SEARCH · Engineering Papers

Results for “Data Reasoning”

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 505 records · Page 28

All sources NASA NTRS DOE OSTI

DOE OSTI2023

Modular machine learning-based elastoplasticity: Generalization in the context of limited data

The development of highly accurate constitutive models for materials that undergo path-dependent processes continues to be a complex challenge in computational solid mechanics. Challenges arise both in considering the appropriate model assumptions and from the viewpoint of data availability, verification, and validation. Recently, data-driven modeling approaches have been proposed that aim to establish stress-evolution laws that avoid user-chosen functional forms by relying on machine learning representations and algorithms. However, these approaches not only require a significant amount of data but also need data that probes the full stress space with a variety of complex loading paths. Furthermore, they rarely enforce all necessary thermodynamic principles as hard constraints. Hence, they are in particular not suitable for low-data or limited-data regimes, where the first arises from the cost of obtaining the data and the latter from the experimental limitations of obtaining labeled data, which is commonly the case in engineering applications. In this work, we discuss a hybrid framework that can work on a variable amount of data by relying on the modularity of the elastoplasticity formulation where each component of the model can be chosen to be either a classical phenomenological or a data-driven model depending on the amount of available information and the complexity of the response. The method is tested on synthetic uniaxial data coming from simulations as well as cyclic experimental data for structural materials. The discovered material models are found to not only interpolate well but also allow for accurate extrapolation in a thermodynamically consistent manner far outside the domain of the training data. This ability to extrapolate from limited data was the main reason for the early and continued success of phenomenological models and the main shortcoming in machine learning-enabled constitutive modeling approaches. Training aspects and details of the implementation of these models into Finite Element simulations are discussed and analyzed.

42 ENGINEERING↗

NASA NTRS2010

CFD Models of a Serpentine Inlet, Fan, and Nozzle

Several computational fluid dynamics (CFD) codes were used to analyze the Versatile Integrated Inlet Propulsion Aerodynamics Rig (VIIPAR) located at NASA Glenn Research Center. The rig consists of a serpentine inlet, a rake assembly, inlet guide vanes, a 12-in. diameter tip-turbine driven fan stage, exit rakes or probes, and an exhaust nozzle with a translating centerbody. The analyses were done to develop computational capabilities for modeling inlet/fan interaction and to help interpret experimental data. Three-dimensional Reynolds averaged Navier-Stokes (RANS) calculations of the fan stage were used to predict the operating line of the stage, the effects of leakage from the turbine stream, and the effects of inlet guide vane (IGV) setting angle. Coupled axisymmetric calculations of a bellmouth, fan, and nozzle were used to develop techniques for coupling codes together and to investigate possible effects of the nozzle on the fan. RANS calculations of the serpentine inlet were coupled to Euler calculations of the fan to investigate the complete inlet/fan system. Computed wall static pressures along the inlet centerline agreed reasonably well with experimental data but computed total pressures at the aerodynamic interface plane (AIP) showed significant differences from the data. Inlet distortion was shown to reduce the fan corrected flow and pressure ratio, and was not completely eliminated by passage through the fan

Chima, R. V.↗

NASA NTRS2001

Auroral, Polar Cap, and Polar Cusp Modeling and Data Analysis for the IMAGE Mission and LENA Instrument

One of the chief mysteries in the LENA perigee pass data is the lack of an apparent auroral oval in the images. Another is that in some cases ENA are seen from any direction near the Earth regardless of the latitude of the spacecraft. These facts lead one to ask a fundamental question: Is the instrument responding to ENA primarily? One possible way to get out of the "ambiguity" of the data is to assume that at least part of the signal is produced by something other than ENA. The two main candidates for this "something else" are UV light and energetic charged particles. UV light could only effect the instrument when its fan shaped aperture points toward the source. The most intense of which will be the sun, with day glow being the second strongest and the auroral zone, the third. We can rule out UV light as a prime source of counts in the perigee pass data for the following reasons: 1) The perigee pass signal is different in form and much stronger than the sun pulse signal seen just before or just after perigee; 2) There is no indication of the auroral zone, which would produce at least two peaks in the counts versus spin phase curve; 3) Mike Colliers' analysis of the sun pulse signal shows that it varies with the flux of the solar wind and not with variations in the solar UV flux. Charge particles that enter the aperture of the instrument and produce counts would show up when the instrument looks in the direction from which they come. In all of the data I am analyzing voltages were being applied to the collimators so that most charged particles should have been excluded from the instrument but this effect could still show up where the flux of energetic particles is high enough. The most likely place would be in the auroral zone where energetic electrons and protons precipitate. If these particles are producing counts then they should be seen when the instrument looks in and near the zenith direction. In nearly all of the perigee passes the zenith direction is devoid of counts.

Wilson, Gordon R.↗

NASA NTRS1982

A simplified design procedure for life prediction of rocket thrust chambers

An analytical procedure for predicting thrust chamber life is developed. The hot-gas-wall ligaments separating the coolant and combustion gas are subjected to pressure loading and severe thermal cycling. The resulting stresses interact during plastic straining causing incremental bulging of the ligaments during each firing cycle. This mechanism of plastic ratcheting is analyzed and a method using a yield surface for combined bending and membrane loading developed for determining the incremental permanent deflection and progressive thinning near the center of the ligaments. Fatigue and tensile instability are analyzed as possible failure modes. Results of the simplified analyses compare favorably with available experimental data and finite element analysis results for OFHC (Oxygen Free High Conductivity) copper. They are also in reasonably good agreement with experimental data for NARloy Z, a copper-zirconium-silver alloy developed by the Rocketdyne Division of Rockwell International.

Porowski, J. S.↗

NASA NTRS2003

Computational Study of Axisymmetric Off-Design Nozzle Flows

Computational Fluid Dynamics (CFD) analyses of axisymmetric circular-arc boattail nozzles operating off-design at transonic Mach numbers have been completed. These computations span the very difficult transonic flight regime with shock-induced separations and strong adverse pressure gradients. External afterbody and internal nozzle pressure distributions computed with the Wind code are compared with experimental data. A range of turbulence models were examined, including the Explicit Algebraic Stress model. Computations have been completed at freestream Mach numbers of 0.9 and 1.2, and nozzle pressure ratios (NPR) of 4 and 6. Calculations completed with variable time-stepping (steady-state) did not converge to a true steady-state solution. Calculations obtained using constant timestepping (timeaccurate) indicate less variations in flow properties compared with steady-state solutions. This failure to converge to a steady-state solution was the result of using variable time-stepping with large-scale separations present in the flow. Nevertheless, time-averaged boattail surface pressure coefficient and internal nozzle pressures show reasonable agreement with experimental data. The SST turbulence model demonstrates the best overall agreement with experimental data.

DalBello, Teryn↗

DOE OSTI2022

DIAS: A Data-Informed Active Subspace Regularization Framework for Inverse Problems

This paper presents a regularization framework that aims to improve the fidelity of Tikhonov inverse solutions. At the heart of the framework is the data-informed regularization idea that only data-uninformed parameters need to be regularized, while the data-informed parameters, on which data and forward model are integrated, should remain untouched. We propose to employ the active subspace method to determine the data-informativeness of a parameter. The resulting framework is thus called a data-informed (DI) active subspace (DIAS) regularization. Four proposed DIAS variants are rigorously analyzed, shown to be robust with the regularization parameter and capable of avoiding polluting solution features informed by the data. They are thus well suited for problems with small or reasonably small noise corruptions in the data. Furthermore, the DIAS approaches can effectively reuse any Tikhonov regularization codes/libraries. Though they are readily applicable for nonlinear inverse problems, we focus on linear problems in this paper in order to gain insights into the framework. Various numerical results for linear inverse problems are presented to verify theoretical findings and to demonstrate advantages of the DIAS framework over the Tikhonov, truncated SVD, and the TSVD-based DI approaches.

Nguyen, Hai↗

NASA NTRS2004

Fabrication of a novel gigabit/second free-space optical interconnect - photodetector characterization and testing and system development

The time when computing power is limited by the copper wire inherent in the computer system and not the speed of the microprocessor is rapidly approaching. With constant advances in computer technology, many researchers believe that in only a few years, optical interconnects will begin to replace copper wires in your Central Processing Unit (CPU). On a more macroscopic scale, the telecommunications industry has already made the switch to optical data transmission as, to date, fiber optic technology is the only reasonable method of reliable, long range data transmission. Within the span of a decade, we will see optical technologies move from the macroscopic world of the telecommunications industry to the microscopic world of the computer chip. Already, the communications industry is marketing commercially available optical links to connect two personal computers, thereby eliminating the need for standard and comparatively slow wired and wireless Ethernet transfers and greatly increasing the distance the computers can be separated. As processing demands continue to increase, the realm of optical communications will continue to move closer to the microprocessor and quite possibly onto the microprocessor itself. A day may come when copper connections are used only to supply power, not transfer data. This summer s work marks some of the beginning stages of a 5 to 10 year, long-term research project to create and study a free-space, 1 Gigabit/sec optical interconnect. The research will result in a novel fabricated, chip-to-chip interconnect consisting of a Vertical Cavity Surface Emitting Laser (VCSEL) Diode linked through free space to a Metal- Semiconductor-Metal (MSM) Photodetector with the possible integration of microlenses for signal focusing and Micro-Electromechanical Systems (MEMS) devices for optical signal steering. The advantages, disadvantages, and practicality of incorporating flip-chip mounting technologies will also be addressed. My work began with the design and construction of a test setup for the experiment and then appropriate characterization of the test system. Specifically, I am involved in the characterization of a commercially available 1550nm wavelength, 5mW diode laser and a study of its modulation bandwidth. Commercially produced photodetectors as well as the incorporation of microwave technology, in the form of RF input and output, are used in the characterization procedure. The next stage involves the use of a probe station and network analyzer to characterize and test a series of photodetectors fabricated on a 2 inch, Indium Gallium Arsenide (InGaAs) wafer in the Branch s microlithography lab. Other project responsibilities include, but are not limited to the incorporation of a transimpedance amplifier to the photodetector circuit; a study of VCSEL technology; bit error rate analysis of an optical interconnect system; and analysis of free space divergence of the VCSEL, optical path length of the interconnect; and any other pertinent optical properties of the one gigabit per second interconnect for fabrication and testing.

Savich, Gregory R.↗

NASA NTRS1994

H2O-H2SO4 system in Venus' clouds and OCS, CO, and H2SO4 profiles in Venus' troposphere

A coupled problem of diffusion and condensation is solved for the H2SO4-H2O system in Venus' cloud layer. The position of the lower cloud boundary and profiles of the H2O and H2SO4 vapor mixing ratios and of the H2O/H2SO4 ratio of sulfuric acid aerosol and its flux are calculated as functions of the column photochemical production rate of sulfuric acid, Phi(sub H2SO4). Variations of the lower cloud boundary are considered. Our basic model, which is constrained to yield f(sub H2O) (30 km) = 30 ppm (Pollack et al. 1993), predicts the position of the lower cloud boundary at 48.4 km coinciding with the mean Pioneer Venus value, the peak H2SO4 mixing ratio of 5.4 ppm, and the H2SO4 production rate Phi(sub H2SO4) = 2.2 x 10(exp 12)/sq cm/s. The sulfur to sulfuric acid mass flux ratio in the clouds is 1:27 in this model, and the mass loading ratio may be larger than this value if sulfur particles are smaller than those of sulfuric acid. The model suggests that the extinction coefficient of sulfuric acid particles with radius 3.7 micrometers (mode 3) is equal to 0.3/km in the middle cloud layer. The downward flux of CO is equal to 1.7 x 10(exp 12)/sq cm/s in this model. Our second model, which is constrained to yield f(sub H2O) = 10 ppm at the lower cloud boundary, close to the value measured by the Magellan radiooccultations, predicts the position of this boundary to be at 46.5 km, which agrees with the Magellan data; f(sub H2O) (30 km) = 90 ppm, close to the data of Moroz et al. (1983) at this altitude; Phi(sub H2SO4) = 6.4 x 10(exp 12)/sq cm/s; and Phi(sub co) = 4.2 x 10(exp 12)/sq cm/s. The S/H2SO4 flux mass ratio is 1:18, and the extinction coefficient of the mode 3 sulfuric acid particles is equal to 0.9 km in the middle cloud layer. A strong gradient of the H2SO4 vapor mixing ratio near the bottom of the cloud layer drives a large upward flux of H2SO4, which condenses and forms the excessive downward flux of liquid sulfuric acid, which is larger by a factor of 4-7 than the flux in the middle cloud layer. This is the mechanism of formation of the lower cloud layer. Variations of the lower cloud layer are discussed. Our modeling of the OCS and CO profiles in the lower atmosphere measured by Pollack et al. (1993) provides a reasonable explanation of these data and shows that the rate coefficient of the reaction SO3 + CO yields CO2 + SO2 is equal to 10(exp -11) exp(-(13,100 +/- 1000)/T) cc/s. The main channel of the reaction between SO3 and OCS is CO2 + (SO)2, and its rate coefficient is equal to 10(exp -11) exp(-(8900 +/- 500)/T) cc/s. In the conditions of Venus' lower atmosphere, (SO)2 is removed by the reaction (SO)2 + OCS yields CO + S2 + SO2. The model predicts an OCS mixing ratio of 28 ppm near the surface.

Krasnopolsky, V. A.↗

NASA NTRSDate not supplied

A Study of the Swelling and Bursting of the High Pressure Coupling (HPC) of a Two-Stage Light Gas Gun

• Need for data at muzzle velocities of 7.0 km/s and above to simulate impacts of space debris and naturally occurring meteoroids • The two-stage light gas gun is a well known way of achieving these velocities • There is a danger of significant swelling and even bursting of the HPC (high pressure coupling). • A consistent picture of the swelling and bursting of the HPCs of two Ames two-stage light gas guns is developed, based on the following: 1. Experimental measurements of the swelling and bursting of the HPCs of two two-stage light gas guns. 2. Stress-strain calculations of the swelling and bursting of thick walled steel tubes. 3. Experimental measurements swelling and bursting of thick walled steel tubes. 4. CFD calculations of the maximum pressures in the HPC of a two-stage light gas gun using the code LGGUN. • It is shown that model developed predicts swelling and bursting that agrees reasonably well with measured data. • Hence, it is suggested that the model could be used in the design of a two-stage light gas guns to assess the risk of swelling and bursting and to give maximum suggested operating conditions for an existing gun.

Ballistic Range↗

DOE OSTI2022

Machine learning predictions for local electronic properties of disordered correlated electron systems

We present a scalable machine learning (ML) model to predict local electronic properties such as on-site electron number and double occupation for disordered correlated electron systems. Our approach is based on the locality principle, or the nearsightedness nature, of many-electron systems, which means local electronic properties depend mainly on the immediate environment. A ML model is developed to encode this complex dependence of local quantities on the neighborhood. We demonstrate our approach using the square-lattice Anderson-Hubbard model, which is a paradigmatic system for studying the interplay between Mott transition and Anderson localization. We develop a lattice descriptor based on the group-theoretical method to represent the on-site random potentials within a finite region. The resultant feature variables are used as input to a multilayer fully connected neural network, which is trained from data sets of variational Monte Carlo (VMC) simulations on small systems. We show that the ML predictions agree reasonably well with the VMC data. Our work underscores the promising potential of ML methods for multiscale modeling of correlated electron systems.

36 MATERIALS SCIENCE↗

NASA NTRS1985

Computations of internal turbulent flow with large separated flow regions

An implicit two-equation turbulence solver in generalized coordinates has been developed and is used in conjunction with the three-dimensional incompressible Navier-Stokes solver, INS3D, to calculate the internal flow in one channel and in an additional channel with a sudden 2:3 expansion. A new and consistent boundary procedure for a low Reynolds number form of the kappa-epsilon turbulence model is chosen to integrate the equations up to the wall. The high Reynolds number form of the equations are integrated using wall functions. The latter approach yields a faster convergence to the steady state solution than the former. For the case of channel flow, both the wall function and wall boundary condition approaches yield results in good agreement with the experimental data. the back-step (sudden expansion) flow is calculated using wall function approach. The predictions are in reasonable agreement with the experimental data.

Kaul, U. K.↗

NASA NTRS1960

Shock Losses in Transonic Compressor Blade Rows

The rather extensive study of the shock losses in transonic compressors can he summarized by the following remarks: 1. A simple flow model can be used to estimate shock losses at the design point for transonic compressor blade rows and results iii reasonable correlation of loss data. It is indicated that shock losses can constitute a sizable portion of the total losses in it transonic compressor rotor. This includes all blade elements at which sonic or higher relative velocities are obtained. 2. Shock losses can he shown to exist across the blade passage (free-stream loss) and by the method of superposition with the blade profile losses result in an estimated design total loss coefficient. 3. The shock configuration was experimentally determined by the rapid pressure rise between the blades as measured by the use of barium titanate crystals. At the minimum loss operating conditions the shock is very similar to that assumed in the simple How model. 4. Shock losses obtained from a more detailed flow model were compared with the losses obtained by the simple flow model. Measured loss distribution from blade to blade closely approaches the analytical shock loss distribution. The measured distribution shows the effect of a shock boundary layer interaction. 5. The analytical method (from the detailed flow model) of determining the shock location ahead of the blade seems to apply reasonably well over a range of incidence angles. The analytical shock losses do not vary a great deal with blade element incidence angles.

Miller, Genevieve R.↗

DOE OSTI2025

Transformation rate maps of dissolved organic carbon in the contiguous US

Riverine dissolved organic carbon (DOC) plays a vital role in regional and global carbon cycles. However, the processes of DOC conversion from soil organic carbon (SOC) and leaching into rivers are insufficiently understood, inconsistently represented, and poorly parameterized, particularly in land surface and Earth system models. As a first attempt to fill this gap, we propose a generic formula that directly connects SOC concentration with DOC concentration in headwater streams, where a single parameter, the transformation rate from SOC in the soil to DOC leaching flux (P r ), accounts for the overall processes governing SOC conversion to DOC and leaching from soils (along with runoff) into headwater streams. We then derive high-resolution P r maps over the contiguous US (CONUS) using SOC data from two different sources: the Harmonized World Soil Database v1.2 (HWSD) and SoilGrids 2.0. Both maps are developed following the same five major steps: (1) selecting independent catchments where observed riverine DOC data are available with reasonable quality; (2) estimating catchment-average SOC for the independent catchments; (3) estimating the P r values for these catchments based on the generic formula and catchment-average SOC; (4) developing a predictive model of P r with machine learning (ML) techniques and catchment-scale climate, hydrology, geology, and other attributes; and (5) deriving a national map of P r based on the ML model. For evaluation, we compare the DOC concentration derived using the P r map and the observed DOC concentration values at evaluation catchments. The resulting mean absolute scaled error and coefficient of determination are 0.73 and 0.47 for the HWSD-based model and 0.58 and 0.72 for the SoilGrids-based model, respectively, suggesting the effectiveness of the overall methodology. Efforts to constrain uncertainty and evaluate sensitivity of P r to different factors are discussed. To illustrate the use of such maps, we derive a riverine DOC concentration reanalysis dataset over CONUS. The two P r maps, robustly derived and empirically validated, lay a critical cornerstone for better simulating the terrestrial carbon cycle in land surface and Earth system models. Our findings not only set a foundation for improving our predictive understanding of the terrestrial carbon cycle at the regional and global scales, but also hold promises for informing policy decisions related to decarbonization and climate change mitigation. The data presented in this study are publicly available at https://doi.org/10.5281/zenodo.14563816 (Li et al., 2024).

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2021

TRANSIENT THERMAL MODELING OF THE HIGH BURNUP DEMONSTRATION RESEARCH PROJECT CASK USING STAR-CCM+ AND COBRA-SFS

The Department of Energy in collaboration with the Electric Power Research Institute is in the process of conducting the High Burnup Demonstration Project. Where the objective is to characterize the performance of high-burnup fuel in long term storage. As part of this demonstration, a TN-32B dry storage cask was instrumented and loaded with spent nuclear fuel at North Anna Nuclear Generating Station in November 2017. The project cask provides a unique opportunity to gain information on spent fuel and cask performance from an in-service operational system. The cask was instrumented with thermocouples inside of the fuel assemblies, then loaded and dried using normal procedures. After the drying process a thermal soak period where the cask was left indoors was used to obtain steady temperatures for model comparison along with surface temperatures. This paper details thermal modeling validation work that was done to model both the steady state and transient cases. Two modeling tools were used to predict temperatures in the cask. The general purpose CFD and heat transfer code STAR-CCM+ was used with both a detailed pin-by-pin model was used along with a more efficient k-effective simplification. The other code is COBRA-SFS, a purpose-built detailed thermal modeling tool developed and maintained at PNNL. Results from all the modeling tools were compared blind to the test data. Each model utilized general design information and compared reasonably well to the blind steady state data using convection and surface temperature boundary conditions. Post-test adjustments were later made to better reflect the “as built” conditions of the cask. Based on lessons learned from the steady state analysis the vacuum drying process was modeled in both codes. Utilizing the measured loading conditions the transient models were able to compare very well with the measured data. Overall the project showed an ability to model spent fuel storage conditions very well and future work is planned to generalize the methodology used for vacuum drying.

Thermal Analysis, Spent Fuel, Dry Storage↗

NASA NTRS1973

A description of a general use automated data reduction system

A data system developed for use in tests in arc jet facilities has the flexibility to accommodate various types of data, time resolution, and number of data signals. Data are tape-recorded, because a direct link to the IBM 360/67 computer is prohibited for safety reasons. The system is analogous to manual data reduction to promote user understanding and acceptance. Data computations may be performed in an interactive ('conversational') mode, for minimum delay in data reduction, or by card input. Reduced data are listed and also stored for plotting of further computational use. Operational costs of the system are reasonable, and both large and small data reduction tasks are handled readily.

Wakefield, R. M.↗

NASA NTRS1994

Day 640 infrared line and continuum measurements: Dust formation in SN 1987A

We have measured day 640-645 line and continuum spectra of (Ni II) 6.6 micrometer (Ne II) 12.8 micrometer (line emission was not detected), and (Fe II) 17.9 and 26.0 micrometer from SN 1987A. The high velocity feature at v(sub HVF) approximately 3900 km/sec found in both of our day 410 (Fe II) spectra is again detected in the day 640 (Ni II) spectrum, although the signal-to-noise of the day 640 (Fe II) spectra is insufficient to show this feature. The continuum fluxes provide clear evidence for the formation of dust between day 410 and day 640 and are best fitted by a graybody spectrum with a temperature of 342 +/- 17 K at day 640 and a surface area corresponding to a minimum dust velocity v(sub dust) = 1910 +/- 170 km/sec. Optically thin dust emissivity laws proportional to lambda(exp -1) or lambda(exp -2) are inconsistent with the data. Either the dust grains are large (radius a much greater than 4 micrometer and radiate like individual blackbodies, or else they are located in clumps optically thick in the 6-26 micrometer range. The (Ni II) 6.6 micrometer line flux yields a minimum Ni(+) mass of 5.8 +/- 1.6 x 10(exp -4) solar mass and a Ni/Fe abundance ratio of 0.06 +/- 0.02, equal to the solar value. The ratio of the two (Fe II) line profiles implies a gas temperature 2600 +/- 700 K, a drop of 1800 +/- 800 K from our day 410 measurement. The (Fe II) 26.0 micrometer line flux has decreased by a factor of 2 and the day 640 (Ni II) profile is blueshifted by -440 +/- 270 km/sec, relative to observations before day 500. We show that the decrease in the (Fe II) flux and the blueshift are not produced by a decrease in electron scattering optical depth, electron density, or temperature, but rather are probably due to obscuration by the same dust which produces the infrared continuum. This supports the interpretation that the dust spectrum is produced by optically thick clumps. We discuss possible explanations for the discrepancy between the mass of Fe(+) detected and the total iron mass required to power the light curve. The decrease in the (Fe II) fluxes relative to the decrease required to account for the blueshifts of optical lines from non-iron-group elements and the similarity between v(sub dust) and the Ni(+) expansion velocity imply a spatial association between the dust clumps and the iron-group elements. In addition, the larger blueshift observed for the near and far-infrared, heavy metal transitions relative to non-iron-group lines suggests that the iron-group elements are somewhat segregated from lighter elements such as the Mg(sup 0) and O(sup 0) responsible for shorter wavelength lines. We speculate that FeS may be an important constituent of the dust. A comparison of our line profiles with radiative transfer models shows that while power law and exponential density distributions yield reasonable fits to the data, polytrope distributions provided significantly worse agreement. The best fits require a substantial fraction of the iron to be undetectable, and are consistent with maximum expansion velocities of v(sub max) approximately 3000 km/sec.

Colgan, Sean W. J.↗

DOE OSTI2021

Watermarks in stream processing systems: semantics and comparative analysis of Apache Flink and Google cloud dataflow

Streaming data processing is an exercise in taming disorder: from oftentimes huge torrents of information, we hope to extract powerful and timely analyses. But when dealing with streaming data, the unbounded and temporally disordered nature of real-world streams introduces a critical challenge: how does one reason about the completeness of a stream that never ends? In this paper, we present a comprehensive definition and analysis of watermarks, a key tool for reasoning about temporal completeness in infinite streams.First, we describe what watermarks are and why they are important, highlighting how they address a suite of stream processing needs that are poorly served by eventually-consistent approaches:• Computing a single correct answer, as in notifications.• Reasoning about a lack of data, as in dip detection.• Performing non-incremental processing over temporal subsets of an infinite stream, as in statistical anomaly detection with cubic spline models.• Safely and punctually garbage collecting obsolete inputs and intermediate state.• Surfacing a reliable signal of overall pipeline health.Second, we describe, evaluate, and compare the semantically equivalent, but starkly different, watermark implementations in two modern stream processing engines: Apache Flink and Google Cloud Dataflow.

Akidau, Tyler↗

NASA NTRS1974

The application of ERTS imagery to mapping snow cover in the western United States

The author has identified the following significant results. In much of the western United States a large part of the utilized water comes from accumulated mountain snowpacks; thus, accurate measurements of snow distributions are required for input to streamflow prediction models. The application of ERTS-1 imagery for mapping snow has been evaluated for two geographic areas, the Salt-Verde watershed in central Arizona and the southern Sierra Nevada in California. Techniques have been developed to identify snow and to differentiate between snow and cloud. The snow extent for these two drainage areas has been mapped from the MSS-5 (0.6 - 0.7 microns) imagery and compared with aerial survey snow charts, aircraft photography, and ground-based snow measurements. The results indicate that ERTS imagery has substantial practical applications for snow mapping. Snow extent can be mapped from ERTS-1 imagery in more detail than is depicted on aerial survey snow charts. Moreover, in Arizona and southern California cloud obscuration does not appear to be a serious deterrent to the use of satellite data for snow survey. The costs involved in deriving snow maps from ERTS-1 imagery appear to be very reasonable in comparison with existing data collection methods.

Barnes, J. C.↗