Search NASA⌕ Search

SEARCH · Search NASA

Results for “Bayesian analysis”

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 1,099 records · Page 61

THE 4 Ms CHANDRA DEEP FIELD-SOUTH NUMBER COUNTS APPORTIONED BY SOURCE CLASS: PERVASIVE ACTIVE GALACTIC NUCLEI AND THE ASCENT OF NORMAL GALAXIES

We present 0.5-2 keV, 2-8 keV, 4-8 keV, and 0.5-8 keV (hereafter soft, hard, ultra-hard, and full bands, respectively) cumulative and differential number-count (log N-log S) measurements for the recently completed Almost-Equal-To 4 Ms Chandra Deep Field-South (CDF-S) survey, the deepest X-ray survey to date. We implement a new Bayesian approach, which allows reliable calculation of number counts down to flux limits that are factors of Almost-Equal-To 1.9-4.3 times fainter than the previously deepest number-count investigations. In the soft band (SB), the most sensitive bandpass in our analysis, the Almost-Equal-To 4 Ms CDF-S reaches a maximum source density of Almost-Equal-To 27,800 deg{sup -2}. By virtue of the exquisite X-ray and multiwavelength data available in the CDF-S, we are able to measure the number counts from a variety of source populations (active galactic nuclei (AGNs), normal galaxies, and Galactic stars) and subpopulations (as a function of redshift, AGN absorption, luminosity, and galaxy morphology) and test models that describe their evolution. We find that AGNs still dominate the X-ray number counts down to the faintest flux levels for all bands and reach a limiting SB source density of Almost-Equal-To 14,900 deg{sup -2}, the highest reliable AGN source density measured at any wavelength. We find that the normal-galaxy counts rise rapidly near the flux limits and, at the limiting SB flux, reach source densities of Almost-Equal-To 12,700 deg{sup -2} and make up 46% {+-} 5% of the total number counts. The rapid rise of the galaxy counts toward faint fluxes, as well as significant normal-galaxy contributions to the overall number counts, indicates that normal galaxies will overtake AGNs just below the Almost-Equal-To 4 Ms SB flux limit and will provide a numerically significant new X-ray source population in future surveys that reach below the Almost-Equal-To 4 Ms sensitivity limit. We show that a future Almost-Equal-To 10 Ms CDF-S would allow for a significant increase in X-ray-detected sources, with many of the new sources being cosmologically distant (z {approx}> 0.6) normal galaxies.

79 ASTRONOMY AND ASTROPHYSICS↗

Fables of reconstruction: controlling bias in the dark energy equation of state

We develop an efficient, non-parametric Bayesian method for reconstructing the time evolution of the dark energy equation of state w(z) from observational data. Of particular importance is the choice of prior, which must be chosen carefully to minimise variance and bias in the reconstruction. Using a principal component analysis, we show how a correlated prior can be used to create a smooth reconstruction and also avoid bias in the mean behaviour of w(z). We test our method using Wiener reconstructions based on Fisher matrix projections, and also against more realistic MCMC analyses of simulated data sets for Planck and a future space-based dark energy mission. While the accuracy of our reconstruction depends on the smoothness of the assumed w(z), the relative error for typical dark energy models is ∼<10% out to redshift z = 1.5.

79 ASTRONOMY AND ASTROPHYSICS↗

Vibration-Based Techniques for Concrete Structural Health Monitoring. Light Water Reactor Sustainability Program report

The objectives of this ongoing research project focus on health monitoring and data analytics of concrete slabs containing reactive aggregates and thus subjected to degradation due to alkali-silica reaction (ASR). A controlled concrete slab with four pockets of reactive aggregates (pure silica, and reactive aggregates from three different quarries) was cast at Vanderbilt University and cured in representative conditions to accelerate degradation due to ASR. A set of four concrete samples were also cast and cured at the University of Alabama for ASR testing. Of these four samples, two slabs contained reactive aggregates while the other two had the non-reactive aggregate counterparts mixed throughout the samples. Four additional slabs were cast and cured at the University of Nebraska- Lincoln. Of these four samples; one contained non-reactive aggregates and no reinforcement, one contained reactive aggregates and no reinforcement, one contained reactive aggregates and rebar reinforcement in one direction, and the last contained reactive aggregates and reinforcement in two directions. Single frequency and dual frequency tests were used on these slabs to locate ASR damage within the specimens containing reactive aggregates. The vibration tests were conducted with different sensor and actuator configurations to facilitate damage detection and localization. The results from the data analysis of the dynamic testing on concrete slabs are discussed in detail in this report. Results for damage localization are dependent on multiple parameters used in the tests. This report investigates different methods of data fusion including voting, raw averaging, damage index averaging, and Bayesian fusion.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

A Differentiable Hydrology Approach for Modeling With Time-Varying Parameters

In the analysis of observational data with mathematical models, it is often desirable to let one or more model parameters vary with time to account for changing environmental conditions or to offer more flexibility. A standard approach for dynamic models is to allow for the number of parameters to grow as a function of time, leading to a high-dimensional inference problem for long data records. We propose using Hamiltonian Monte Carlo, a gradient-based Markov chain Monte Carlo (MCMC) method, to perform Bayesian inference for time-varying parameters (TVP) in hydrology models. As derivatives of model error functions with regard to parameters are not available in closed form, we implemented the GR4J rainfall-runoff models in Theano/PyMC3, and Jax/NumPyro, allowing for the application of automatic differentiation techniques. A simulation experiment assessing the viability of these methods for recovering underlying temporal variation in parameters indicates that a discrete Gaussian random walk-based prior appears to be best suited across a range of scenarios while the continuous-time Gaussian process compares relatively poorly. In an analysis of 20 years of daily real-world streamflow records from the Model Parameter Estimation Experiment (MOPEX) hydrology data set with increasing data sparsity and at multiple temporal resolutions, we find that TVP inference with gradient-based MCMC is a flexible and reliable approach for analyzing hydrology models with dynamic parameter sets.

54 ENVIRONMENTAL SCIENCES↗

Stan : A Probabilistic Programming Language

Stan is a probabilistic programming language for specifying statistical models. A Stan program imperatively defines a log probability function over parameters conditioned on specified data and constants. As of version 2.14.0, Stan provides full Bayesian inference for continuous-variable models through Markov chain Monte Carlo methods such as the No-U-Turn sampler, an adaptive form of Hamiltonian Monte Carlo sampling. Penalized maximum likelihood estimates are calculated using optimization methods such as the limited memory Broyden-Fletcher-Goldfarb-Shanno algorithm. Stan is also a platform for computing log densities and their gradients and Hessians, which can be used in alternative algorithms such as variational Bayes, expectation propagation, and marginal inference using approximate integration. To this end, Stan is set up so that the densities, gradients, and Hessians, along with intermediate quantities of the algorithm such as acceptance probabilities, are easily accessible. Stan can also be called from the command line using the cmdstan package, through R using the rstan package, and through Python using the pystan package. All three interfaces support sampling and optimization-based inference with diagnostics and posterior analysis. rstan and pystan also provide access to log probabilities, gradients, Hessians, parameter transforms, and specialized plotting.

97 MATHEMATICS AND COMPUTING↗

Assessing and benchmarking the fidelity of posterior inference methods for astrophysics data analysis

In this era of large and complex astronomical survey data, interpreting, validating, and comparing inference techniques becomes increasingly difficult. This is particularly critical for emerging inference methods like Simulation-Based Inference (SBI), which offer significant speedup potential and posterior modeling flexibility, especially when deep learning is incorporated. We present a study to assess and compare the performance and uncertainty prediction capability of Bayesian inference algorithms – from traditional MCMC sampling of analytic functions to deep learning-enabled SBI. We focus on testing the capacity of hierarchical inference modeling in those scenarios. Before we extend this study to cosmology, we first use astrophysical simulation data to ensure interpretability. We demonstrate a probabilistic programming implementation of hierarchical and non-hierarchical Bayesian inference using simulations derived from the DeepBench software library, a benchmarking tool developed by our group that generates simple and controllable astrophysical objects from first principles. This study will enable astronomers and physicists to harness the inference potential of these methods with confidence.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

A Bayesian Monte Carlo inversion of spatial auto-correlation (SPAC) for near-surface Vs structure applied to both broad-band and geophone data

SUMMARY We propose a new Bayesian method to reveal the Vs structure of the near surface of the earth using spatial autocorrelation (SPAC) functions and apply this new method to synthetic, broad-band and geophone data sets. The principle of SPAC is introduced, and an implementation of the Bayesian Monte Carlo inversion (BMCI) for modelling SPAC coherency functions is described. To demonstrate its effectiveness, BMCI is applied to synthetic tests, data from 14 SPAC array sites in the Salt Lake Valley (SLV), Utah, and two arrays (one broad-band and one geophone) located in south central Utah. The Vs models derived from previous SPAC analysis of the 14 SLV sites differ by 10 per cent at most from those determined by BMCI and lie within uncertainties determined for the BMCI models. These agreements demonstrate the effectiveness of the BMCI method. The synthetic tests and applications to the SLV SPAC data show BMCI has great potential to resolve Vs structure down to at least 400 m. To achieve resolution for deeper Vs structure, longer duration deployments, wider array apertures, and additional seismometers or geophones can be used. Additionally, when the target frequencies are greater than 0.1 Hz, there is no apparent disadvantage in using geophone data for BMCI compared to broad-band data. Most significantly, BMCI places a quantifiable constraint on the uncertainties of the Vs models as well as Vs30.

58 GEOSCIENCES↗

Inverter Model Validation and Calibration Using Phasor Measurement Unit Data

As the penetration of inverter-based renewable energy resources increases in the power grid, especially at the distribution and microgrid levels, the need to accurately represent them in planning studies increases as well. However, due to the lack of well-established standard procedures, and vendor reluctance towards the detailed sharing of proprietary models, automated dynamic model validation and parameter calibration tools for inverter based resources (IBRs) remain scarce. This work presents a model validation and parameter calibration platform for representing IBRs with generic phasor-domain models. Phasor measurements of power system events are used for continuous validation using the data playback method, and model parameters are re-calibrated if a significant mismatch between measurements and model response is observed. Unique features of the proposed platform include- (a) an iterative Bayesian optimization approach towards parameter calibration to address a possible mismatch between the structures of generic models implemented in simulation softwares and actual commercial inverters, (b) error metrics designed to account for a possible mismatch between the time resolution of simulation and measurements, and (c) analysis of the measurement-simulation mismatch to provide guidance to engineering personnel regarding model shortcomings. The performance of the platform has been illustrated using both simulated data and field measurements to validate/calibrate inverter models in GridLAB-D.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Voltron Compatible Whole Building Root-Fault Detection and Diagnosis

Buildings consume more than 40% of primary energy in the U.S. Malfunctioning sensors, components, and control systems, as well as degrading components in Heating, Ventilating and Air-conditioning (HVAC) and lighting systems are main reasons for energy waste and unsatisfactory indoor environment. Field studies have shown that an energy saving of 5-30% of total building energy consumption and improved indoor air quality can be achieved by simply applying automated fault detection and diagnosis (AFDD), followed by corrections even if these are not done in real time. Extensive research has been made on the development of component level AFDD tools in the past two decades. However, in recent years, the development of method for a whole building level AFDD has received much attention. Here, a whole building fault refers to a fault happening in one subsystem but impact on more than one subsystems, or has a significant impact on building performances, such as building’s energy consumption. Existing component level AFDD tools could fail to detect or give false alarm for faults that trigger abnormalities in multiple subsystems. Isolating such whole building faults, which have impacts on multiple subsystems, is also challenging by simply using component based AFDD solutions. On the other hand, it might be more cost-effective for medium- or small-sized commercial buildings that have subsystems (e.g. primary cooling subsystem, air loop subsystem, heating subsystem, lighting subsystem, etc.) to install one whole building level AFDD tool, rather than many component level AFDD tools, to focus on significant faults. The focus of this project is firstly to provide training opportunities in building AFDD research area for undergraduate students and to develop a workforce pipeline in building sciences. The objective of the proposed research activities is to develop cost-effective and scalable AFDD solutions for whole building level faults. The developed AFDD solution is VOLTTRONTM compatible to further increase its plug-n-play capacity and the market penetration. VOLTTRONTM is an open-source innovative distributed control and sensing software platform developed by the Pacific Northwest National Laboratory. A VOLTTRONTM compatible AFDD tool brings the benefits such as low-cost deployment, and better building to grid integration, etc. The performance of the developed whole building AFDD tools is evaluated using data collected from a mixed-use medium-sized university campus building. Another significant goal of this project is to engage undergraduate students in research activities and expose them to building control and AFDD fields. A demonstration building – Nesbitt Hall at Drexel University is identified for this study. This demonstration building has a typical HVAC system configuration, which includes a water-cooled chiller system, three variable air volume (VAV) air handling unit (AHU) systems, and a hydronic heating system. Faults, that are expected to have a whole building level impact, are artificially implemented in this demonstration building for three different seasons. Building data have been collected in this project, which include fault free building automation system (BAS) data and BAS data that contain system behaviors caused by artificially implemented and naturally occurred faults. These collected data are used to evaluate the developed whole building AFDD tools. A data-driven method, which includes a weather and schedule-based pattern matching (WPM) method and Feature based Principal Component Analysis (FPCA) method, is developed for whole building level fault detection. Parameter sensitivity tests, including snapshot window size, data sample searching pool size, etc., are implemented to evaluate the impacts of various WPM-FPCA parameters. A data-driven and expert knowledge/rule-based method using Bayesian Network (BN) and WPM is developed for whole building level fault diagnosis. The developed WPM-BN method, which includes a two-layer BN structure model and BN parameters, are either learned from baseline data or developed from expert knowledge. The developed whole building AFDD tools are evaluated using BAS data collected from the demonstration building. Using the evaluation data set, the WPM-FPCA fault detection method is evaluated under two different Principle Component (PC) retention rates, i.e., 0.65 retention rate and 0.95 retention rate. It is observed that when using 0.65 PC retention rate, the WPM-FPCA fault detection method achieves a 79% fault detection rate (this rate reaches to 85% when including the six make-up fault test cases discussed in the next bullet), while when using 0.95 PC retention rate, the WPM-FPCA fault detection method achieves 100% fault detection rate (this rate remains to be 100% when including the six make-up fault test cases). Under both PC retention rates, the WPM-FPCA fault detection method demonstrates 0% false alarm rates. Using the evaluation data set, the developed WPM-BN fault diagnosis method has successfully diagnosed thirteen (13) faults out of the fourteen (14) fault test cases. The one mis-diagnosed case exhibits fewer fault symptoms in different subsystems, when compared with other fault test cases. Overall speaking, the developed whole building AFDD solutions demonstrate satisfactory fault detection/diagnosis accuracy with very low false alarm rate. Lastly, a survey and market study has been performed to understand the existing AFDD market, including its gaps and needs.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Holographic dark energy models: a comparison from the latest observational data

The holographic principle of quantum gravity theory has been applied to the dark energy (DE) problem, and so far three holographic DE models have been proposed: the original holographic dark energy (HDE) model, the agegraphic dark energy (ADE) model, and the holographic Ricci dark energy (RDE) model. In this work, we perform the best-fit analysis on these three models, by using the latest observational data including the Union+CFA3 sample of 397 Type Ia supernovae (SNIa), the shift parameter of the cosmic microwave background (CMB) given by the five-year Wilkinson Microwave Anisotropy Probe (WMAP5) observations, and the baryon acoustic oscillation (BAO) measurement from the Sloan Digital Sky Survey (SDSS). The analysis shows that for HDE, χ{sub min}{sup 2} = 465.912; for RDE, χ{sub min}{sup 2} = 483.130; for ADE, χ{sub min}{sup 2} = 481.694. Among these models, HDE model can give the smallest χ{sup 2}{sub min}. Besides, we also use the Bayesian evidence (BE) as a model selection criterion to make a comparison. It is found that for HDE, ADE, and RDE, Δln BE = −0.86, −5.17, and −8.14, respectively. So, it seems that the HDE model is more favored by the observational data.

79 ASTRONOMY AND ASTROPHYSICS↗

Accidental Father-to-Son HIV-1 Transmission During the Seroconversion Period

A 4-year-old child born to an HIV-1 seronegative mother was diagnosed with HIV-1, the main risk factor being transmission from the child's father who was seroconverting at the time of the child's birth. In the context of a forensic investigation, here we aimed to identify the source of infection of the child and date of the transmission event. Samples were collected from the father and child at two time points about 4 years after the child's birth. Partial segments of three HIV-1 genes (gag, pol, and env) were sequenced and maximum likelihood (ML) and Bayesian methods were used to determine direction and estimate date of transmission. Neutralizing antibodies were determined using a single cycle assay. Bayesian trees displayed a paraphyletic–monophyletic topology in all three genomic regions, with the father's host label at the root, which is consistent with father-to-son transmission. ML trees found similar topologies in gag and pol and a monophyletic–monophyletic topology in env. Analysis of the time of the most recent common ancestor of each HIV-1 gene population indicated that the child was infected shortly after the father. Consistent with the infection history, both father and son developed broad and potent HIV-specific neutralizing antibody responses. In conclusion, the direction of transmission implicated the father as the source of transmission. Transmission occurred during the seroconversion period when the father was unaware of the infection and was likely accidental. This case shows how genetic, phylogenetic, and serological data can contribute for the forensic investigation of HIV transmission.

60 APPLIED LIFE SCIENCES↗

Study of the top quark electric charge at the CDF experiment

We report on the measurement of the top quark electric charge using the jet charge tagging method on events containing a single lepton collected by the CDF II detector at Fermilab between February 2002 and February 2010 at the center-of-mass energy √s = 1.96 TeV. There are three main components to this measurement: determining the charge of the W (using the charge of the lepton), pairing the W with the b-jet to ensure that they are from the same top decay branch and finally determining the charge of the b-jet using the Jet Charge algorithm. We found, on a sample of 5.6 fb -1 of data, that the p-value under the standard model hypothesis is equal to 13.4%, while the p-value under the exotic model hypothesis is equal to 0.014%. Using the a priori criteria generally accepted by the CDF collaboration, we can say that the result is consistent with the standard model, while we exclude an exotic quark hypothesis with 95% confidence. Using the Bayesian approach, we obtain for the Bayes factor (2ln(BF)) a value of 19.6, that favors very strongly the SM hypothesis over the XM one. The presented method has the highest sensitivity to the top quark electric charge among the presented so far top quark charge analysis.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

In Situ Inference for Earth System Predictability

An understanding of future evolution in precipitation extremes is critical to numerous DOE mission questions. Extreme events are by nature short time-scale events that are difficult to diagnose in available model data. Accurate modeling of extreme events necessarily requires high spatial resolution at the storm scale locally. However, the environment in which storms grow is dependent on global, remote, processes. These complex spatiotemporal relationships are impossible to diagnose at resolutions required to accurately model storms responsible for extreme precipitation. At exascale, climate simulations will produce results at fine enough resolution to investigate these relationships. However, the resulting data from these simulations will be far too large to save for post-simulation analysis. We advocate for fitting statistical models inside the simulations as they run, a context known as in situ, which will facilitate scientific investigations using the full fine-scale data stream. Figure 1 shows an example of the type of model we could consider, a Bayesian hierarchical spatial regression model. Precipitation extremes at each grid cell are modeled using extreme value distributions. Since extremes are rare, fitting models to individual grid cells can result in high variance and poor estimates. Instead, the model can be made more robust by smoothing the parameters of the extreme value model across space. Additionally, the parameters themselves can be functionally linked to other variables elsewhere in the simulation. Thus, we can use the fine-scale data to build more robust models for extremes that link extreme behavior to other climate patterns.

54 ENVIRONMENTAL SCIENCES↗

In Situ Inference for Earth System Predictability

Focal Area: Focal Area 3: Insight gleaned from complex simulated data using AI, big data analytics, and other advanced methods, including explainable AI and physics- or knowledge-guided AI. Science Challenge: An understanding of future evolution in precipitation extremes is critical to numerous DOE mission questions. Extreme events are by nature short time-scale events that are difficult to diagnose in available model data. Accurate modeling of extreme events necessarily requires high spatial resolution at the storm scale locally. However, the environment in which storms grow is dependent on global, remote, processes. These complex spatiotemporal relationships are impossible to diagnose at resolutions required to accurately model storms responsible for extreme precipitation. At exascale, climate simulations will produce results at fine enough resolution to investigate these relationships. However, the resulting data from these simulations will be far too large to save for post-simulation analysis. We advocate for fitting statistical models inside the simulations as they run, a context known as in situ, which will facilitate scientific investigations using the full fine-scale data stream. Figure 1 shows an example of the type of model we could consider, a Bayesian hierarchical spatial regression model. Precipitation extremes at each grid cell are modeled using extreme value distributions. Since extremes are rare, fitting models to individual grid cells can result in high variance and poor estimates. Instead, the model can be made more robust by smoothing the parameters of the extreme value model across space. Additionally, the parameters themselves can be functionally linked to other variables elsewhere in the simulation. Thus, we can use the fine-scale data to build more robust models for extremes that link extreme behavior to other climate patterns.

54 ENVIRONMENTAL SCIENCES↗

Embedded Error Bayesian Calibration of Thermal Decomposition of Organic Materials

Organic materials are an attractive choice for structural components due to their light weight and versatility. However, because they decompose at low temperatures relative to tradiational materials they pose a safety risk due to fire and loss of structural integrity. To quantify this risk, analysts use chemical kinetics models to describe the material pyrolysis and oxidation using thermogravimetric analysis. This process requires the calibration of many model parameters to closely match experimental data. Previous efforts in this field have largely been limited to finding a single best-fit set of parameters even though the experimental data may be very noisy. Furthermore the chemical kinetics models are often simplified representations of the true de- composition process. The simplification induces model-form errors that the fitting process cannot capture. In this work we propose a methodology for calibrating decomposition models to thermogravimetric analysis data that accounts for uncertainty in the model-form and experimental data simultaneously. The methodology is applied to the decomposition of a carbon fiber epoxy composite with a three-stage reaction network and Arrhenius kinetics. The results show a good overlap between the model predictions and thermogravimetric analysis data. Uncertainty bounds capture devia- tions of the model from the data. The calibrated parameter distributions are also presented. In conclusion, the distributions may be used in forward propagation of uncertainty in models that leverage this material.

36 MATERIALS SCIENCE↗

Play fairway analysis of geothermal resources across the state of Hawaii: 2. Resource probability mapping

We develop a new geostatistical method to combine evidence provided by diverse geological data sets and produce maps of geothermal resource probability. The application is to the State of Hawaii, and the data sets include the locations and ages of mapped volcanic centers, gravity and magnetotelluric measurements, groundwater temperature and geochemistry, ground surface deformation, seismicity, water table elevation, and groundwater recharge. Using the basic principles of Bayesian statistics, these data and expert knowledge about the effects and importance of the data are used to compute the probabilities of the primary resource qualities of elevated subsurface heat, reservoir permeability, and reservoir fluid content. The product of these marginal probabilities estimates the joint probability of all three qualities and hence the probability of a successful geothermal prospect at each map point. An analogous set of algorithms is used to quantify the confidence in the probability at each point. Not surprisingly, we find that successful geothermal prospects are most probable on the active volcanoes of Hawaii Island, including the area of Hawaii’s single geothermal energy plant. Probability decreases primarily with shield volcano age, being relatively moderate in select locations on Maui and Lanai, relatively low on Oahu, and minimal on Kauai. Future exploration efforts should consider these results as well as the practical, societal, and economic conditions that influence development viability. Here, the difficulties of interisland power transmission mean that even areas with moderate to low probabilities are worth investigating on islands with population centers.

15 GEOTHERMAL ENERGY↗

Measuring the robustness of predictive probability for early stopping in two-group comparisons

We report physical experiments are often expensive and time-consuming. Test engineers must certify the compatibility of aircraft and their weapon systems before they can be deployed in the field, but the testing required is time consuming, expensive, and resource limited. Adopting Bayesian adaptive designs is a promising way to borrow from the successes seen in the clinical trials domain. The use of predictive probability (PP) to stop testing early and make faster decisions is particularly appealing given the aforementioned constraints. Given the high-consequence nature of the tests performed in the national security space, a strong understanding of new methods is required before being deployed. Although PP has been thoroughly studied for binary data, there is less work with continuous data, where many reliability studies are interested in certifying the specification limits of components. A simulation study evaluating the robustness of this approach indicates early stopping based on PP is reasonably robust to minor assumption violations, especially when only a few interim analyses are conducted. The simulation study also compares PP to conditional power, showing its relative strengths and weaknesses. A post-hoc analysis exploring whether release requirements of a weapon system from an aircraft are within specification with desired reliability resulted in stopping the experiment early and saving 33% of the experimental runs.

97 MATHEMATICS AND COMPUTING↗

Attributing runoff changes to climate variability and human activities: Uncertainty analysis using four monthly water balance models

Hydrological simulations to delineate the impacts of climate variability and human activities are subjected to uncertainties related to both parameter and structure of the hydrological models. To analyze the impact of these uncertainties on the model performance and to yield more reliable simulation results, a global calibration and multimodel combination method that integrates the Shuffled Complex Evolution Metropolis (SCEM) and Bayesian Model Averaging (BMA) of four monthly water balance models was proposed. The method was applied to the Weihe River Basin (WRB), the largest tributary of the Yellow River, to determine the contribution of climate variability and human activities to runoff changes. The change point, which was used to determine the baseline period (1956-1990) and human-impacted period (1991-2009), was derived using both cumulative curve and Pettitt’s test. Results show that the combination method from SCEM provides more skillful deterministic predictions than the best calibrated individual model, resulting in the smallest uncertainty interval of runoff changes attributed to climate variability and human activities. This combination methodology provides a practical and flexible tool for attribution of runoff changes to climate variability and human activities by hydrological models.

runoff change, uncertainty assessment, monthly wat↗