Search NASA⌕ Search

SEARCH · Search NASA

Results for “gradient”

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 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
NASA NTRS1994

Colors and color gradients in bulges of galaxies

We have obtained surface photometry in U, B, R, and I for a complete optically selected sample of 45 early-type spiral galaxies, to investigate the colors and color gradients of spiral bulges. Color profiles in U-R, B-R, U-B, and R-I have been determined in wedges opening on the semiminor axes. Based on several criteria, like the smoothness of the color profiles, the absence of dust lanes, and the central colors, we have defined a subsample of 18 objects whose colors are largely unaffected by dust. We believe such colors are suitable for inferring properties of the stellar populations of bulges. We find that the colors of bulges are predominantly bluer than those of ellipticals. This result holds even when bulges are compared to ellipticals of the same luminosity, and indicates that bulges are younger and/or more metal poor than old elliptical galaxies. Most bulges do not reach solar metallicities. Bulges show predominantly negative color gradients (bluer outward). For bright bulges (M(sup Bulge sub R) is less than -20.0), the magnitude of the gradient increases with bulge luminosity. For fainter bulges, gradients scatter around large negative values. The behavior of color gradients as a function of bulge luminosity suggests different formation mechanisms for faint and bright spheroids. For bright bulges, the scaling of gradients with luminosity suggests a formation process involving dissipation. The similarity with ellipticals suggests that the formation of the disk did not affect the stellar populations of the bulge in a major way. For small bulges (M(sub R) is greater than -20), the existence of pronounced color gradients suggests a different formation mechanism. For these objects, the presence of the disk may have severely affected the radial population distribution in the bulge.

Balcells, Marc↗
NASA NTRS1995

Anomalous cosmic ray oxygen gradients throughout the heliosphere

We have used data from the Solar, Anomalous, and Magnetospheric Particle Explorer (SAMPEX), Ulysses, Voyager 1, Voyager 2, and Pioneer 10 spacecraft to determine the radial and latitudinal gradients of anomalous cosmic ray oxygen at 10 MeV/nuc during the last half of 1993. These five spacecraft cover radial distances from 1 AU (SAMPEX) to 58 AU (P10) and latitudes to 41 deg S (Ulysses) and 32 deg N (V1). We find that the radial gradient is a decreasing function of radial distance, approximately r(exp -n), with n = 1.7 +/- 0.7. The large-scale radial gradient between the inner and outer heliosphere is much smaller than it was during the last solar minimum period in approximately 1987. The latitudinal gradient is small and positive, 1.3 +/- 0.4 %/deg, as opposed to the large and negative latitudinal gradients found during 1987, but similar to the small positive latitudinal gradient measured during 1976 for anomalous cosmic ray helium. These observations confirm that effects of curvature and gradient drift in the large scale magnetic field of the Sun are important for establishing the three-dimensional intensity distributions of these particles in the heliosphere during periods of solar minimum conditions.

Cummings, A. C.↗
NASA NTRS1997

Effect of Pressure Gradients on Plate Response and Radiation in a Supersonic Turbulent Boundary Layer

Using the model developed by the author for zero-pressure gradient turbulent boundary layers, results are obtained for adverse and favorable pressure gradients. It is shown that when a flexible plate is located in an adverse pressure gradient area, it vibrates more than if it were in a favorable pressure gradient one. Therefore the noise generated by the plate in an adverse pressure gradient is much greater than that due to the plate in a favorable pressure gradient. The effects of Reynolds number and boundary layer thickness are also analyzed and found to have the same effect in both adverse and favorable pressure gradient cases. Increasing the Reynolds number is found to increase the loading on the plate and therefore acoustic radiation. An increase in boundary layer thickness is found to decrease the level of the high frequencies and therefore the response and radiation at these frequencies. The results are in good qualitative agreement with experimental measurements.

Frendi, Abdelkader↗
NASA NTRS2000

A Study of Wake Development and Structure in Constant Pressure Gradients

Motivated by the application to high-lift aerodynamics for commercial transport aircraft, a systematic investigation into the response of symmetric/asymmetric planar turbulent wake development to constant adverse, zero, and favorable pressure gradients has been conducted. The experiments are performed at a Reynolds number of 2.4 million based on the chord of the wake generator. A unique feature of this wake study is that the pressure gradients imposed on the wake flow field are held constant. The experimental measurements involve both conventional LDV and hot wire flow field surveys of mean and turbulent quantities including the turbulent kinetic energy budget. In addition, similarity analysis and numerical simulation have also been conducted for this wake study. A focus of the research has been to isolate the effects of both pressure gradient and initial wake asymmetry on the wake development. Experimental results reveal that the pressure gradient has a tremendous influence on the wake development, despite the relatively modest pressure gradients imposed. For a given pressure gradient, the development of an initially asymmetric wake is different from the initially symmetric wake. An explicit similarity solution for the shape parameters of the symmetric wake is obtained and agrees with the experimental results. The turbulent kinetic energy budget measurements of the symmetric wake demonstrate that except for the convection term, the imposed pressure gradient does not change the fundamental flow physics of turbulent kinetic energy transport. Based on the turbulent kinetic energy budget measurements, an approach to correct the bias error associated with the notoriously difficult dissipation estimate is proposed and validated through the comparison of the experimental estimate with a direct numerical simulation result.

Thomas, Flint O.↗
NASA NTRS1997

Arterial Pressure Gradients during Upright Posture and 30 deg Head Down Tilt

Gravity alters local blood pressure within the body so that arterial pressures in the head and foot are lower and higher, respectively, than that at heart level. Furthermore, vascular responses to local alterations of arterial pressure are probably important to maintain orthostatic tolerance upon return to the Earth after space flight. However, it has been difficult to evaluate the body's arterial pressure gradient due to the lack of noninvasive technology. This study was therefore designed to investigate whether finger arterial pressure (FAP), measured noninvasively, follows a normal hydrostatic pressure gradient above and below heart level during upright posture and 30 deg head down tilt (HDT). Seven healthy subjects gave informed consent and were 19 to 52 years old with a height range of 158 to 181 cm. A Finapres device measured arterial pressure at different levels of the body by moving the hand from 36 cm below heart level (BH) to 72 cm above heart level (AH) in upright posture and from 36 cm BH to 48 cm AH during HDT in increments of 12 cm. Mean FAP creased by 85 mmHg transitioning from BH to AH in upright posture, and the pressure gradient calculated from hydrostatic pressure difference (rho(gh)) was 84 mmHg. In HDT, mean FAP decreased by 65 mmHg from BH to AH, and the calculated pressure gradient was also 65 mmHg. There was no significant difference between the measured FAP gradient and the calculated pressure gradient, although a significant (p = 0.023) offset was seen for absolute arterial pressure in upright posture. These results indicate that arterial pressure at various levels can be obtained from the blood pressure at heart level by calculating rho(gh) + an offset. The offset equals the difference between heart level and the site of measurement. In summary, we conclude that local blood pressure gradients can be measured by noninvasive studies of FAP.

Sanchez, E. R↗
NASA NTRS2002

Motion of Drops on Surfaces with Wettability Gradients

A liquid drop present on a solid surface can move because of a gradient in wettability along the surface, as manifested by a gradient in the contact angle. The contact angle at a given point on the contact line between a solid and a liquid in a gaseous medium is the angle between the tangent planes to the liquid and the solid surfaces at that point and is measured within the liquid side, by convention. The motion of the drop occurs in the direction of increasing wettability. The cause of the motion is the net force exerted on the drop by the solid surface because of the variation of the contact angle around the periphery. This force causes acceleration of an initially stationary drop, and leads to its motion in the direction of decreasing contact angle. The nature of the motion is determined by the balance between the motivating force and the resisting hydrodynamic force from the solid surface and the surrounding gaseous medium. A wettability gradient can be chemically induced as shown by Chaudhury and Whitesides who provided unambiguous experimental evidence that drops can move in such gradients. The phenomenon can be important in heat transfer applications in low gravity, such as when condensation occurs on a surface. Daniel et al have demonstrated that the velocity of a drop on a surface due to a wettability gradient in the presence of condensation can be more than two orders of magnitude larger than that observed in the absence of condensation. In the present research program, we have begun to study the motion of a drop in a wettability gradient systematically using a model system. Our initial efforts will be restricted to a system in which no condensation occurs. The experiments are performed as follows. First, a rectangular strip of approximate dimensions 10 x 20 mm is cut out of a silicon wafer. The strip is cleaned thoroughly and its surface is exposed to the vapor from an alkylchlorosilane for a period lasting between one and two minutes inside a desiccator. This is done using an approximate line source of the vapor in the form of a string soaked in the alkylchlorosilane. Ordinarily, many fluids, including water, wet the surface of silicon quite well. This means that the contact angle is small. But the silanized surface resists wetting, with contact angles that are as large as 100 degs. Therefore, a gradient of wettability is formed on the silicon surface. The region near the string is highly hydrophobic, and the contact angle decreases gradually toward a small value at the hydrophilic end away from this region. The change in wettability occurs over a distance of several mm. The strip is placed on a platform within a Plexiglas cell. Drops of a suitable liquid are introduced on top of the strip near the hydrophobic end. An optical system attached to a video camera is trained on the drop so that images of the moving drop can be captured on videotape for subsequent analysis. We have performed preliminary experiments with water as well as ethylene glycol drops. Results from these experiments will be presented in the poster. Future plans include the refinement of the experimental system so as to permit images to be recorded from the side as well as the top, and the conduct of a systematic study in which the drop size is varied over a good range. Experiments will be conducted with different fluids so as to obtain the largest possible range of suitably defined Reynolds and Capillary numbers. Also, an effort will be initiated on theoretical modeling of this motion. The challenges in the development of the theoretical description lie in the proper analysis of the region in the vicinity of the contact line, as well as in the free boundary nature of the problem. It is known that continuum models assuming the no slip condition all the way to the contact line fail by predicting that the stress on the solid surface becomes singular as the contact line is approached. One approach for dealing with this issue has been to relax the no-slip boundary condition using the Navier model. Molecular dynamics simulations of the contact line region show that for a non-polar liquid on a solid surface, the no-slip boundary condition is in fact incorrect near the contact line. Furthermore, the same simulations also show that the usual relationship between stress and the rate of deformation breaks down in the vicinity of the contact line. In developing continuum theoretical models of the system, we shall accommodate this knowledge to the extent possible.

Subramanian, R. Shankar↗
NASA NTRS1991

The role of extracellular free-calcium gradients in gravitropic signalling in maize roots

Gravitropism in roots has been proposed to depend on a downward redistribution of calcium across the root cap. However, because of the many calcium-binding sites in the apoplast, redistribution might not result in a physiologically effective change in the apoplasmic calcium activity. To test whether there is such a change, we measured the effect of gravistimulation on the calcium activity of statocyte cell walls with calcium-specific microelectrodes. Such a measurement must be made on a tissue with gravity sensing cells at the surface. To obtain such a tissue, decapped maize roots (Zea mays L. cv. Golden Cross Bantam) were grown for 31 h to regenerate gravitropic sensitivity, but not root caps. The calcium activity in the apoplasm surrounding the gravity-sensing cells could then be measured. The initial pCa was 2.60 +/- 0.28 (approx 2.5 mM). The calcium activity on the upper side of the root tip remained constant for 10 min after gravistimulation, then decreased 1.7-fold. On the lower side, after a similar lag the calcium activity increased 1.6-fold. Control roots, which were decapped but measured before recovering gravisensitivity (19 h), showed no change in calcium activity. To test whether this gradient is necessary for gravitropic curvature, we eliminated the calcium activity gradient during gravitropism by applying a mobile calcium-binding site (dinitro-BAPTA; 1,2-bis(2-amino-5-nitro-phenoxy)ethane-N,N,N',N'-tetraacetic acid) to the root cap; this treatment eliminated gravicurvature. A calcium gradient may be formed by proton-induced calcium desorption if there is a proton gradient. Preventing the formation of apoplastic pH gradients, using 10 and 50 mM 2-(N-morpholino)ethanesulfonic acid (Mes) buffer or 10 mM fusicoccin to stimulate proton excretion maximally, did not inhibit curvature; therefore the calcium gradient is not a secondary effect of a proton gradient. We have found a distinct and rapid differential in the apoplasmic calcium activity between the upper and lower sides of gravistimulated maize root tips which is necessary for gravitropism.

NASA Discipline Number 40-50↗
NASA NTRS2005

Gradient Driven Fluctuations

We have worked with our collaborators at the University of Milan (Professor Marzio Giglio and his group-supported by ASI) to define the science required to measure gradient driven fluctuations in the microgravity environment. Such a study would provide an accurate test of the extent to which the theory of fluctuating hydrodynamics can be used to predict the properties of fluids maintained in a stressed, non-equilibrium state. As mentioned above, the results should also provide direct visual insight into the behavior of a variety of fluid systems containing gradients or interfaces, when placed in the microgravity environment. With support from the current grant, we have identified three key systems for detailed investigation. These three systems are: 1) A single-component fluid to be studied in the presence of a temperature gradient; 2) A mixture of two organic liquids to be studied both in the presence of a temperature gradient, which induces a steady-state concentration gradient, and with the temperature gradient removed, but while the concentration gradient is dying by means of diffusion; 3) Various pairs of liquids undergoing free diffusion, including a proteidbuffer solution and pairs of mixtures having different concentrations, to allow us to vary the differences in fluid properties in a controlled manner.

Cannell, David↗
NASA NTRS2011

Investigation of the Crust of the Pannonian Basin, Hungary Using Low-Altitude CHAMP Horizontal Gradient Magnetic Anomalies

The Pannonian Basin is a deep intra-continental basin that formed as part of the Alpine orogeny. It is some 600 by 500 km in area and centered on Hungary. This area was chosen since it has one of the thinnest continental crusts in Europe and is the region of complex tectonic structures. In order to study the nature of the crustal basement we used the long-wavelength magnetic anomalies acquired by the CHAMP satellite. The SWARM constellation, scheduled to be launched next year, will have two lower altitude satellites flying abreast, with a separation of between ca. 150 to 200 km. to record the horizontal magnetic gradient. Since the CHAMP satellite has been in orbit for eight years and has obtained an extensive range of data, both vertically and horizontally there is a large enough data base to compute the horizontal magnetic gradients over the Pannonian Basin region using these many CHAMP orbits. We recomputed a satellite magnetic anomaly map, using the spherical-cap method of Haines (1985), the technique of Alsdorf et al. (1994) and from spherical harmonic coefficients of MF6 (Maus et aI., 2008) employing the latest and lowest altitude CHAMP data. We then computed the horizontal magnetic anomaly gradients (Kis and Puszta, 2006) in order to determine how these component data will improve our interpretation and to preview what the SW ARM mission will reveal with reference to the horizontal gradient anomalies. The gradient amplitude of an 1000 km northeast-southwest profile through our horizontal component anomaly map varied from 0 to 0.025 nT/km with twin positive anomalies (0.025 and 0.023 nT/km) separated by a sharp anomaly negative at o nT/km. Horizontal gradient indicate major magnetization boundaries in the crust (Dole and Jordan, 1978 and Cordell and Grauch, 1985). Our gradient anomaly was modeled with a twodimensional body and the anomaly, of some 200 km, correlates with a 200 km area of crustal thinning in the southwestern Pannonian Basin.

Taylor, Patrick T.↗
DOE OSTI2026

Optimality of Gradient-MUSIC for Spectral Estimation

We introduce the Gradient-MUSIC algorithm for estimating the unknown frequencies and amplitudes of a nonharmonic signal from noisy time samples. While the classical MUSIC algorithm performs a computationally expensive search over a fine grid, Gradient-MUSIC is significantly more efficient and eliminates the need for discretization over a fine grid by using optimization techniques. It coarsely scans the 1D landscape to find initialization simultaneously for all frequencies followed by parallelizable local refinement via gradient descent. We also analyze its performance when the noise level is sufficiently small and the signal frequencies are separated by at least 8π/m, where π/m is the standard resolution of this problem. Even though the 1D landscape is nonconvex, we prove a global convergence result for Gradient-MUSIC: coarse scanning provably finds suitable initialization and gradient descent converges at a linear rate. In addition to convergence results, we also upper bound the error between the true signal frequencies and amplitudes with those found by Gradient-MUSIC. For example, if the noise has $\ell^\infty$ norm at most ϵ, then the frequencies and amplitudes are recovered up to error at most Cϵ/m and Cϵ respectively, which are minimax optimal in m and ϵ. Our theory can also handle stochastic noise with performance guarantees under nonstationary independent Gaussian noise. Our main approach is a comprehensive geometric analysis of the landscape, a perspective that has not been explored before.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Analytical gradient-based optimization of CALPHAD model parameters

The calibration of CALPHAD (CALculation of PHAse Diagrams) models involves the solution of a very challenging high-dimensional multiobjective optimization problem. Traditional approaches to parameter fitting predominantly rely on gradient-free methods, which while robust, are computationally inefficient and often scale poorly with model complexity. In this work, we introduce and demonstrate a generalizable framework for analytic gradient-based optimization of the parameters of the CALPHAD model enabled by the recently formalized Jansson derivative technique. This method allows for efficient evaluation of gradients of thermodynamic properties at equilibrium with respect to model parameters, even in the presence of arbitrarily complex internal degrees of freedom. Leveraging these semi-analytic gradients, we employ the conjugate gradient (CG) method to optimize thermodynamic model parameters for four binary alloy systems: Cu-Mg, Fe-Ni, Cr-Ni, and Cr-Fe. Across all systems, CG achieves comparable or superior optimality relative to Bayesian ensemble Markov Chain Monte Carlo (MCMC) with improvements in computational efficiency ranging from one to three orders of magnitude. Furthermore, our results establish a new paradigm for CALPHAD assessments in which high fidelity data-rich model calibration becomes tractable using deterministic gradient-informed algorithms.

CALPHAD↗
DOE OSTI2026

Integrated-photonics-based systems for polarization-gradient cooling of trapped ions

Trapped ions are a promising modality for quantum systems, with demonstrated utility as the basis for quantum processors and optical clocks. However, traditional trapped-ion systems are implemented using complex free-space optical configurations, whose large size and susceptibility to vibrations and drift inhibit scaling to large numbers of qubits. In recent years, integrated-photonics-based systems have been demonstrated as an avenue to address the challenge of scaling trapped-ion systems while maintaining high fidelities. While these previous demonstrations have implemented both Doppler and resolved-sideband cooling of trapped ions, these cooling techniques are fundamentally limited in efficiency. In contrast, polarization-gradient cooling can enable faster and more power-efficient cooling and, therefore, improved computational efficiencies in trapped-ion systems. While free-space implementations of polarization-gradient cooling have demonstrated advantages over other cooling mechanisms, polarization-gradient cooling has never previously been implemented using integrated photonics. In this paper, we design and experimentally demonstrate key polarization-diverse integrated-photonics devices and utilize them to implement a variety of integrated-photonics-based polarization-gradient-cooling systems, culminating in the first experimental demonstration of polarization-gradient cooling of a trapped ion by an integrated-photonics-based system. By demonstrating polarization-gradient cooling using an integrated-photonics-based system and, in general, opening up the field of polarization-diverse integrated-photonics-based devices and systems for trapped ions, this work facilitates new capabilities for integrated-photonics-based trapped-ion platforms.

atom optics↗
DOE OSTI2025

Vertical canopy gradients of respiration drive plant carbon budgets and leaf area index

Despite its importance for determining global carbon fluxes, leaf respiration remains poorly constrained in land surface models (LSMs). We tested the sensitivity of the Energy Exascale Earth System Model Land Model – Functionally Assembled Terrestrial Ecosystem Simulator (ELM-FATES) to variation in the canopy gradients of leaf maintenance respiration (R dark ). We ran global and point simulations varying the canopy gradient of R dark to explore the impacts on forest structure, composition, and carbon cycling. In global simulations, steeper canopy gradients of R dark lead to increased understory survival and leaf biomass. Leaf area index (LAI) increased up to 77% in tropical regions compared with the default parameterization, improving alignment with remotely sensed benchmarks. Global vegetation carbon varied from 308 Pg C to 449 Pg C across the ensemble. In tropical forest simulations, steeper gradients of R dark had a large impact on successional dynamics. Results show the importance of canopy gradients in leaf traits and fluxes for determining plant carbon budgets and emergent ecosystem properties such as competitive dynamics, LAI, and vegetation carbon. The high-model sensitivity to canopy gradients in R dark highlights the need for more observations of how leaf traits and fluxes vary along light micro-environments to inform critical dynamics in LSMs.

59 BASIC BIOLOGICAL SCIENCES↗
NASA NTRS1975

Cosmic ray intensity gradients in the solar system

Recent progress in the determination of cosmic-ray intensity gradients is reviewed. Direct satellite measurements of the integral gradient are described together with various types of indirect measurements, including measurements of the Ar-37/Ar-39 ratio in samples from the Lost City meteorite, studies of anisotropies in neutron-monitor counting rates, and analysis of the sidereal diurnal anisotropy observed at a single point on earth. Nucleonic radial gradients and electron gradients measured by satellites in differential energy windows are discussed, and theoretical studies of the physical processes involved in these gradients are summarized. Observations of intensity gradients in heliographic latitude are reported.

Mckibben, R. B.↗
NASA NTRS1976

Coronal temperatures and temperature gradients from OSO-7 spectroheliograms

Temperatures and temperature gradients for the outer corona are obtained from brightness gradients of EUV lines that were measured with the spectroheliograph on OSO-7. Brightness gradients show considerable deviations from isothermal model calculations that include collisional excitation and photoexcitation. A negative temperature gradient that gives both positive and negative ion-abundance gradients appears to be able to account for the discrepancy. For the 284-A of Fe XV, perhaps the strongest line from the outer corona, measurements during 1972 appear to be consistent with a temperature near 2.3 million K near the equator at about 1.3 solar radii from the solar center and with temperature-gradient values near -0.7 that extend from as low as 1.2 to about 1.8 solar radii. Temperatures from strong lines of Fe XIV and Fe XVI indicate that variations of about 200,000 K exist along lines of sight where emission is appreciable. There appears to be some agreement between these results and temperature measurements from ion abundances in the solar wind and the Doppler width of the 5303-A line.

Nakada, M. P.↗
NASA NTRS1978

Theoretical study of the effect of wind velocity gradients on longitudinal stability and control in climbing and level flight

A change in the wind vector over a short distance along the flight path (wind gradient) has caused several severe airplane accidents during take-off and landing. Results of a previous study showed that, in descending flight, a positive wind gradient (decreasing head wind) caused severe divergent motion and a negative wind gradient (decreasing tail wind) caused oscillatory motion which should not create a control problem. The results obtained when the same method of analysis was applied to climbing and to straight and level flight are reported. In straight and level flight, a wind gradient was found to cause effects similar to those found in descending flight. In climbing flight, it was found that a negative wind gradient caused a slightly divergent oscillation that presented no control problems and a positive wind gradient caused oscillatory divergence. Results of motion studies indicated that adequate control of the airplane motions can be provided by automatic control systems.

Sherman, W. L.↗
NASA NTRS1983

Latitudinal and field-aligned cosmic ray gradients 2 to 5 AU Voyagers 1 and 2 and IMP 8

Results are presented showing that the contemporary picture of a relatively small heliocentric radial gradient is generally correct. They emphasize that such a gradient should really be thought of in terms of field-aligned structures since the cosmic ray populations tend to 'corotate' locally with the interplanetary field structure. Nonuniform latitudinal gradients of approximately 2 to 5 percent/deg in structures lasting 10 to 30 days are identified unambiguously. Additional evidence is seen for somewhat smaller latitudinal gradients often north to south and probably mixed with small field-aligned gradients of less than 1 percent/AU that persist for several solar rotations. It is suggested that these smallest ephemeral latitudinal gradients manifest themselves by affecting the degree of irregularity in diurnal variations of high latitude neutron monitors.

Roelof, E. C.↗
NASA NTRS1989

Numerical optimization in Hilbert space using inexact function and gradient evaluations

Trust region algorithms provide a robust iterative technique for solving non-convex unstrained optimization problems, but in many instances it is prohibitively expensive to compute high accuracy function and gradient values for the method. Of particular interest are inverse and parameter estimation problems, since function and gradient evaluations involve numerically solving large systems of differential equations. A global convergence theory is presented for trust region algorithms in which neither function nor gradient values are known exactly. The theory is formulated in a Hilbert space setting so that it can be applied to variational problems as well as the finite dimensional problems normally seen in trust region literature. The conditions concerning allowable error are remarkably relaxed: relative errors in the gradient error condition is automatically satisfied if the error is orthogonal to the gradient approximation. A technique for estimating gradient error and improving the approximation is also presented.

Carter, Richard G.↗