Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “sparse sampling”

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 631 records · Page 35
All sourcesNASA NTRSDOE OSTI
NASA NTRS1994

MME-based attitude dynamics identification and estimation for SAMPEX

A method is described for obtaining optimal attitude estimation algorithms for spacecraft lacking attitude rate measurement devices (rate gyros), and then demonstrated using actual flight data from the Solar, Anomalous, and Magnetospheric Particle Explorer (SAMPEX) spacecraft. SAMPEX does not have on-board rate sensing, and relies on sun sensors and a three-axis magnetometer for attitude determination. Problems arise since typical attitude estimation is accomplished by filtering measurements of both attitude and attitude rates. Rates are nearly always sampled much more densely than are attitudes. Thus, the absence/loss of rate data normally reduces both the total amount of data available and the sampling density (in time) by a substantial fraction. As a result, the sensitivity of the estimates to model uncertainty and to measurement noise increases. In order to maintain accuracy in the attitude estimates, there is increased need for accurate models of the rotational dynamics. The proposed approach is based on the minimum model error (MME) optimal estimation strategy, which has been successfully applied to estimation of poorly modeled dynamic systems which are relatively sparsely and/or noisily measured. The MME estimates may be used to construct accurate models of the system dynamics (i.e. perform system model identification). Thus, an MME-based approach directly addresses the problems created by absence of attitude rate measurements.

Depena, Juan↗
DOE OSTI2023

Estimation of hydraulic conductivity in a watershed using sparse multi-source data via Gaussian process regression and Bayesian experimental design

Enhanced water management systems depend on accurate estimation of subsurface hydraulic properties. However, geologic formations can vary significantly, so information from a single source (e.g., widely spaced boreholes) is insufficient in characterizing subsurface aquifer properties. Therefore, multiple sources of information are needed to complement the hydrogeology understanding of a region. Here, this study presents a numerical framework in which information from different measurement sources is combined to characterize the 3D random field in a multi-fidelity prediction model. Coupled with the model, a Bayesian experimental design was used to determine the best future sampling locations. The Upper Sangamon watershed in east-central Illinois was selected as the case study site, where the multi-fidelity Gaussian process model was used to estimate the hydraulic conductivity in the region of interest. Multi-source observation data were obtained from electrical resistivity and borehole pumping tests. The accuracy of the model prediction is dependent on the locations and the distribution of both high- and low-fidelity data. Furthermore, the multi-fidelity model was compared with the single-fidelity model. The uncertainties and confidence in the measurements and parameter estimates were quantified and used to design future cycles of data collection to further improve the confidence intervals.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2015

OPEN CLUSTERS AS PROBES OF THE GALACTIC MAGNETIC FIELD. I. CLUSTER PROPERTIES

Stars in open clusters are powerful probes of the intervening Galactic magnetic field via background starlight polarimetry because they provide constraints on the magnetic field distances. We use 2MASS photometric data for a sample of 31 clusters in the outer Galaxy for which near-IR polarimetric data were obtained to determine the cluster distances, ages, and reddenings via fitting theoretical isochrones to cluster color–magnitude diagrams. The fitting approach uses an objective χ{sup 2} minimization technique to derive the cluster properties and their uncertainties. We found the ages, distances, and reddenings for 24 of the clusters, and the distances and reddenings for 6 additional clusters that were either sparse or faint in the near-IR. The derived ranges of log(age), distance, and E(B−V) were 7.25–9.63, ∼670–6160 pc, and 0.02–1.46 mag, respectively. The distance uncertainties ranged from ∼8% to 20%. The derived parameters were compared to previous studies, and most cluster parameters agree within our uncertainties. To test the accuracy of the fitting technique, synthetic clusters with 50, 100, or 200 cluster members and a wide range of ages were fit. These tests recovered the input parameters within their uncertainties for more than 90% of the individual synthetic cluster parameters. These results indicate that the fitting technique likely provides reliable estimates of cluster properties. The distances derived will be used in an upcoming study of the Galactic magnetic field in the outer Galaxy.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

An Online Approach to Solve the Dynamic Vehicle Routing Problem with Stochastic Trip Requests for Paratransit Services

Many transit agencies operating paratransit and microtransit services have to respond to trip requests that arrive in real-time, which entails solving hard combinatorial and sequential decision-making problems under uncertainty. To avoid decisions that lead to significant inefficiency in the long term, vehicles should be allocated to requests by optimizing a non-myopic utility function or by batching requests together and optimizing a myopic utility function. While the former approach is typically offline, the latter can be performed online. We point out two major issues with such approaches when applied to paratransit services in practice. First, it is difficult to batch paratransit requests together as they are temporally sparse. Second, the environment in which transit agencies operate changes dynamically (e.g., traffic conditions can change over time), causing the estimates that are learned offline to become stale. To address these challenges, we propose a fully online approach to solve the dynamic vehicle routing problem (DVRP) with time windows and stochastic trip requests that is robust to changing environmental dynamics by construction. We focus on scenarios where requests are relatively sparse—our problem is motivated by applications to paratransit services. We formulate DVRP as a Markov decision process and use Monte Carlo tree search to evaluate actions for any given state. Accounting for stochastic requests while optimizing a non-myopic utility function is computationally challenging; indeed, the action space for such a problem is intractably large in practice. To tackle the large action space, we leverage the structure of the problem to design heuristics that can sample promising actions for the tree search. Our experiments using real-world data from our partner agency show that the proposed approach outperforms existing state-of-the-art approaches both in terms of performance and robustness.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
NASA NTRS2010

Transfer of Satellite Rainfall Uncertainty from Gauged to Ungauged Regions at Regional and Seasonal Timescales

Hydrologists and other users need to know the uncertainty of the satellite rainfall data sets across the range of time/space scales over the whole domain of the data set. Here, uncertainty' refers to the general concept of the deviation' of an estimate from the reference (or ground truth) where the deviation may be defined in multiple ways. This uncertainty information can provide insight to the user on the realistic limits of utility, such as hydrologic predictability, that can be achieved with these satellite rainfall data sets. However, satellite rainfall uncertainty estimation requires ground validation (GV) precipitation data. On the other hand, satellite data will be most useful over regions that lack GV data, for example developing countries. This paper addresses the open issues for developing an appropriate uncertainty transfer scheme that can routinely estimate various uncertainty metrics across the globe by leveraging a combination of spatially-dense GV data and temporally sparse surrogate (or proxy) GV data, such as the Tropical Rainfall Measuring Mission (TRMM) Precipitation Radar and the Global Precipitation Measurement (GPM) mission Dual-Frequency Precipitation Radar. The TRMM Multi-satellite Precipitation Analysis (TMPA) products over the US spanning a record of 6 years are used as a representative example of satellite rainfall. It is shown that there exists a quantifiable spatial structure in the uncertainty of satellite data for spatial interpolation. Probabilistic analysis of sampling offered by the existing constellation of passive microwave sensors indicate that transfer of uncertainty for hydrologic applications may be effective at daily time scales or higher during the GPM era. Finally, a commonly used spatial interpolation technique (kriging), that leverages the spatial correlation of estimation uncertainty, is assessed at climatologic, seasonal, monthly and weekly timescales. It is found that the effectiveness of kriging is sensitive to the type of uncertainty metric, time scale of transfer and the density of GV data within the transfer domain. Transfer accuracy is lowest at weekly timescales with the error doubling from monthly to weekly.However, at very low GV data density (<20% of the domain), the transfer accuracy is too low to show any distinction as a function of the timescale of transfer.

Tang, Ling↗
DOE OSTI2017

Quantum algorithms for Gibbs sampling and hitting-time estimation

In this paper, we present quantum algorithms for solving two problems regarding stochastic processes. The first algorithm prepares the thermal Gibbs state of a quantum system and runs in time almost linear in √Nβ/Ζ and polynomial in log(1/ϵ), where N is the Hilbert space dimension, β is the inverse temperature, Ζ is the partition function, and ϵ is the desired precision of the output state. Our quantum algorithm exponentially improves the dependence on 1/ϵ and quadratically improves the dependence on β of known quantum algorithms for this problem. The second algorithm estimates the hitting time of a Markov chain. For a sparse stochastic matrix Ρ, it runs in time almost linear in 1/(ϵΔ 3/2 ), where ϵ is the absolute precision in the estimation and Δ is a parameter determined by Ρ, and whose inverse is an upper bound of the hitting time. Our quantum algorithm quadratically improves the dependence on 1/ϵ and 1/Δ of the analog classical algorithm for hitting-time estimation. Finally, both algorithms use tools recently developed in the context of Hamiltonian simulation, spectral gap amplification, and solving linear systems of equations.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Preserving nonlinear constraints in variational flow filtering data assimilation

Data assimilation aims to estimate the states of a dynamical system by optimally combining sparse and noisy observations of the physical system with uncertain forecasts produced by a computational model. The states of many dynamical systems of interest obey nonlinear physical constraints, and the corresponding dynamics is confined to a certain sub-manifold of the state space. Standard data assimilation techniques applied to such systems yield posterior states lying outside the manifold, violating the physical constraints. This work focuses on particle flow filters which use stochastic differential equations to evolve state samples from a prior distribution to samples from an observation-informed posterior distribution. The variational Fokker-Planck (VFP)—a generic particle flow filtering framework—is extended to incorporate non-linear, equality state constraints in the analysis. To this end, two algorithmic approaches that modify the VFP stochastic differential equation are discussed: (i) VFPSTAB, to inexactly preserve constraints with the addition of a stabilizing drift term, and (ii) VFPDAE, to exactly preserve constraints by treating the VFP dynamics as a stochastic differential-algebraic equation (SDAE). Additionally, an implicit-explicit time integrator is developed to evolve the VFPDAE dynamics. The strength of the proposed approach for constraint preservation in data assimilation is demonstrated on three test problems: the double pendulum, Korteweg-de-Vries, and the incompressible Navier-Stokes equations.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

SUPERLUMINOUS SPIRAL GALAXIES

We report the discovery of spiral galaxies that are as optically luminous as elliptical brightest cluster galaxies, with r-band monochromatic luminosity L{sub r} = 8–14L* (4.3–7.5 × 10{sup 44} erg s{sup −1}). These super spiral galaxies are also giant and massive, with diameter D = 57–134 kpc and stellar mass M{sub stars} = 0.3–3.4 × 10{sup 11}M{sub ⊙}. We find 53 super spirals out of a complete sample of 1616 SDSS galaxies with redshift z < 0.3 and L{sub r} > 8L*. The closest example is found at z = 0.089. We use existing photometry to estimate their stellar masses and star formation rates (SFRs). The SDSS and Wide-field Infrared Survey Explorer colors are consistent with normal star-forming spirals on the blue sequence. However, the extreme masses and rapid SFRs of 5–65 M{sub ⊙} yr{sup −1} place super spirals in a sparsely populated region of parameter space, above the star-forming main sequence of disk galaxies. Super spirals occupy a diverse range of environments, from isolation to cluster centers. We find four super spiral galaxy systems that are late-stage major mergers—a possible clue to their formation. We suggest that super spirals are a remnant population of unquenched, massive disk galaxies. They may eventually become massive lenticular galaxies after they are cut off from their gas supply and their disks fade.

79 ASTRONOMY AND ASTROPHYSICS↗
NASA NTRS1982

Output statistics of laser anemometers in sparsely seeded flows

It is noted that until very recently, research on this topic concentrated on the particle arrival statistics and the influence of the optical parameters on them. Little attention has been paid to the influence of subsequent processing on the measurement statistics. There is also controversy over whether the effects of the particle statistics can be measured. It is shown here that some of the confusion derives from a lack of understanding of the experimental parameters that are to be controlled or known. A rigorous framework is presented for examining the measurement statistics of such systems. To provide examples, two problems are then addressed. The first has to do with a sample and hold processor, the second with what is called a saturable processor. The sample and hold processor converts the output to a continuous signal by holding the last reading until a new one is obtained. The saturable system is one where the maximum processable rate is arrived at by the dead time of some unit in the system. At high particle rates, the processed rate is determined through the dead time.

Edwards, R. V.↗
NASA NTRS2017

Effective Use Of Scatterometer Winds In Current and Future GMAO Reanalysis

Scatterometer-derived near-surface ocean vector wind retrievals provide global measurements complementary to the sparse conventional observing system which primarily consists of ships and buoys over water surfaces. The RapidScat instrument was flown on the International Space Station as a quick and low cost replacement of QuikScat and as a continuation of the NASA scatterometry data record. A unique characteristic of RapidScat was that it flew in a non-sun synchronous orbit at an inclination of 51.6 degrees. This orbit allowed for the collocation of measurements with other scatterometers as well as an ability to sample diurnal signals. In the Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2) reanalysis, the scatterometry record began with the ESA European Remote Sensing (ERS) scatterometer on 5 Aug 1991 and continued through today with the EUMETSAT Metop Advanced Scatterometer (ASCAT). RapidScat, however, was not used in the MERRA-2 system as development had been completed prior to the beginning of its data record. In this presentation, the RapidScat ocean vector winds will be compared to MERRA-2, both in terms of the analysis fields and in the context of its global observing system, to assess the viability of using the data in future reanalysis systems developed by the Global Modeling and Assimilation Office (GMAO) at NASA Goddard Space Flight Center.

Chattopadhyay, Mohar↗
NASA NTRS2007

NASA Tech Briefs, January 2007

Topics covered include: Flexible Skins Containing Integrated Sensors and Circuitry; Artificial Hair Cells for Sensing Flows; Video Guidance Sensor and Time-of-Flight Rangefinder; Optical Beam-Shear Sensors; Multiple-Agent Air/Ground Autonomous Exploration Systems; A 640 512-Pixel Portable Long-Wavelength Infrared Camera; An Array of Optical Receivers for Deep-Space Communications; Microstrip Antenna Arrays on Multilayer LCP Substrates; Applications for Subvocal Speech; Multiloop Rapid-Rise/Rapid Fall High-Voltage Power Supply; The PICWidget; Fusing Symbolic and Numerical Diagnostic Computations; Probabilistic Reasoning for Robustness in Automated Planning; Short-Term Forecasting of Radiation Belt and Ring Current; JMS Proxy and C/C++ Client SDK; XML Flight/Ground Data Dictionary Management; Cross-Compiler for Modeling Space-Flight Systems; Composite Elastic Skins for Shape-Changing Structures; Glass/Ceramic Composites for Sealing Solid Oxide Fuel Cells; Aligning Optical Fibers by Means of Actuated MEMS Wedges; Manufacturing Large Membrane Mirrors at Low Cost; Double-Vacuum-Bag Process for Making Resin- Matrix Composites; Surface Bacterial-Spore Assay Using Tb3+/DPA Luminescence; Simplified Microarray Technique for Identifying mRNA in Rare Samples; High-Resolution, Wide-Field-of-View Scanning Telescope; Multispectral Imager With Improved Filter Wheel and Optics; Integral Radiator and Storage Tank; Compensation for Phase Anisotropy of a Metal Reflector; Optical Characterization of Molecular Contaminant Films; Integrated Hardware and Software for No-Loss Computing; Decision-Tree Formulation With Order-1 Lateral Execution; GIS Methodology for Planning Planetary-Rover Operations; Optimal Calibration of the Spitzer Space Telescope; Automated Detection of Events of Scientific Interest; Representation-Independent Iteration of Sparse Data Arrays; Mission Operations of the Mars Exploration Rovers; and More About Software for No-Loss Computing.

Source record↗
NASA NTRSDate not supplied

Lessons from Earth Bioaerosol Sampling for Venus Habitability & Biosignature Science

Venus’s clouds are often discussed as a potential habitat. Several high-priority science goals – cloud aerosol composition, internal radiative flux, and aerosol residence time and circulation models – will improve our understanding of Venus in an astrobiology context. Lessons from Earth’s aerobiosphere can inform these requirements and interpretation of results. A stable Venus aerobiosphere would need microorganisms to reproduce faster than they settle out due to gravity. This is a joint constraint of the potential nutrients and energy influx, bioenergetic costs such as desiccation and radiation damage, and aerosol dynamics. Though life cycles involving S- and Fe-based redox metabolism have been proposed, at an optimistic estimate of 75% H2SO4, Venus aerosol water activity (aw) is still ~0.02, far below the observed microbial growth limit of ~0.6. Long-term desiccation with brief spurts of repair, growth and reproduction in response to transient water influx (perhaps volcanism) is the most likely model for Earth-like life on Venus – a ‘desert bloom’ scenario. In Earth’s troposphere, clouds can carry 103 – 105 cells/mL, some metabolically active. A better analogue may be Earth’s stratospheric sulfate aerosol layer: supercooled sulfuric acid aerosols (acid weight fraction 0.6 – 0.85, 0.1 – 1 µm diameter, 0.1 – 1 cm-3) with little water activity, long residence times, high UV radiation, and only sporadic influx from surface particle sources. Viable cells isolated from this region are rare (~102 cells/m3), though ‘hot spots’ at tropospheric mixing occur. Cells are primarily in inactive forms such as spores, and it is unclear if any are associated with sulfate aerosols or simply co-located. Given the heterogeneities in Venus clouds and the model of a sparse, largely dormant ecosystem, a single transect on descent would most likely pass through a low-water, inactive region, missing potential signs of habitability or biosignatures. Even a passive aerial platform (balloon) may end up following a single air mass over its lifetime. A better strategy for Venus would target an aerial region with some upwelling from surface sources, and take multiple transects separated in time and space. Earth’s stratospheric bioaerosol population could provide a starting point for estimating the required lower volume and detection limits.

Earth↗
DOE OSTI2025

The imprint of cosmic voids from the DESI Legacy Survey DR9 Luminous Red Galaxies in the Planck 2018 lensing map through spectroscopically calibrated mocks

The cross-correlation of cosmic voids with the lensing convergence (κ) map of the Cosmic Microwave Background (CMB) fluctuations provides a powerful tool to refine our understanding of the current cosmological model. However, several studies have reported a moderate tension (up to ∼2σ) between the lensing imprint of cosmic voids on the observed CMB and the ΛCDM signal predicted by simulations. To address this “lensing-is-low” tension and to obtain new, precise measurements of the signal, we exploit the large DESI Legacy Survey Luminous Red Galaxy (LRG) data set, covering approximately 19 500 deg 2 of the sky and including about 10 million LRGs at z < 1.05. Our ΛCDM template was created using the Buzzard mocks, which we specifically calibrated to match the clustering properties of the observed galaxy sample by exploiting more than one million DESI spectra. We identified our catalogs of 3D voids in the range 0.35 < z < 0.95 and cross-correlated them through a stacking methodology, dividing the sample into bins according to the redshift and λv values of the voids. For the full void sample, we report a 14σ detection of the lensing signal, with A κ = 1.016 ± 0.054, which increases to 17σ when considering the void-in-void (A κ = 0.944 ± 0.064) and the void-in-cloud (A κ = 0.975 ± 0.060) populations individually, the highest detection significance for studies of this kind. We observe a full agreement between observations and ΛCDM mocks across all redshift bins, sky regions, and void populations considered. In addition to these findings, our analysis highlights the importance of accurately matching sparseness and redshift error distributions between mocks and observations, as well as the role of λ v in enhancing the signal-to-noise ratio through void population discrimination.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2020

LiDAR-Aided Interior Orientation Parameters Refinement Strategy for Consumer-Grade Cameras Onboard UAV Remote Sensing Systems

Unmanned aerial vehicles (UAVs) are quickly emerging as a popular platform for 3D reconstruction/modeling in various applications such as precision agriculture, coastal monitoring, and emergency management. For such applications, LiDAR and frame cameras are the two most commonly used sensors for 3D mapping of the object space. For example, point clouds for the area of interest can be directly derived from LiDAR sensors onboard UAVs equipped with integrated global navigation satellite systems and inertial navigation systems (GNSS/INS). Imagery-based mapping, on the other hand, is considered to be a cost-effective and practical option and is often conducted by generating point clouds and orthophotos using structure from motion (SfM) techniques. Mapping with photogrammetric approaches requires accurate camera interior orientation parameters (IOPs), especially when direct georeferencing is utilized. Most state-of-the-art approaches for determining/refining camera IOPs depend on ground control points (GCPs). However, establishing GCPs is expensive and labor-intensive, and more importantly, the distribution and number of GCPs are usually less than optimal to provide adequate control for determining and/or refining camera IOPs. Moreover, consumer-grade cameras with unstable IOPs have been widely used for mapping applications. Therefore, in such scenarios, where frequent camera calibration or IOP refinement is required, GCP-based approaches are impractical. To eliminate the need for GCPs, this study uses LiDAR data as a reference surface to perform in situ refinement of camera IOPs. The proposed refinement strategy is conducted in three main steps. An image-based sparse point cloud is first generated via a GNSS/INS-assisted SfM strategy. Then, LiDAR points corresponding to the resultant image-based sparse point cloud are identified through an iterative plane fitting approach and are referred to as LiDAR control points (LCPs). Finally, IOPs of the utilized camera are refined through a GNSS/INS-assisted bundle adjustment procedure using LCPs. Seven datasets over two study sites with a variety of geomorphic features are used to evaluate the performance of the developed strategy. The results illustrate the ability of the proposed approach to achieve an object space absolute accuracy of 3–5 cm (i.e., 5–10 times the ground sampling distance) at a 41 m flying height.

47 OTHER INSTRUMENTATION↗
DOE OSTI2019

Application of a new hybrid particle swarm optimization-mixed kernels function-based support vector machine model for reservoir porosity prediction: A case study in Jacksonburg-Stringtown oil field, West Virginia, USA

Porosity is a fundamental property that characterizes the storage capability of fluid and gas-bearing formations in a reservoir. An accurate porosity value can be measured from core samples in the laboratory; however, core analysis is expensive and time consuming. Well-log data can be used to calculate porosity, but the availability of log suites is often limited in mature fields. Therefore, robust porosity prediction requires integration of core-measured porosity with available well-log suites to control for changes in lithology and fluid content. A support vector machine (SVM) model with mixed kernel function (MKF) is used to construct the relationship between limited conventional well-log suites and sparse core data. Porosity is the desired output, and two conventional well-log responses (gamma ray [GR] and bulk density) and three well-log-derived parameters (the slope of GR, the slope of density, and [Formula: see text]) are input parameters. A global stochastic searching algorithm, particle swarm optimization (PSO), is applied to improve the efficiency of locating the appropriate values of five control parameters in MKF-SVM model. The results of SVM with different traditional kernel functions were compared, and the MKF-SVM model provided an improvement over the traditional SVM model. To confirm the advantage of the hybrid PSO-MKF-SVM model, the results from three models: (1) radial basis function (RBF)-based least-squares SVM, (2) multilayer perceptron artificial neural network (ANN), and (3) RBF ANN, are compared with the result of the hybrid PSO-MKF-SVM model. The results indicate that the hybrid PSO-MKF-SVM model improves porosity prediction with the highest correlation coefficient ([Formula: see text] of 0.9560), the highest coefficient of determination ([Formula: see text] of 0.9140), the lowest root-mean-square error (1.6505), average absolute error value (1.4050), and maximum absolute error (2.717).

Geochemistry & Geophysics↗
DOE OSTI2021

Geo Thermal Cloud: Cloud Fusion of Big Data and Multi-Physics Models using Machine Learning for Discovery, Exploration, and Development of Hidden Geothermal Resources

The project is motivated by the challenges, risks, and costs associated with geothermal exploration and production. Many processes and parameters impacting geothermal conditions are poorly understood. Diverse datasets are available to help characterize subsurface geothermal conditions (public and proprietary; satellite, airborne surveys, vegetation/water sampling, geological, geophysical, etc.). Yet, it is not clear how to properly leverage these datasets for geothermal exploration due to an incomplete understanding of how physical processes impacting subsurface geothermal conditions are represented in these observations. Recent advancements in machine learning (ML) provide great promise to resolve these issues. The tremendous challenges and risks of geothermal exploration and production bring the demand for novel ML methods and tools that can (1) analyze large field datasets, (2) assimilate model simulations (large inputs and outputs), (3) process sparse datasets, (4) perform transfer learning (between sites with different exploratory levels), (5) extract hidden geothermal signatures in the field and simulation data, (6) label geothermal resources and processes, (7) identify high-value data acquisition targets, and (8) guide geothermal exploration and production by selecting optimal exploration, production, and drilling strategies. Our goals and work under Phases 1 and 2 (as proposed) of this project address all these needs.

15 GEOTHERMAL ENERGY↗
DOE OSTI2020

How Low Can You Go? Using Synthetic 3D Imagery to Drastically Reduce Real-World Training Data for Object Detection

Deep convolutional neural networks (DCNNs) currently provide state-of-the-art performance on image classification and object detection tasks, and there are many global security mission areas where such models could be extremely useful. Crucially, the success of these models is driven in large part by the widespread availability of high-quality open source data sets such as Image Net, Common Objects in Context (COCO), and KITTI, which contain millions of images with thousands of unique labels. However, global security relevant objects-of-interest can be difficult to obtain: relevant events are low frequency and high consequence; the content of relevant images is sensitive; and adversaries and proliferators seek to obscure their activities. For these cases where exemplar data is hard to come-by, even fine-tuning an existing model with available data can be effectively impossible. Recent work demonstrated that models can be trained using a combination of real-world and synthetic images generated from 3D representations; that such models can exceed the performance of models trained using real-world data alone; and that the generated images need not be perfectly realistic (Tremblay, et al., 2018). However, this approach still required hundreds to thousands of real-world images for training and fine tuning, which for sparse, global security-relevant datasets can be an unrealistic hurdle. In this research, we validate the performance and behavior of DCNN models as we drive the number of real-world images used for training object detection tasks down to a minimal set. We perform multiple experiments to identify the best approach to train DCNNs from an extremely small set of real-world images. In doing so, we: Develop state-of-the-art, parameterized 3D models based on real-world images and sample from their parameters to increase the variance in synthetic image training data; Use machine learning explainability techniques to highlight and correct through targeted training the biases that result from training using completely synthetic images; and Validate our results by comparing the performance of the models trained on synthetic data to one another, and to a control model created by fine-tuning an existing ImageNet-trained model with a limited number (hundreds) of real-world images.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Adversarial methods to reduce simulation bias in neutrino interaction event filtering at liquid argon time projection chambers

For current and future neutrino oscillation experiments using large liquid argon time projection chambers (LAr-TPCs), a key challenge is identifying neutrino interactions from the pervading cosmic-ray background. Rejection of such background is often possible using traditional cut-based selections, but this typically requires the prior use of computationally expensive reconstruction algorithms. This work demonstrates an alternative approach of using a 3D submanifold sparse convolutional network trained on low-level information from the scintillation light signal of interactions inside LAr-TPCs. This technique is applied to example simulations from ICARUS, the far detector of the short baseline neutrino program at Fermilab. The results of the network, show that cosmic background is reduced by up to 76.3% whilst neutrino interaction selection efficiency remains over 98.9%. We further present a way to mitigate potential biases from imperfect input simulations by applying domain adversarial neural networks (DANNs), for which modified simulated samples are introduced to imitate real data and a small portion of them are used for adversarial training. A series of mock-data studies are performed and demonstrate the effectiveness of using DANNs to mitigate biases, showing neutrino interaction selection efficiency performances significantly better than that achieved without the adversarial training.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗