SEARCH · Search NASA

Results for “synthetic data sets”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 235 records · Page 13

All sources NASA NTRS DOE OSTI

DOE OSTI2024

Bayesian inference of structured latent spaces from neural population activity with the orthogonal stochastic linear mixing model

The brain produces diverse functions, from perceiving sounds to producing arm reaches, through the collective activity of populations of many neurons. Determining if and how the features of these exogenous variables (e.g., sound frequency, reach angle) are reflected in population neural activity is important for understanding how the brain operates. Often, high-dimensional neural population activity is confined to low-dimensional latent spaces. However, many current methods fail to extract latent spaces that are clearly structured by exogenous variables. This has contributed to a debate about whether or not brains should be thought of as dynamical systems or representational systems. Here, we developed a new latent process Bayesian regression framework, the orthogonal stochastic linear mixing model (OSLMM) which introduces an orthogonality constraint amongst time-varying mixture coefficients, and provide Markov chain Monte Carlo inference procedures. We demonstrate superior performance of OSLMM on latent trajectory recovery in synthetic experiments and show superior computational efficiency and prediction performance on several real-world benchmark data sets. We primarily focus on demonstrating the utility of OSLMM in two neural data sets: μ ECoG recordings from rat auditory cortex during presentation of pure tones and multi-single unit recordings form monkey motor cortex during complex arm reaching. We show that OSLMM achieves superior or comparable predictive accuracy of neural data and decoding of external variables (e.g., reach velocity). Most importantly, in both experimental contexts, we demonstrate that OSLMM latent trajectories directly reflect features of the sounds and reaches, demonstrating that neural dynamics are structured by neural representations. Together, these results demonstrate that OSLMM will be useful for the analysis of diverse, large-scale biological time-series datasets.

59 BASIC BIOLOGICAL SCIENCES↗

NASA NTRS1989

Adaptation of the ISCCP cloud detection algorithm to combined AVHRR and SMMR arctic data

The International Satellite Cloud Climatology Project (ISCCP) cloud detection algorithm is applied to artic data, and modifications are suggested. Both Advanced Very High Resolution Radiometer (AVHRR) and Scanning Multichannel Microwave Radiometer (SMMR) data are examined. Synthetic AVHRR and SMMR data are also generated. Modifications suggested include the use of snow and ice data sets for the estimation of surface parameters, additional AVHRR channels, and surface class characteristic values when clear sky values cannot be obtained. Greatest improvement in computed cloud fraction is realized over snow and ice surfaces; over other surfaces all versions perform similarly. Since the use of SMMR for surface analysis increases the computational burden, its use may be justified only over snow and ice-covered regions.

Key, J.↗

NASA NTRS1992

More About Calibration Of Polarimetric SAR

Azimuthal symmetry is not assumed. Two sets of equations and algorithms for radiometric and polarimetric calibration of complex quadpolarization data gathered by synthetic-aperture radar (SAR) based partly on known scattering characteristics of artificial targets in scene, and partly on assumption that all other targets scatter reciprocally.

Klein, Jeffrey D.↗

NASA NTRS1992

Interaction types and their like-polarization phase-angle difference signatures

The JPL Airborne Synthetic Aperture Radar (AIRSAR) measures the full polarimetric properties of scene objects. AIRSAR data contain a complete set of magnitudes (for all linearly-polarized combinations (viz., VV, HH, and CS (cross Polarization))) and phase-angle differences (for all pairs of linearly-polarized components) for each image pixel. Radar images users have been greatly interested in the like-polarization (LK) phase-angle difference parameter, delta phi(sub LK). Usually, it is associated with the type of interaction between microwaves and scene-object scattering-elements. One value of delta phi(sub LK) is said to be produced by a dominant single scattering (or reflection) event, and another value is said to be the result of a dominant double-scattering (or double reflection) event. these are often called 'single bounce' and 'double bounce' interaction signatures, respectively. There appears to be a significant amount of confusion about how the interaction processes produce the characteristic values of delta phi(sub LK) and about what value of delta phi(sub LK) should be associated with each specific type. This issue is addressed.

Paris, Jack F.↗

NASA NTRSDate not supplied

Aeroacoustic Computations of a Generic Low Boom Concept in Landing Configuration: Part 3 - Aerodynamic Validation and Noise Source Identification

Wind tunnel test data were used to validate the predicted aerodynamic behavior of a 15%-scale version of a generic, low-boom aircraft. The test was conducted in the NASA Langley 14- by 22-Foot Subsonic Tunnel to determine the low-speed aerodynamic characteristics of the model. Measured steady surface pressures and global forces were used to validate predicted aerodynamic results obtained from high-fidelity simulations of the model as installed in the tunnel. Very good agreement between predicted and measured aerodynamic trends was demonstrated, providing the impetus to proceed with companion airframe noise simulations that were conducted in a free-air setting. Computed near-field flow variables acquired on a permeable data surface were used to generate synthetic pressure records for two 800-element phased microphone arrays positioned overhead and to the side of the model. The array data were beamformed to generate noise source localization maps for the clean model and several landing configurations. Primary and secondary airframe sources were identified and their relative strengths were determined. Far-field integrated noise spectra for the full aircraft, as well as individual components, were obtained from the source maps via integration of tailored regions. The analysis showed that noise produced by the landing gear was the dominant contributor to the far-field acoustic signature of the model, followed by flap noise. The effects of permeable data surface end caps, spatial resolution, array orientation, angle of attack, component interaction, velocity scaling, and numerical precision on synthetic far-field spectra were also evaluated.

low boom↗

NASA NTRS1983

The absolute amplitude calibration of the SEASAT synthetic aperture radar - An intercomparison with other L-band radar systems

The absolute amplitude calibration of the spaceborne Seasat SAR data set is presented based on previous relative calibration studies. A scale factor making it possible to express the perceived radar brightness of a scene in units of sigma-zero is established. The system components are analyzed for error contribution, and the calibration techniques are introduced for each stage. These include: A/D converter saturation tests; prevention of clipping in the processing step; and converting the digital image into the units of received power. Experimental verification was performed by screening and processing the data of the lava flow surrounding the Pisgah Crater in Southern California, for which previous C-130 airborne scatterometer data were available. The average backscatter difference between the two data sets is estimated to be 2 dB in the brighter, and 4 dB in the dimmer regions. For the SAR a calculated uncertainty of 3 dB is expected.

Held, D.↗

DOE OSTI2022

Deep Learning At Depth: Estimating subsurface parameters from geophysical monitoring data

Geophysical imaging techniques are a non-invasive way to image the subsurface and understand both subsurface solid (rock/soil) and fluid property distributions and their evolution in time. Inversions of the geophysical data, such as Electrical Resistance Tomography (ERT) data, are solved to estimate the subsurface property distributions, such as conductivity, and many inversion techniques smooth out sharp gradients in rock or fluid property distributions. Sharp gradients in subsurface properties tend to be present in situations with complex subsurface structures, which are common in many subsurface applications. We have successfully demonstrated that it is possible to inform, or constrain, inversions with neural networks trained on synthetic data with complex subsurface structures. Initial results suggest this process may be optimizable to yield property distributions that better represent the true property distributions than the same inversion process without the neural network constraint. Future work would optimize the neural network performance for this application and then apply the synthetic-data trained neural network to real data to understand the utility and performance of this technique for real data sets.

47 OTHER INSTRUMENTATION↗

NASA NTRS1981

The Seasat commercial demonstration program

The background and development of the Seasat commercial demonstration program are reviewed and the Seasat spacecraft and its sensors (altimeter, wind field scatterometer, synthetic aperture radar, and scanning multichannel microwave radiometer) are described. The satellite data distribution system allows for selected sets of data, reformatted or tailored to specific needs and geographical regions, to be available to commercial users. Products include sea level and upper atmospheric pressure, sea surface temperature, marine winds, significant wave heights, primary wave direction and period, and spectral wave data. The results of a set of retrospective case studies performed for the commercial demonstration program are described. These are in areas of application such as marine weather and ocean condition forecasting, offshore resource exploration and development, commercial fishing, and marine transportation.

Mccandless, S. W.↗

NASA NTRS1993

Techniques and Applications of SAR interferometry for ERS-1: Topographic Mapping, Change Detection, and Slope Measurement

SAR (Synthetic Aperture Radar) interferometry allows the measurement of high resolution topography of terrain from repeat orbit SAR data sets obtained by the ERS-1 AMI (Active Microwave Instrument). The system parameters which affect the accuracy of the topographic maps are errors in determination of the interferometric baseline, non parallel orbit tracks, decorrelation caused by baseline length, thermal noise, and surface change. Surface change can be observed in the interferograms either through decorrelation of the interferometric phase, or coherent phase shifts caused by locally uniform surface displacements. Phase gradient maps can be derived directly from the complex interferograms and can be transformed into surface slopes mapped onto a geometrically corrected grid.

Charles L Werner↗

DOE OSTI2024

Utilization of Synthetic Near-Infrared Spectra via Generative Adversarial Network to Improve Wood Stiffness Prediction

Near-infrared (NIR) spectroscopy is widely used as a nondestructive evaluation (NDE) tool for predicting wood properties. When deploying NIR models, one faces challenges in ensuring representative training data, which large datasets can mitigate but often at a significant cost. Machine learning and deep learning NIR models are at an even greater disadvantage because they typically require higher sample sizes for training. In this study, NIR spectra were collected to predict the modulus of elasticity (MOE) of southern pine lumber (training set = 573 samples, testing set = 145 samples). To account for the limited size of the training data, this study employed a generative adversarial network (GAN) to generate synthetic NIR spectra. The training dataset was fed into a GAN to generate 313, 573, and 1000 synthetic spectra. The original and enhanced datasets were used to train artificial neural networks (ANNs), convolutional neural networks (CNNs), and light gradient boosting machines (LGBMs) for MOE prediction. Overall, results showed that data augmentation using GAN improved the coefficient of determination (R 2 ) by up to 7.02% and reduced the error of predictions by up to 4.29%. ANNs and CNNs benefited more from synthetic spectra than LGBMs, which only yielded slight improvement. All models showed optimal performance when 313 synthetic spectra were added to the original training data; further additions did not improve model performance because the quality of the datapoints generated by GAN beyond a certain threshold is poor, and one of the main reasons for this can be the size of the initial training data fed into the GAN. LGBMs showed superior performances than ANNs and CNNs on both the original and enhanced training datasets, which highlights the significance of selecting an appropriate machine learning or deep learning model for NIR spectral-data analysis. The results highlighted the positive impact of GAN on the predictive performance of models utilizing NIR spectroscopy as an NDE technique and monitoring tool for wood mechanical-property evaluation. Further studies should investigate the impact of the initial size of training data, the optimal number of generated synthetic spectra, and machine learning or deep learning models that could benefit more from data augmentation using GANs.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2023

A Joint Roman Space Telescope and Rubin Observatory Synthetic Wide-Field Imaging Survey

We present and validate 20 deg(2) of overlapping synthetic imaging surveys representing the full depth of the Nancy Grace Roman Space Telescope High-Latitude Imaging Survey (HLIS) and 5 yr of observations of the Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST). The two synthetic surveys are summarized, with reference to the existing 300 deg(2) of LSST simulated imaging produced as part of Dark Energy Science Collaboration (DESC) Data Challenge 2 (DC2). Both synthetic surveys observe the same simulated DESC DC2 universe. For the synthetic Roman survey, we simulate for the first time fully chromatic images along with the detailed physics of the Sensor Chip Assemblies derived from lab measurements using the flight detectors. The simulated imaging and resulting pixel-level measurements of photometric properties of objects span a wavelength range of similar to 0.3-2.0 mu m. We also describe updates to the Roman simulation pipeline, changes in how astrophysical objects are simulated relative to the original DC2 simulations, and the resulting simulated Roman data products. We use these simulations to explore the relative fraction of unrecognized blends in LSST images, finding that 20-30 per cent of objects identified in LSST images with i-band magnitudes brighter than 25 can be identified as multiple objects in Roman images. These simulations provide a unique testing ground for the development and validation of joint pixel-level analysis techniques of ground- and space-based imaging data sets in the second half of the 2020s - in particular the case of joint Roman-LSST analyses.

gravitational lensing↗

DOE OSTI2023

Accelerating full-waveform inversion using source stacking: synthetic experiments at the global scale in a realistic 3-D earth model

SUMMARY The spectral element method is currently the method of choice for computing accurate synthetic seismic wavefields in realistic 3-D earth models at the global scale. However, it requires significantly more computational time, compared to normal mode-based approximate methods. Source stacking, whereby multiple earthquake sources are aligned on their origin time and simultaneously triggered, can reduce the computational costs by several orders of magnitude. We present the results of synthetic tests performed on a realistic radially anisotropic 3-D model, slightly modified from model SEMUCB-WM1 with three component synthetic waveform ‘data’ for a duration of 10 000 s, and filtered at periods longer than 60 s, for a set of 273 events and 515 stations. We consider two definitions of the misfit function, one based on the stacked records at individual stations and another based on station-pair cross-correlations of the stacked records. The inverse step is performed using a Gauss–Newton approach where the gradient and Hessian are computed using normal mode perturbation theory. We investigate the retrieval of radially anisotropic long wavelength structure in the upper mantle in the depth range 100–800 km, after fixing the crust and uppermost mantle structure constrained by fundamental mode Love and Rayleigh wave dispersion data. The results show good performance using both definitions of the misfit function, even in the presence of realistic noise, with degraded amplitudes of lateral variations in the anisotropic parameter ξ. Interestingly, we show that we can retrieve the long wavelength structure in the upper mantle, when considering one or the other of three portions of the cross-correlation time series, corresponding to where we expect the energy from surface wave overtone, fundamental mode or a mixture of the two to be dominant, respectively. We also considered the issue of missing data, by randomly removing a successively larger proportion of the available synthetic data. We replace the missing data by synthetics computed in the current 3-D model using normal mode perturbation theory. The inversion results degrade with the proportion of missing data, especially for ξ, and we find that a data availability of 45 per cent or more leads to acceptable results. We also present a strategy for grouping events and stations to minimize the number of missing data in each group. This leads to an increased number of computations but can be significantly more efficient than conventional single-event-at-a-time inversion. We apply the grouping strategy to a real picking scenario, and show promising resolution capability despite the use of fewer waveforms and uneven ray path distribution. Source stacking approach can be used to rapidly obtain a starting 3-D model for more conventional full-waveform inversion at higher resolution, and to investigate assumptions made in the inversion, such as trade-offs between isotropic, anisotropic or anelastic structure, different model parametrizations or how crustal structure is accounted for.

Geochemistry & Geophysics↗

DOE OSTI2024

Simulation-based inference for parameter estimation of complex watershed simulators

High-resolution, spatially distributed process-based (PB) simulators are widely employed in the study of complex catchment processes and their responses to a changing climate. However, calibrating these PB simulators using observed data remains a significant challenge due to several persistent issues, including the following: (1) intractability stemming from the computational demands and complex responses of simulators, which renders infeasible calculation of the conditional probability of parameters and data, and (2) uncertainty stemming from the choice of simplified representations of complex natural hydrologic processes. Here, we demonstrate how simulation-based inference (SBI) can help address both of these challenges with respect to parameter estimation. SBI uses a learned mapping between the parameter space and observed data to estimate parameters for the generation of calibrated simulations. To demonstrate the potential of SBI in hydrologic modeling, we conduct a set of synthetic experiments to infer two common physical parameters – Manning's coefficient and hydraulic conductivity – using a representation of a snowmelt-dominated catchment in Colorado, USA. We introduce novel deep-learning (DL) components to the SBI approach, including an “emulator” as a surrogate for the PB simulator to rapidly explore parameter responses. We also employ a density-based neural network to represent the joint probability of parameters and data without strong assumptions about its functional form. While addressing intractability, we also show that, if the simulator does not represent the system under study well enough, SBI can yield unreliable parameter estimates. Approaches to adopting the SBI framework for cases in which multiple simulator(s) may be adequate are introduced using a performance-weighting approach. The synthetic experiments presented here test the performance of SBI, using the relationship between the surrogate and PB simulators as a proxy for the real case.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2025

DESI 2024 II: sample definitions, characteristics, and two-point clustering statistics

We present the samples of galaxies and quasars used for DESI 2024 cosmological analyses, drawn from the DESI Data Release 1 (DR1). We describe the construction of largescale structure (LSS) catalogs from these samples, which include matched sets of synthetic reference ‘randoms’ and weights that account for variations in the observed density of the samples due to experimental design and varying instrument performance. We detail how we correct for variations in observational completeness, the input ‘target’ densities due to imaging systematics, and the ability to confidently measure redshifts from DESI spectra. We then summarize how remaining uncertainties in the corrections can be translated to systematic uncertainties for particular analyses. We describe the weights added to maximize the signalto-noise of DESI DR1 2-point clustering measurements. We detail measurement pipelines applied to the LSS catalogs that obtain 2-point clustering measurements in configuration and Fourier space. The resulting 2-point measurements depend on window functions and normalization constraints particular to each sample, and we present the corrections required to match models to the data. We compare the configuration- and Fourier-space 2-point clustering of the data samples to that recovered from simulations of DESI DR1 and find they are, generally, in statistical agreement to within 2% in the inferred real-space over-density field. The LSS catalogs, 2-point measurements, and their covariance matrices will be released publicly with DESI DR1.

79 ASTRONOMY AND ASTROPHYSICS↗

NASA NTRS1987

Parametric analysis of synthetic aperture radar data for characterization of deciduous forest stands

The SAR sensor parameters that affect the estimation of deciduous forest stand characteristics were examined using data sets for the Gulf Coastal Plain region, acquired by the NASA/JPL multipolarization airborne SAR. In the regression analysis, the mean digital-number values of the three polarization data are used as the independent variables to estimate the average tree height (HT), basal area (BA), and total-tree biomass (TBM). The following results were obtained: (1) in the case of simple regression and using 28 plots, vertical-vertical (VV) polarization yielded the largest correlation coefficients (r) in estimating HT, BA, and TBM; (2) in the case of multiple regression, the horizontal-horizontal (HH) and VV polarization combination yielded the largest r value in estimating HT, while the VH and HH polarization combination yielded the largest r values in estimating BA and TBM. With the addition of a third polarization, the increase in r values is insignificant.

Wu, Shih-Tseng↗

DOE OSTI2026

Challenges and Vision for Standardization of Biopolymer Data Sets for Machine Learning

Machine learning (ML) is transforming materials research, yet potential for biopolymer discovery remains constrained by fragmented data and nonstandardized reporting. Biopolymers differ significantly from synthetic polymers, requiring specialized approaches to represent their biosynthetic origins, hierarchical structures, and application-specific metrics. In this Perspective, we identify three core challenges limiting biopolymer representation: information encoding, data quality, and data sharing. We describe the most pressing issues and propose commensurate approaches to address each key challenge. Recommendations include the design and adoption of biopolymer-specific fingerprinting and representation frameworks, development of hybrid human-large language model (LLM) data extraction strategies, and expanding Findable, Accessible, Interoperable, Reusable (FAIR)-compliant repositories. We propose a robust foundation to define interoperable, high-quality data sets that capture the full context of biopolymer materials. Standardized metadata, shared ontologies, and community-driven infrastructure would enable scalable, reproducible workflows and accelerate the ML-driven development of biopolymers.

36 MATERIALS SCIENCE↗

NASA NTRS1991

Land cover stratification using Landsat Thematic Mapper data in Sahelian and Sudanian woodland and wooded grassland

A standard methodology for thematic mapping of natural vegetation using remotely sensed imagery and digital image processing was modified to account for the spatial and spectral properties of semi-arid landscapes, and tested in study areas in the Sahelian and Sudanian zones, Mali. A principal components transformation of registered wet and dry season Landsat TM images produced a set of synthetic spectral channels differentiating vegetation cover between seasons, and allowed areas with annual grass growth to be distinguished from areas with woody cover. The transformed data were statistically clustered and clusters were assigned to vegetation type and density categories. In a separate step, the images were manually interpreted to differentiate broad soil classes. Four statistics were compared to evaluate the accuracy of the maps based on sample points from air photos. For the relatively detailed categories initially defined, map accuracies were substandard; however, when vegetation density classes were aggregated, overall accuracy was around 90 percent, and class accuracy was greater than 80 percent for most classes. This method is suitable for stratification and inventory of woody biomass at a regional scale in semi-arid woodland and wooded grassland.

Franklin, J.↗

DOE OSTI2025

Applying Transfer Learning for Street-Scale Nuisance Flood Forecasting in Coastal-Urban Cities

An important challenge with Machine Learning (ML) is its transferability; that is, whether a ML model trained on one set of data can be applied to a second set of data without requiring a full re-training of the model. Transfer Learning (TL) addresses this challenge by transferring knowledge learned in the source domain (the data it was trained on) to the target domain (a second set of data that is statistically different but related, which the model was not trained on). This study investigates the use of TL for street-scale nuisance flood forecasting by exploring whether a ML model trained on data collected for one set of streets can effectively forecast flooding for another set of streets in the same city using TL. The envisioned use case is a city deploying a new flood depth monitoring sensor on a street and using TL to apply a ML model, trained on sensor data from an existing flood depth sensor network, to this new street. Eventually, the new flood depth sensor will have a sufficient dataset for training its own ML model, but TL can be used to fill the gap in time while this new dataset is being generated. This method is explored using a Long Short-Term Memory (LSTM) model trained on data for the flood-prone streets of Norfolk City, Virginia. The data used for training includes environmental time series (rainfall, tide), topographic features (Digital Elevation Model (DEM), Topographic Wetness Index (TWI), Depth To Water (DTW)), and street-scale flood depth time series obtained from a high-fidelity physics-based model, acting as a synthetic street-scale stream depth sensor dataset since actual stream depth sensor data is generally unavailable for most cities. A set of 180 flood-prone streets was used to train a base model, while another set of 180 flood-prone streets was used to re-train that model using different TL strategies. The results show that full-weight re-training proved most effective and minimal re-training of only the output layer was insufficient. The advantage of TL was most pronounced when target data was limited, meaning data collected at the new water depth sensor location included generally less than 18 flood events. As target data increased beyond 18 flood events, the benefit of TL diminished relative to training a ML model directly on the local flood events. These findings can assist cities as they implement street-scale flood sensing systems to create accurate forecasts for new sensing locations that do not yet have sufficient data records to train a local ML model.

Roy, Binata [Univ. of Virginia, Charlottesville, V↗