algorithm-based optimization techniques.

Machine learning, active learning, design optimiza↗

Combining multi-objective optimization and bayesian model averaging to calibrate forecast ensembles of soil hydraulic models

Most studies in vadose zone hydrology use a single conceptual model for predictive inference and analysis. Focusing on the outcome of a single model is prone to statistical bias and underestimation of uncertainty. In this study, we combine multi-objective optimization and Bayesian Model Averaging (BMA) to generate forecast ensembles of soil hydraulic models. To illustrate our method, we use observed tensiometric pressure head data at three different depths in a layered vadose zone of volcanic origin in New Zealand. A set of seven different soil hydraulic models is calibrated using a multi-objective formulation with three different objective functions that each measure the mismatch between observed and predicted soil water pressure head at one specific depth. The Pareto solution space corresponding to these three objectives is estimated with AMALGAM, and used to generate four different model ensembles. These ensembles are post-processed with BMA and used for predictive analysis and uncertainty estimation. Our most important conclusions for the vadose zone under consideration are: (1) the mean BMA forecast exhibits similar predictive capabilities as the best individual performing soil hydraulic model, (2) the size of the BMA uncertainty ranges increase with increasing depth and dryness in the soil profile, (3) the best performing ensemble corresponds to the compromise (or balanced) solution of the three-objective Pareto surface, and (4) the combined multi-objective optimization and BMA framework proposed in this paper is very useful to generate forecast ensembles of soil hydraulic models.

58 GEOSCIENCES↗