Search NASA⌕ Search

SEARCH · Search NASA

Results for “distributed state estimation”

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 919 records · Page 51

Improved Bayesian Infrasonic Source Localization for regional infrasound

The Bayesian Infrasonic Source Localization (BISL) methodology is examined and simplified providing a generalized method of estimating the source location and time for an infrasonic event and the mathematical framework is used therein. The likelihood function describing an infrasonic detection used in BISL has been redefined to include the von Mises distribution developed in directional statistics and propagation-based, physically derived celerity-range and azimuth deviation models. Frameworks for constructing propagation-based celerity-range and azimuth deviation statistics are presented to demonstrate how stochastic propagation modelling methods can be used to improve the precision and accuracy of the posterior probability density function describing the source localization. Infrasonic signals recorded at a number of arrays in the western United States produced by rocket motor detonations at the Utah Test and Training Range are used to demonstrate the application of the new mathematical framework and to quantify the improvement obtained by using the stochastic propagation modelling methods. Moreover, using propagation-based priors, the spatial and temporal confidence bounds of the source decreased by more than 40 per cent in all cases and by as much as 80 per cent in one case. Further, the accuracy of the estimates remained high, keeping the ground truth within the 99 per cent confidence bounds for all cases.

58 GEOSCIENCES↗

Forest Carbon Storage in the Western United States: Distribution, Drivers, and Trends

Abstract Forests are a large carbon sink and could serve as natural climate solutions that help moderate future warming. Thus, establishing forest carbon baselines is essential for tracking climate‐mitigation targets. Western US forests are natural climate solution hotspots but are profoundly threatened by drought and altered disturbance regimes. How these factors shape spatial patterns of carbon storage and carbon change over time is poorly resolved. Here, we estimate live and dead forest carbon density in 19 forested western US ecoregions with national inventory data (2005–2019) to determine: (a) current carbon distributions, (b) underpinning drivers, and (c) recent trends. Potential drivers of current carbon included harvest, wildfire, insect and disease, topography, and climate. Using random forests, we evaluated driver importance and relationships with current live and dead carbon within ecoregions. We assessed trends using linear models. Pacific Northwest (PNW) and Southwest (SW) ecoregions were most and least carbon dense, respectively. Climate was an important carbon driver in the SW and Lower Rockies. Fire reduced live and increased dead carbon, and was most important in the Upper Rockies and California. No ecoregion was unaffected by fire. Harvest and private ownership reduced carbon, particularly in the PNW. Since 2005, live carbon declined across much of the western US, likely from drought and fire. Carbon has increased in PNW ecoregions, likely recovering from past harvest, but recent record fire years may alter trajectories. Our results provide insight into western US forest carbon function and future vulnerabilities, which is vital for effective climate change mitigation strategies.

Environmental Sciences & Ecology↗

Modelling Groundwater Depletion at Regional and Global Scales: Present State and Future Prospects.

Except for frozen water in ice and glaciers, groundwater is the world's largest distributed store of freshwater and has strategic importance to global food and water security. In this paper, the most recent advances quantifying groundwater depletion (GWD) are comprehensively reviewed. This paper critically evaluates the recently advanced modeling approaches estimating GWD at regional and global scales, and the evidence of feedbacks to the Earth system including sea-level rise associated with GWD. Finally, critical challenges and opportunities in the use of groundwater are identified for the adaption to growing food demand and uncertain climate.

Groundwater Depletion (GWD)↗

The state of meteoritic material on the moon

Velocity distributions obtained from earth-based meteor observations and corrected to constant meteoroid mass are transformed to meteoroid impact velocity distributions at the moon. Average velocities at the moon thus obtained range from 13 to 18 km/sec. This information is then combined with impact heating estimates to determine the amounts of resident lunar material and incoming meteoritic material that is melted or vaporized. It is found that about 5% of the resulting melt and about 25% of the impact generated vapor should be of meteoritic content. The uncertainties in these estimates are rather large, however, because of uncertainties in the various experimental results required to establish the parameters used in this analysis. The impact generated melts and impact generated vapor deposits, if the latter can be found, should provide additional means for chemically separating out the meteoritic component in the lunar soil.

Zook, H. A.↗

Karhunen–Loève deep learning method for surrogate modeling and approximate Bayesian parameter estimation

We evaluate the performance of the Karhunen-Loève Deep Neural Network (KL-DNN) framework for surrogate modeling and approximate Bayesian parameter estimation in partial differential equation models. In the surrogate model, the Karhunen-Loève (KL) expansions are used for the dimensionality reduction of the number of unknown parameters and variables, and a deep neural network is employed to relate the reduced space of parameters to that of the state variables. The KL-DNN surrogate model is used to formulate a maximum-a-posteriori-like least-squares problem, which is randomized to draw samples of the posterior distribution of the parameters. We test the proposed framework for a hypothetical unconfined aquifer via comparison with the forward MODFLOW and inverse PEST++ iterative ensemble smoother (IES) solutions as well as the state-of-the-art Fourier neural operator (FNO) and deep operator networks (DeepONets) operator learning surrogate models. Our results show that the KL-DNN surrogate model outperforms FNO and DeepONet for forward predictions. For solving inverse problems, the randomized algorithm provides the same or more accurate Bayesian predictions of the parameters than IES as evidenced by the higher log-predictive probability of both the estimated parameter field and the forecast hydraulic head. The posterior mean obtained from the randomized algorithm is closer to the reference parameter field than that obtained with FNO as the maximum a posteriori estimate.

Approximate Bayesian inference↗

Distributions of methyl group rotational barriers in polycrystalline organic solids

We bring together solid state {sup 1}H spin-lattice relaxation rate measurements, scanning electron microscopy, single crystal X-ray diffraction, and electronic structure calculations for two methyl substituted organic compounds to investigate methyl group (CH{sub 3}) rotational dynamics in the solid state. Methyl group rotational barrier heights are computed using electronic structure calculations, both in isolated molecules and in molecular clusters mimicking a perfect single crystal environment. The calculations are performed on suitable clusters built from the X-ray diffraction studies. These calculations allow for an estimate of the intramolecular and the intermolecular contributions to the barrier heights. The {sup 1}H relaxation measurements, on the other hand, are performed with polycrystalline samples which have been investigated with scanning electron microscopy. The {sup 1}H relaxation measurements are best fitted with a distribution of activation energies for methyl group rotation and we propose, based on the scanning electron microscopy images, that this distribution arises from molecules near crystallite surfaces or near other crystal imperfections (vacancies, dislocations, etc.). An activation energy characterizing this distribution is compared with a barrier height determined from the electronic structure calculations and a consistent model for methyl group rotation is developed. The compounds are 1,6-dimethylphenanthrene and 1,8-dimethylphenanthrene and the methyl group barriers being discussed and compared are in the 2–12 kJ mol{sup −1} range.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

An ensemble Kalman filter for statistical estimation of physics constrained nonlinear regression models

A central issue in contemporary science is the development of nonlinear data driven statistical–dynamical models for time series of noisy partial observations from nature or a complex model. It has been established recently that ad-hoc quadratic multi-level regression models can have finite-time blow-up of statistical solutions and/or pathological behavior of their invariant measure. Recently, a new class of physics constrained nonlinear regression models were developed to ameliorate this pathological behavior. Here a new finite ensemble Kalman filtering algorithm is developed for estimating the state, the linear and nonlinear model coefficients, the model and the observation noise covariances from available partial noisy observations of the state. Several stringent tests and applications of the method are developed here. In the most complex application, the perfect model has 57 degrees of freedom involving a zonal (east–west) jet, two topographic Rossby waves, and 54 nonlinearly interacting Rossby waves; the perfect model has significant non-Gaussian statistics in the zonal jet with blocked and unblocked regimes and a non-Gaussian skewed distribution due to interaction with the other 56 modes. We only observe the zonal jet contaminated by noise and apply the ensemble filter algorithm for estimation. Numerically, we find that a three dimensional nonlinear stochastic model with one level of memory mimics the statistical effect of the other 56 modes on the zonal jet in an accurate fashion, including the skew non-Gaussian distribution and autocorrelation decay. On the other hand, a similar stochastic model with zero memory levels fails to capture the crucial non-Gaussian behavior of the zonal jet from the perfect 57-mode model.

97 MATHEMATICS AND COMPUTING↗

6D Simulations of PIP-II Booster Injection

The PIP-II superconducting linac will deliver 2 mA average H⁻ beam current at 800 MeV to the existing Booster synchrotron over a period of 0.55 ms (285 turns). As a result, the injected beam power will quadruple to 17 kW. Safe operation at the increased beam power implies careful attention to the origin, magnitude, and distribution of both controlled and uncontrolled losses. Uncontrolled losses are due to neutral ions in excited states stripped in downstream magnets and large angle scattered protons from parasitic foil hits. The relative magnitudes of these loss mechanisms is used to determine the optimal foil thickness. A transverse painting scheme involving closed orbit motion will be used to mitigate space charge effects and minimize parasitic foil hits. Using a detailed full 6D simulation of the injection process, we compute large angle scattering losses and compare results to back of the envelope estimates. We investigate possible impact of space charge on the emittance and beam distribution both during and at the conclusion of the injection period.

43 PARTICLE ACCELERATORS↗

2012 Market Report on U.S. Wind Technologies in Distributed Applications

At the end of 2012, U.S. wind turbines in distributed applications reached a 10-year cumulative installed capacity of more than 812 MW from more than 69,000 units across all 50 states. In 2012 alone, nearly 3,800 wind turbines totaling 175 MW of distributed wind capacity were documented in 40 states and in the U.S. Virgin Islands, with 138 MW using utility-scale turbines (i.e., greater than 1 MW in size), 19 MW using mid-size turbines (i.e., 101 kW to 1 MW in size), and 18.4 MW using small turbines (i.e., up to 100 kW in size). Distributed wind is defined in terms of technology application based on a wind project’s location relative to end-use and power-distribution infrastructure, rather than on technology size or project size. Distributed wind systems are either connected on the customer side of the meter (to meet the onsite load) or directly to distribution or micro grids (to support grid operations or offset large loads nearby). Estimated capacity-weighted average costs for 2012 U.S. distributed wind installations was $\$$2,540/kW for utility-scale wind turbines, $\$$2,810/kW for mid-sized wind turbines, and $\$$6,960/kW for newly manufactured (domestic and imported) small wind turbines. An emerging trend observed in 2012 was an increased use of refurbished turbines. The estimated capacity-weighted average cost of refurbished small wind turbines installed in 2012 was $\$$4,080/kW. As a result of multiple projects using utility-scale turbines, Iowa deployed the most new overall distributed wind capacity, 37 MW, in 2012. Nevada deployed the most small wind capacity in 2012, with nearly 8 MW of small wind turbines installed in distributed applications. In the case of mid-size turbines, Ohio led all states in 2012 with 4.9 MW installed in distributed applications. State and federal policies and incentives continued to play a substantial role in the development of distributed wind projects. In 2012, U.S. Treasury Section 1603 payments and grants and loans from the U.S. Department of Agriculture’s Rural Energy for America Program were the main sources of federal funding for distributed wind projects. State and local funding varied across the country, from rebates to loans, tax credits, and other incentives. Reducing utility bills and hedging against potentially rising electricity rates remain drivers of distributed wind installations. In 2012, other drivers included taking advantage of the expiring U.S. Treasury Section 1603 program and a prosperous year for farmers. While 2012 saw a large addition of distributed wind capacity, considerable barriers and challenges remain, such as a weak domestic economy, inconsistent state incentives, and very competitive solar photovoltaic and natural gas prices. The industry remains committed to improving the distributed wind marketplace by advancing the third-party certification process and introducing alternative financing models, such as third-party power purchase agreements and lease-to-own agreements more typical in the solar photovoltaic market. Continued growth is expected in 2013.

distributed wind↗

Generalized Dalitz plot analysis of the near-threshold pp{yields}ppK{sup +}K{sup -} reaction in view of the K{sup +}K{sup -} final state interaction

The excitation function for the pp{yields}ppK{sup +}K{sup -} reaction revealed a significant enhancement close to threshold that may plausibly be assigned to the influence of the pK{sup -} and K{sup +}K{sup -} final-state interactions. In an improved reanalysis of COSY-11 data for the pp{yields}ppK{sup +}K{sup -} reaction at excess energies of Q=10 and 28 MeV, including the proton-K{sup -} interaction, the enhancement is confirmed. Invariant mass distributions for the two- and three-particle subsystems allow us to test at low excess energies the ansatz and parameters for the description of the interaction in the ppK{sup +}K{sup -} system as derived from the COSY-ANKE data. Finally, based for the first time on the low-energy K{sup +}K{sup -} invariant mass distributions and the generalized Dalitz plot analysis, we estimate the scattering length for the K{sup +}K{sup -} interaction to be |Re(a{sub K{sup +}}{sub K{sup -}})|=0.5{sub -0.5}{sup +4.0} fm and Im(a{sub K{sup +}}{sub K{sup -}})=3.0{+-}3.0 fm.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Estimating residential energy consumption in metropolitan areas: A microsimulation approach

Prior research has shown that land use patterns and the spatial configurations of cities have a significant impact on residential energy demand. Given the pressing issues surrounding energy security and climate change, there is renewed interest in developing and retrofitting cities to make them more energy efficient. Yet deriving micro-scale residential energy footprints of metropolitan areas is challenging because high resolution data from energy providers is generally unavailable. In this study, a bottom-up model is proposed to estimate residential energy demand using datasets that are commonly available in the United States. The model applies novel machine learning methods to match records in the Residential Energy Consumption Survey with Public Use Microdata samples. This matching and machine learning produce a synthetic household energy distribution at a neighborhood scale. The model was tested and validated with data from the Atlanta metropolitan region to demonstrate its application and promise.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Boil-off losses along LH2 pathway

Losses along the LH 2 pathway are intrinsic to the utilization of a cryogenic fluid. They occur when the molecule is transferred between 2 vessels (liquefaction plant to trailer, trailer to station storage, station storage to pump or compressor, then fuel cell electric vehicles …) and when the fluid is warmed up due to heat transfer with the environment. Those losses can be estimated with good accuracy using thermodynamic models based on conservation of mass and energy, providing the thermodynamic states are correctly described. Indeed, the fluid undergoes various changes as it moves along the entire pathway (2 phase transition, super-heated warming, non-uniform temperature distributions across the saturation film) and accurate equations of state and 2 phase behavior implementations are essential. The balances of mass and energy during the various dynamics processes then enable to quantify the boil-off losses. In this work, a MATLAB code previously developed by NASA to simulate rocket loading is used as the basis for the LH 2 transfer model. This code implements complex physical phenomena such as the competition between condensation and evaporation and the convection vs. conduction heat transfer as a function of the relative temperatures on both sides of the saturated film. The original code was modified to consider real gas equations of state, and some semi-empirical relationships, such as between the heat of vaporization and the critical temperature, were also replaced by a REFPROP equivalent expression, assumed to be more accurate. Non-constant liquid temperature equations were added to simulate sub-cooled conditions. It is shown that although trailer depressurization and transfer losses from the receiving vessel during a LH 2 delivery may constitute the largest sources of losses, those can be greatly reduced and even eliminated if CFR/DOT regulations are properly applied and if no-vent fill methods are implemented, particularly for a stationary LH2 storage operating at reasonably low pressure (45 psia or so). As such, it is expected that the only remaining boil-off losses for a refueling station would come from the LH 2 pump (utilization and idling), pump vessel cool down/warm up, and environment heat transfer. Based on experimental data measured at LLNL on the Linde 875 bar LH 2 cryo-pump and results from the model, and extrapolating for refueling stations of various sizes, it can be shown that boil-off losses can vary from 15% of delivered LH 2 for a 100 kg/day, 5% at 400 kg/day, and down to less than 2% for stations above 1,800 kg/day (for a system similar to the one at LLNL). Less boil-off is to be expected for LH 2 pumps dispensing at 350 bar, so that less than 0.7% can be expected above 1,800 kg/day station capacities. At last, boil-off recovery solutions are briefly analyzed, including compressors, cryo-coolers and stationary fuel cell. Their economic value is strongly related to the costs of H2 and of the electricity needed to perform the recovery, on top of the initial capital expenditures. Effective boil-off could be reduced to 0.6-1% for a 2,000 kg/day station if adequate recovery solutions are used.

08 HYDROGEN↗

Spatial Heterogeneity in Cloud Condensation Nuclei (SHINE) Field Campaign Report

Multiple land-atmosphere coupling processes influence the spatial distribution of boundary-layer energy, moisture, aerosol, or cloud condensation nuclei (CCN) properties. Accurately quantifying the spatial and temporal distributions of CCN concentrations is essential for investigating aerosol-cloud interactions and reducing uncertainties in aerosol radiative forcing estimates of Earth's system.

54 ENVIRONMENTAL SCIENCES↗

Using probabilistic analysis to assess the reliability of predicted SRB aft-skirt stresses

Probabilistic failure analysis is a tool to predict the reliability of a part or system. Probabalistic techniques were used to predict critical stresses which occur in the solid rocket booster aft-skirt during main engine buildup, immediately prior to lift-off. More than any other hold down post (HDP) load component, the Z loads are sensitive to variations in strains and calibration constants. Also, predicted aft-skirt stresses are strongly affected by HDP load variations. Therefore, the instrumented HDP are not effective load transducers for Z loads, and, when used with aft skirt stress indicator equations, yield estimates with large uncertainty. Monte Carlo simulation proved to be a straight forward way of studying the overlapping effects of multiple parameters on predicted equipment performance. An advantage of probabilistic analysis is the degree of uncertainty of each parameter as stated explicitly by its probability distribution. It was noted, however, that the choice of parameter distribution had a large effect on the simulation results. Many times these distributions must be assumed. The engineer who is designing the part should be responsible for the choice of parameter distribution.

Richardson, James A.↗

Improving the SMAP Level-4 Soil Moisture Product

The NASA Soil Moisture Active Passive (SMAP) mission generates, among other data sets, the Level 4 Soil Moisture (L4 SM) product. The L4 SM algorithm is based on the assimilation of SMAP radiometer brightness temperature observations into the NASA Catchment land surface model using a spatially distributed ensemble Kalman filter. The L4 SM data are published with a mean latency of approx. 2.5 days from the time of observation and provide global, three-hourly, 9 km resolution estimates of surface and root-zone soil moisture and related land surface states and fluxes. In 2018, the product was upgraded from Version 3 to Version 4. Underlying the new version is a revised modeling system that includes improved input parameter datasets for land cover, topography, and vegetation height that are based on recent, high quality space-borne remote sensing observations. Land cover inputs were updated to the GlobCover2009 product, which is based on satellite observations from the Medium Resolution Imaging Spectrometer. Topographic statistics now rely on observations from the Shuttle Radar Topography Mission. Finally, vegetation height inputs are derived from space-borne Lidar measurements. Additionally, SMAP Level-2 soil moisture retrievals and in situ soil moisture measurements were used to calibrate a particular Catchment model parameter that governs the recharge of soil moisture from the models root-zone excess reservoir into the surface excess reservoir. Specifically, the replenishment of soil moisture near the surface from below under non-equilibrium conditions was substantially reduced, which brings the models surface soil moisture more in line with the SMAP Level 2 and in situ soil moisture. Finally, the calibration of the assimilated SMAP brightness temperatures changed substantially from Version 3 to Version 4. Considerable effort went into the version upgrade, creating an expectation that the new version is improved over the old version. Indeed, some aspects of the new version are clearly better. However, other aspects are not, and on balance, the overall improvement is modest at best. In this presentation we summarize the skill of the new and old versions vs. independent in situ measurements and in terms of data assimilation diagnostics, including, for example, the statistics of the (soil moisture) analysis increments and the observation minus forecast (brightness temperatures) residuals. We share our experience with trying to improve to the L4 SM product and the lessons learned from the effort.

Reichle, Rolf↗

Quaternion Averaging

Many applications require an algorithm that averages quaternions in an optimal manner. For example, when combining the quaternion outputs of multiple star trackers having this output capability, it is desirable to properly average the quaternions without recomputing the attitude from the the raw star tracker data. Other applications requiring some sort of optimal quaternion averaging include particle filtering and multiple-model adaptive estimation, where weighted quaternions are used to determine the quaternion estimate. For spacecraft attitude estimation applications, derives an optimal averaging scheme to compute the average of a set of weighted attitude matrices using the singular value decomposition method. Focusing on a 4-dimensional quaternion Gaussian distribution on the unit hypersphere, provides an approach to computing the average quaternion by minimizing a quaternion cost function that is equivalent to the attitude matrix cost function Motivated by and extending its results, this Note derives an algorithm that deterniines an optimal average quaternion from a set of scalar- or matrix-weighted quaternions. Rirthermore, a sufficient condition for the uniqueness of the average quaternion, and the equivalence of the mininiization problem, stated herein, to maximum likelihood estimation, are shown.

Markley, F. Landis↗

Adversarial autoencoder ensemble for fast and probabilistic reconstructions of few-shot photon correlation functions for solid-state quantum emitters

Second-order photon correlation measurements [g (2) (τ) functions] are widely used to classify single-photon emission purity in quantum emitters or to measure the multiexciton quantum yield of emitters that can simultaneously host multiple excitations – such as quantum dots – by evaluating the value of g (2) (τ = 0). Accumulating enough photons to accurately calculate this value is time consuming and could be accelerated by fitting of few-shot photon correlations. Here, we develop an uncertainty-aware, deep adversarial autoencoder ensemble (AAE) that reconstructs noise-free g (2) (τ) functions from noise-dominated, few-shot inputs. The model is trained with simulated g (2) (τ) functions that are facilely generated by Poisson sampling time bins. The AAE reconstructions are performed orders-of-magnitude faster, with reconstruction errors and estimates of g (2) (τ = 0) that are lower in variance and similar in accuracy compared to Maximum likelihood estimation and Levenberg-Marquardt least-squares fitting approaches, for simulated and experimentally measured few-shot g (2) (τ) functions (~100 two-photon events) of InP/ZnS/ZnSe and CdS/CdSe/CdS quantum dots. The deep-ensemble model comprises eight individual autoencoders, allowing for probabilistic reconstructions of noise-free g (2) (τ) functions, and we show that the predicted variance scales inversely with number of shots, with comparable uncertainties to computationally intensive Markov chain Monte Carlo sampling. Furthermore, this work demonstrates the advantage of machine learning models to perform uncertainty-aware, fast, and accurate reconstructions of simple Poisson-distributed photon correlation functions, allowing for on-the-fly reconstructions and accelerated materials characterization of solid-state quantum emitters.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗