Search NASA⌕ Search

SEARCH · Search NASA

Results for “Explanability”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

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

At least 1,603 records · Page 89

Screening in nanowires and nanocontacts: field emission, adhesion force, and contact resistance

The explanations of several nanoscale phenomena such as the field enhancement factor in field emission, the large decay length of the adhesion force between a metallic tip and a surface, and the contact resistance in a nanowire break junction have been elusive. Here we develop an analytical theory of Thomas-Fermi screening in nanoscale structures. We demonstrate that nanoscale dimensions give rise to an effective screening length that depends on geometry and physical boundary conditions. The above phenomena are shown to be manifestations of the effective screening length.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

Ionically-mediated electromechanical hysteresis in transition metal oxides

Electromechanical activity, remanent polarization states, and hysteresis loops in paraelectric TiO2 and SrTiO3 are observed. The coupling between the ionic dynamics and incipient ferroelectricity in these materials is analyzed using extended Ginsburg Landau Devonshire (GLD) theory. The possible origins of electromechanical coupling including ionic dynamics, surface-charge induced electrostriction, and ionically-induced ferroelectricity are identified. For the latter, the ionic contribution can change the sign of first order GLD expansion coefficient, rendering material effectively ferroelectric. These studies provide possible explanation for ferroelectric-like behavior in centrosymmetric transition metal oxides.

Kim, Yunseok↗

Melting behavior of H[subscript 2]O at high pressures and temperatures

Water plays an important role in the physics and chemistry of planetary interiors. In situ high pressure-temperature Raman spectroscopy and synchrotron x-ray diffraction have been used to examine the phase diagram of H{sub 2}O. A discontinuous change in the melting curve of H{sub 2}O is observed at approximately 35 GPa and 1040 K, indicating a triple point on the melting line. The melting curve of H{sub 2}O increases significantly above the triple point and may intersect the isentropes of Neptune and Uranus. Solid ice could therefore form in stratified layers at depth within these icy planets. The extrapolated melting curve may also intersect with the geotherm of Earth's lower mantle above 60 GPa. The presence of solid H{sub 2}O would result in a jump in the viscosity of the mid-lower mantle and provides an additional explanation for the observed higher viscosity of the mid-lower mantle.

43 PARTICLE ACCELERATORS↗

Exploiting diurnal variations to evaluate the ISCCP-FD flux calculations and radiative-flux-analysis-processed surface observations from BSRN, ARM, and SURFRAD

Based on monthly-3-hourly and 3-hourly mean surface radiative fluxes and their associated meteorological parameters for 2004 from the International Satellite Cloud Climatology Project-FD (ISCCP-FD) and the Radiative Flux Analysis method-Produced Surface Observations (RFA-PSO) for 15 high-quality-controlled surface stations, operated by the Baseline Surface Radiation Network (BSRN), the Atmospheric Radiation Measurement (ARM) and the National Oceanic and Atmospheric Administration's Surface Radiation budget network (SURFRAD), this work, goes beyond the previous validation for FD against surface observation by introducing the Meteorological Similarity Comparison Method (MSCM) to make a more precise, mutual evaluation of both FD and PSO products. The comparison results in substantial uncertainty reduction and provides reasonable physical explanations for the flux differences. This approach compares fluxes for cases where the atmospheric and surface physical properties (specifically, the input parameters for radiative transfer model) are as close as possible to the values determined at the observational sites by matching the RFA-produced cloud fraction (CF) and/or optical thickness (Tau), etc., or alternatively, by directly changing the model input variables for FD to match PSO values, and using such-produced matched sub-datasets to make more accurate comparisons based on more similar meteorological environments between FD and PSO. The crucial part is the availability of flux-associated meteorological parameters from RFA-PSO, which was only recently made available that makes this work possible. For surface downwelling shortwave(SW) flux (SWdn) and its two components, diffuse (Dif) and direct (Dir), uncertainty for monthly mean is 15, 15 and 17 W/m 2 , respectively, smaller than the separately estimated uncertainty values from both FD and PSO. When applying MSCM by reducing their CF difference, the differences can be reduced by a factor of 2. The strength of MSCM is particularly shown in the comparisons of diurnal variations. For clear sky, reducing the FD values of aerosol optical depth (AOD) by 50% to approximately match the PSO values brings all downward SW flux components into substantial agreement. For cloudy scenes, when both CF and Tau are matched to within 0.1 – 0.25 and ~10, respectively, the majority of the SW flux components have nearly-perfect agreement between FD and PSO. The best restriction differences are not zero indicates the influence of other parameters that are not accounted for yet. For longwave (LW) fluxes, general evaluation also confirms uncertainty values for FD and PSO less than separately estimated. When applying MSCM to CF and surface air temperature, the agreement is substantially improved. For downwelling LW diurnal variation comparison, FD shows good agreement with PSO for both RFA-defined or true clear sky but overestimates the amplitude for cloudy sky by 3-7 W/m 2 , which may be caused by different sensitivities to cirrus clouds. For upwelling LW diurnal cycle, the situation is reversed; FD now underestimates the diurnal amplitude for all and clear sky but generally agrees for overcast (CF > 0.7). The combined effect of downwelling and upwelling LW fluxes results in FD's underestimates of the diurnal variation of the net-LW-loss for all the scenes by up to 10 W/m 2 , although the daily mean net loss is more accurate. Therefore, in terms of amplitude and phase, both FD and PSO seem to have caught correct diurnal variations.

58 GEOSCIENCES↗

Effect of Spatial Concentration Fluctuations on Effective Kinetics in Diffusion-Reaction Systems

The effect of spatial concentration fluctuations on reaction of two solutes, A+B$\rightarrow$C, is considered. In the absence of fluctuations, concentration of solutes decays as $A_{det}=B_{det} \sim t^{-1}$. Contrary to this, experimental and numerical studies suggest that concentrations decay significantly slower. Existing theory suggests $t^{-d/4}$ scaling in the asymptotic regime ($d$ is the dimensionality of the problem). Here we study the effect of fluctuations using the classical diffusion-reaction equation with random initial conditions. Initial concentrations of the reactants are treated as correlated random fields. We use the method of moment equations to solve the resulting stochastic diffusion-reaction equation and obtain a solution for the average concentrations that deviates from $\sim t^{-1}$ to $\sim t^{-d/4}$ behavior at time $t^*$. Finally, we derive analytical expressions for the time $t^*$ as a function of Damk\"{o}hler number and the coefficient of variation of the initial concentration. Our analytical results support earlier explanations of the deviation of average concentrations from $A_{det}=B_{det} \sim t^{-1}$ behavior, which attributed this change of scaling to the creation of isolated {\it islands} of A and B.

uncertainty quantification, stochastic differentia↗

Modeling Intragranular Diffusion in Low-Connectivity Granular Media

Diffusive exchange of solutes between bulk water in an aquifer and water in the intragranular pores of the solid phase remains confusing after decades of study. In a previous paper, we reviewed some of the explanations, and suggested that the disparities between observation and theory were largely due to low connectivity of the intragranular pores. Low connectivity indicates that a useful conceptual framework is percolation theory, which guided our analysis. The present study was initiated to improve the finite difference (FD) model presented in the previous paper, and to test that new model rigorously against new random walk (RW) simulations of diffusion in low-connectivity porous spheres starting from non-equilibrium. The new FD model calculates diffusion separately in the infinite cluster and the finite clusters, and closely matches the new, more complex RW results. The percolation-theory based description of the new model is fairly simple, and can readily be incorporated into existing FD models. The simulations showed that the combination of low intragranular pore connectivity, and out-diffusion initiated at diffusive non-equilibrium, can produce diffusive behavior that appears as if the solute had undergone slow sorption, even in the absence of any sorption process. This mechanism may help explain some hitherto confusing aspects of intragranular diffusion.

diffusion↗

Do CMIP5 Models Reproduce Observed Low‐Frequency North Atlantic Jet Variability?

Abstract The magnitude of observed multidecadal variations in the North Atlantic Oscillation (NAO) in winter is at the upper end of the range simulated by climate models and a clear explanation for this remains elusive. Recent research shows that observed multidecadal winter NAO variability is more strongly associated with North Atlantic (NA) jet strength than latitude, thus motivating a comprehensive comparison of NA jet and NAO variability across the Coupled Model Intercomparison Project Phase 5 (CMIP5) models. Our results show that the observed peak in multidecadal jet strength variability is even more unusual than NAO variability when compared to the model‐simulated range across 133 historical CMIP5 simulations. Some CMIP5 models appear capable of reproducing the observed low‐frequency peak in jet strength, but there are too few simulations of each model to clearly identify which. It is also found that an observed strong multidecadal correlation between jet strength and NAO since the mid‐nineteenth century may be specific to this period.

Bracegirdle, Thomas J.↗

Kinetic Control on the Depth Distribution of Superdeep Diamonds

Superdeep diamonds contain unique information from the sublithospheric regions of Earth's interior. Recent studies imply that reaction between subducted carbonate and iron metal in the mantle plays an important role in the production of superdeep diamonds. It is unknown if this reaction is kinetically feasible in cold slabs subducted into the deep mantle. In this work we introduce experimental data on real-time tracking of the magnesite-iron reaction at high pressures and high temperatures to demonstrate the production of diamond at the surface conditions of cold slabs in the transition zone and lower mantle. Our data reveal that the diamond production rate has a positive temperature dependence and a negative pressure dependence, and along a slab geotherm it decreases by a factor of three at pressures from 14.4 to 18.4 GPa. This rate reduction provides an explanation for the rarity of superdeep diamonds from the interior of the mantle transition zone.

58 GEOSCIENCES↗

Is Shallow Convection Sensitive to Environmental Heterogeneities?

The key assumption underlying convection parameterizations is that rising plumes develop in a horizontally homogeneous environment. With this in mind, we investigate the impact of environmental cloud layer heterogeneities on shallow convection using large-eddy simulation that applies a master-slave methodology. In the master-slave approach, two independent sets of thermodynamic variables (master and slave) are driven by one dynamics (coupled with the master) to remove the impact of internal variability of the dynamical system considered. The two thermodynamic sets include either a realistic heterogeneous environment or an environment that is homogenized outside clouds. The key is that homogenization also excludes subsiding shells surrounding cumulus clouds. The results show a small impact of the homogenization on cloud field properties. Here, the physical explanation highlights the role of the subsiding shell shielding a shallow cumulus cloud from its environment.

58 GEOSCIENCES↗

Do Surface and Air Temperatures Contain Similar Imprints of Evaporative Conditions?

Abstract Generally, surface and air temperatures seem closely related but we show that they respond differently to evaporative conditions. We evaluate the temperature increase in response to solar radiation for different evaporative fractions, using observations from the Southern Great Plains. The warming rate of air temperature decreases only by 1.7 × 10 −3 K/(W m −2 ) from dry to moist conditions compared to a stronger reduction by 14 × 10 −3 K/(W m −2 ) for surface temperature. The weaker response of air temperature to evaporative fraction is explained by the larger growth of boundary layer on drier days, which suppresses the warming of air. Estimates based on this explanation reproduce the warming rate of air temperature in observations. Our results show that diurnal variations of surface temperatures contain imprints of evapotranspiration while air temperatures do not. These findings appear important to be considered when using, analyzing, or interpreting temperature data in studies dealing with climate change, hydrology, or land‐atmosphere interactions.

Panwar, Annu↗

Stabilizing Effect of High Pore Fluid Pressure on Slip Behaviors of Gouge-bearing Faults

We conducted experiments to investigate the influence of pore fluid pressure on the frictional strength and slip behavior of gouge bearing faults. Saw cut porous sandstone samples with a layer of gouge powders placed between the precut surfaces were deformed in the conventional triaxial loading configuration. A series of velocity-step tests were performed to measure the response of the friction coefficient to variations in sliding velocity. Pore volume changes were monitored during shearing of the gouge. Our results demonstrate that under constant effective pressure, increasing pore pressure stabilizes the frictional slip of faults with all four gouge materials including antigorite, olivine, quartz, and chrysotile. The stabilizing effect is the strongest in antigorite gouge, which shows an evolution of friction parameters from velocity-weakening toward velocity-strengthening behavior with increasing pore pressure. Experiments with controlled pore volume show that the pore volume reduction diminishes under high pore fluid pressures, implying an increasing dilation component at these conditions. The dilatant hardening mechanism can explain the observed strengthening. These results provide a possible explanation to the observed spatial correlation between slow slip events and high pore pressure in many subduction zones.

58 GEOSCIENCES↗

A Thin Film Viscoplastic Theory for Calving Glaciers: Toward a Bound on the Calving Rate of Glaciers

Projections of the growth and demise of ice sheets and glaciers require physical models of the processes governing flow and fracture of ice. The flow of glacier ice has been treated using increasingly sophisticated models. By contrast, fracture, the process ultimately responsible for half of the mass lost from ice sheets through iceberg calving, is often included using ad hoc parameterizations. In this study we seek to bridge this gap by introducing a model where ice obeys a power law rheology appropriate for intact ice below a yield strength. Above the yield strength, we introduce a separate rheology appropriate for the flow of heavily fractured ice, where ice deformation occurs more readily along faults and fractures. We show that, provided the motion of fractured ice is sufficiently rapid compared to that of intact ice, the behavior of glaciers depends solely on the rheology of intact ice and the yield strength of ice and is insensitive to the precise rheology of fractured ice. Moreover, assuming that glacier ice is unyielded allows us to bound the long‐term average rate of terminus advance, providing a first principles estimate of rates of retreat associated with the marine ice cliff instability. We illustrate model behavior using idealized geometries and climate forcing and show that the model not only exhibits realistic patterns of advance and retreat but also has the potential to exhibit hysteresis. This hysteresis could provide an explanation for the sudden onset of rapid retreat observed in marine‐terminating glaciers.

Bassis, J. N.↗

Decreased Soil Organic Matter in a Long-Term Soil Warming Experiment Lowers Soil Water Holding Capacity and Affects Soil Thermal and Hydrological Buffering

Long-term soil warming can decrease soil organic matter (SOM), resulting in self-reinforcing feedback to the global climate system. We investigated additional consequences of SOM reduction for soil water holding capacity (WHC) and soil thermal and hydrological buffering. At a long-term soil warming experiment in a temperate forest in the northeastern United States, we suspended the warming treatment for 104 days during the summer of 2017. The formerly heated plot remained warmer (+0.39 °C) and drier (-0.024 cm 3 H 2 O cm -3 soil) than the control plot throughout the suspension. We measured decreased SOM content (-0.184 g SOM g -1 for O horizon soil, -0.010 g SOM g -1 for A horizon soil) and WHC (-0.82 g H 2 O g -1 for O horizon soil, -0.18 g H2O g -1 for A horizon soil) in the formerly heated plot relative to the control plot. Reduced SOM content accounted for 62% of the WHC reduction in the O horizon and 22% in the A horizon. We investigated differences in SOM composition as a possible explanation for the remaining reductions with Fourier transform infrared (FTIR) spectra. We found FTIR spectra that correlated more strongly with WHC than SOM, but those particular spectra did not differ between the heated and control plots, suggesting that SOM composition affects WHC but does not explain treatment differences in this study. We conclude that SOM reductions due to soil warming can reduce WHC and hydrological and thermal buffering, further warming soil and decreasing SOM. This feedback may operate in parallel, and perhaps synergistically, with carbon cycle feedbacks to climate change.

54 ENVIRONMENTAL SCIENCES↗

A Benchmark to Test Generalization Capabilities of Deep Learning Methods to Classify Severe Convective Storms in a Changing Climate

Abstract This is a test case study assessing the ability of deep learning methods to generalize to a future climate (end of 21st century) when trained to classify thunderstorms in model output representative of the present‐day climate. A convolutional neural network (CNN) was trained to classify strongly rotating thunderstorms from a current climate created using the Weather Research and Forecasting model at high‐resolution, then evaluated against thunderstorms from a future climate and found to perform with skill and comparatively in both climates. Despite training with labels derived from a threshold value of a severe thunderstorm diagnostic (updraft helicity), which was not used as an input attribute, the CNN learned physical characteristics of organized convection and environments that are not captured by the diagnostic heuristic. Physical features were not prescribed but rather learned from the data, such as the importance of dry air at mid‐levels for intense thunderstorm development when low‐level moisture is present (i.e., convective available potential energy). Explanation techniques also revealed that thunderstorms classified as strongly rotating are associated with learned rotation signatures. Results show that the creation of synthetic data with ground truth is a viable alternative to human‐labeled data and that a CNN is able to generalize a target using learned features that would be difficult to encode due to spatial complexity. Most importantly, results from this study show that deep learning is capable of generalizing to future climate extremes and can exhibit out‐of‐sample robustness with hyperparameter tuning in certain applications.

54 ENVIRONMENTAL SCIENCES↗

Identifying Key Drivers of Wildfires in the Contiguous US Using Machine Learning and Game Theory Interpretation

Abstract Understanding the complex interrelationships between wildfire and its environmental and anthropogenic controls is crucial for wildfire modeling and management. Although machine learning (ML) models have yielded significant improvements in wildfire predictions, their limited interpretability has been an obstacle for their use in advancing understanding of wildfires. This study builds an ML model incorporating predictors of local meteorology, land‐surface characteristics, and socioeconomic variables to predict monthly burned area at grid cells of 0.25° × 0.25° resolution over the contiguous United States. Besides these predictors, we construct and include predictors representing the large‐scale circulation patterns conducive to wildfires, which largely improves the temporal correlations in several regions by 14%–44%. The Shapley additive explanation is introduced to quantify the contributions of the predictors to burned area. Results show a key role of longitude and latitude in delineating fire regimes with different temporal patterns of burned area. The model captures the physical relationship between burned area and vapor pressure deficit, relative humidity (RH), and energy release component (ERC), in agreement with the prior findings. Aggregating the contribution of predictor variables of all the grids by region, analyses show that ERC is the major contributor accounting for 14%–27% to large burned areas in the western US. In contrast, there is no leading factor contributing to large burned areas in the eastern US, although large‐scale circulation patterns featuring less active upper‐level ridge‐trough and low RH two months earlier in winter contribute relatively more to large burned areas in spring in the southeastern US.

54 ENVIRONMENTAL SCIENCES↗

Testing the Recycled Gabbro Hypothesis for the Origin of “Ghost Plagioclase” Melt Signatures Using 87 Sr/ 86 Sr of Individual Olivine-Hosted Melt Inclusions From Hawai'i

Melt inclusions with large, positive Sr anomalies have been described in multiple tectonic settings, and the origins of this unusual geochemical feature are debated. Three origins have been proposed, all involving plagioclase as the source of the elevated Sr: (i) direct assimilation of plagioclase-rich lithologies, (ii) recycled lower oceanic gabbro in the mantle source, and (iii) shallow-level diffusive interaction between present day lower oceanic crust (i.e., plagioclase-bearing lithologies) and the percolating melt. A “ghost plagioclase” signature (i.e., a large, positive Sr anomaly without associated high Al 2 O 3 ) is present in melt inclusions from Mauna Loa. Here, we present new 87 Sr/ 86 Sr measurements of individual olivine-hosted melt inclusions from three Hawaiian volcanoes, Mauna Loa, Loihi, and Koolau. The data set includes a Mauna Loa melt inclusion with the highest reported Sr anomaly (or highest (Sr/Ce) N , which is 7.2) for Hawai'i. All melt inclusions have 87 Sr/ 86 Sr values within the range reported previously for the lavas from each volcano. Critically, the 87 Sr/ 86 Sr of the high (Sr/Ce) N melt inclusion lies within the narrow range of 87 Sr/ 86 Sr for Mauna Loa melts that lack high (Sr/Ce) N signatures. Therefore, to explain the high (Sr/Ce) N ratio of the ghost plagioclase signature using an ancient recycled gabbro, the gabbro-infused mantle source would have had to evolve, by chance, to have the same 87 Sr/ 86 Sr as the source of the Mauna Loa melts that lack a recycled gabbro (ghost plagioclase) signature. Alternatively, shallow-level diffusive interactions between Mauna Loa plagioclase-rich cumulates and a percolating mantle-derived melt provides a simpler explanation for the presence of the high (Sr/Ce) N Mauna Loa melts.

58 GEOSCIENCES↗

Dynamic Amplification of Subtropical Extreme Precipitation in a Warming Climate

Projected precipitation changes in a warming climate vary considerably, spatially, and between intensities. The changes can be greater or less than the ~7% K -1 Clausius-Clapeyron (CC) prediction, owing to dynamic effects. Using two global-climate-model large ensembles, we quantify the dynamically induced changes to precipitation extremes from the present (1996–2005) to late-21st-century (2071–2080) climates, as a function of recurrence interval, focusing on the subtropics. We separate non-CC changes into a term proportional to the present-day vertical-velocity spatial pattern (i.e., an amplification or damping thereof by a constant factor) and a residual. The amplitude term varies with recurrence interval, approximately canceling (doubling) CC for moderate (large) extremes, increasing precipitation variability. Contrastingly, the residual is quasi-uniform across recurrence intervals but spatially heterogeneous, weakening extremes over dry zones. This residual may be related to Hadley cell expansion, although this explanation is insufficient to explain many features, and other possible mechanisms are discussed.

54 ENVIRONMENTAL SCIENCES↗

Why Are There so Few Reports of High-Energy Electron Drift Resonances? Role of Radial Phase Space Density Gradients

Models of monochromatic Pc5 (2–7 mHz) ultralow frequency (ULF) wave interactions with high energy (greater than ~1 MeV) electrons predict drift resonant interactions that can cause rapid radial transport and acceleration. There are few reports of electron drift resonance at energies greater than ~1 MeV, in contrast to lower energies; moreover, all previous reports occur in the aftermath of interplanetary shocks. These two facts are difficult to reconcile with theory and numerical simulations predicting that greater than ~1 MeV drift resonances should occur more often and in a wider variety of driving conditions. In this study, we show that a combination of observational sampling biases and nominal radial phase space density gradients is one explanation for this discrepancy between theory and observations. In particular, we examine electron dynamics in two case studies with very similar satellite coverage, solar wind conditions, and Pc5 wave properties, yet with different radial phase space density profiles. Using global wave and particle observations, we show that the events have vastly different particle responses despite having similar wave properties. Placing these results in context with past studies, we further show here that nominal radial PSD gradients near geostationary orbit can mask the expected drift resonance particle response and explain (1) the small number of past greater than ~1 MeV drift resonance reports and (2) the restriction of these reports to interplanetary shock events. We argue that future observational studies characterizing radial transport via drift resonance should examine global particle dynamics, including observations of the radial phase space density profile.

79 ASTRONOMY AND ASTROPHYSICS↗