SEARCH · Engineering Papers

Results for “Diffuse fraction”

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 541 records · Page 30

All sources NASA NTRS DOE OSTI

NASA NTRS1999

Soot Aerosol Properties in Laminar Soot-Emitting Microgravity Nonpremixed Flames

The spatial distributions and morphological properties of the soot aerosol are examined experimentally in a series of 0-g laminar gas-jet nonpremixed flames. The methodology deploys round jet diffusion flames of nitrogen-diluted acetylene fuel burning in quiescent air at atmospheric pressure. Full-field laser-light extinction is utilized to determine transient soot spatial distributions within the flames. Thermophoretic sampling is employed in conjunction with transmission electron microscopy to define soot microstructure within the soot-emitting 0-g flames. The microgravity tests indicate that the 0-g flames attain a quasi-steady state roughly 0.7 s after ignition, and sustain their annular structure even beyond their luminous flame tip. The measured peak soot volume fractions show a complex dependence on burner exit conditions, and decrease in a nonlinear fashion with decreasing characteristic flow residence times. Fuel preheat by approximately 140 K appears to accelerate the formation of soot near the flame axis via enhanced fuel pyrolysis rates. The increased soot presence caused by the elevated fuel injection temperatures triggers higher flame radiative losses, which may account for the premature suppression of soot growth observed along the annular region of preheated-fuel flames. Electron micrographs of soot aggregates collected in 0-g reveal the presence of soot precursor particles near the symmetry axis at midflame height, The observations also verify that soot primary particle sizes are nearly uniform among aggregates present at the same flame location, but vary considerably with radius at a fixed distance from the burner. The maximum primary size in 0-g is found to be by 40% larger than in 1-g, under the same burner exit conditions. Estimates of the number concentration of primary particles and surface area of soot particulate phase per unit volume of the combustion gases are also made for selected in-flame locations.

Konsur, Bogdan↗

NASA NTRS2004

Direct Electrolytic Deposition of Mats of Mn(x)O(y) Nanowires

Mats of free-standing manganese oxide (MnxOy) nanowires have been fabricated as experimental electrode materials for rechargeable electrochemical power cells and capacitors. Because they are free-standing, the wires in these mats are electrochemically accessible. The advantage of the mat-of-nanowires configuration, relative to other configurations of electrode materials, arises from the combination of narrowness and high areal number density of the wires. This combination offers both high surface areas for contact with electrolytes and short paths for diffusion of ions into and out of the electrodes, thereby making it possible to charge and discharge at rates higher than would otherwise be possible and, consequently, to achieve greater power densities. The nanowires are fabricated in an electrolytic process in which there is no need for an electrode binder material. Moreover, there is no need to incorporate an electrically conductive additive into the electrode material; the only electrically conductive material that must be added is a thin substrate contact film at the anchored ends of the nanowires. Hence, the mass fraction of active electrode material is close to 100 percent, as compared with about 85 percent in conventional electrodes made from a slurry of active electrode material, binder, and conductive additive pressed onto a metal foil. The locations and sizes of the nanowires are defined by holes in templates in the form of commercially available porous alumina membranes. In experiments to demonstrate the present process, alumina membranes of various pore sizes and degrees of porosity were used. First, a film of Au was sputtered onto one side of each membrane. The membranes were then attached, variously, to carbon tape or a gold substrate by use of silver or carbon paste. Once thus attached, the membranes were immersed in a plating solution comprising 0.01 M MnSO4 + 0.03 M (NH4)2SO4. The pH of the solution was kept constant at 8 by addition of H2SO4 or NH4OH as needed. Mn(x)O(y) nanowires were potentiostatically electrodeposited in the pores in the alumina templates. Depending on the anodic deposition potentials, Mn(x)O(y) was deposited in various oxidation states [divalent (Mn3O4), trivalent (Mn2O3), or tetravalent (MnO2)]. The Mn(x)O(y) wires were made free-standing (see figure) by dissolving the alumina templates, variously, in KOH or NaOH at a concentration of 20 volume percent.

Myung, Nosang↗

NASA NTRS2021

Extragalactic Magnetism with SOFIA (Legacy Program). I. The Magnetic Field in the Multiphase Interstellar Medium of M51*

The recent availability of high-resolution far-infrared (FIR) polarization observations of galaxies using HAWC+/SOFIA has facilitated studies of extragalactic magnetic fields in the cold and dense molecular disks. We investigate whether any significant structural differences are detectable in the kiloparsec-scale magnetic field of the grand design face-on spiral galaxy M51 when traced within the diffuse (radio) and the dense and cold (FIR) interstellar medium (ISM). Our analysis reveals a complex scenario where radio and FIR polarization observations do not necessarily trace the same magnetic field structure. We find that the magnetic field in the arms is wrapped tighter at 154 μm than at 3 and 6 cm; statistically significant lower values for the magnetic pitch angle are measured at FIR in the outskirts (R ≥ 7 kpc) of the galaxy. This difference is not detected in the interarm region. We find strong correlations of the polarization fraction and total intensity at FIR and radio with the gas column density and 12CO(1–0) velocity dispersion. We conclude that the arms show a relative increase of small-scale turbulent B-fields at regions with increasing column density and dispersion velocities of the molecular gas. No correlations are found with H i neutral gas. The star formation rate shows a clear correlation with the radio polarized intensity, which is not found in FIR, pointing to a small-scale dynamo-driven B-field amplification scenario. This work shows that multiwavelength polarization observations are key to disentangling the interlocked relation between star formation, magnetic fields, and gas kinematics in the multiphase ISM.

Spiral galaxies↗

NASA NTRS1984

An X-ray survey of clusters of galaxies. IV - A survey of southern clusters and a compilation of upper limits for both Abell and southern clusters

The results of the HEAO 1 A-1 X-ray survey of galaxy clusters are reported. X-ray error boxes and intensities are presented for all clusters in the Abell catalog and for the catalog of southern clusters and groups compiled by Duus and Newell (1977). A correlation is derived on the basis of the X-ray luminosity function for 2-6 keV which may be used to calculate the contribution of clusters to the diffuse X-ray background at different energies. The cluster X-ray is estimated to be 9.3 percent (+ 1.9 or - 1.5 percent). Correlations between X-ray luminosity and other cluster properties are exmained, and it is found that the distribution of upper limits may be applied to obtaining a more precise estimate of the average X-ray luminosity of clusters. The Abell richness class and southern cluster concentrations were strongly correlated with X-ray luminosity. Correlations between optical x-ray luminosity and optical radius velocity dispersion, spiral fraction, and radio power are analyzed. The evidence for all these correlations was considered to be weak because of poor scatter in the data.

Kowalski, M. P.↗

NASA NTRS2000

Design and Performance of McRas in SCMs and GEOS I/II GCMs

The design of a prognostic cloud scheme named McRAS (Microphysics of clouds with Relaxed Arakawa-Schubert Scheme) for general circulation models (GCMs) will be discussed. McRAS distinguishes three types of clouds: (1) convective, (2) stratiform, and (3) boundary-layer types. The convective clouds transform and merge into stratiform clouds on an hourly time-scale, while the boundary-layer clouds merge into the stratiform clouds instantly. The cloud condensate converts into precipitation following the auto-conversion equations of Sundqvist that contain a parametric adaptation for the Bergeron-Findeisen process of ice crystal growth and collection of cloud condensate by precipitation. All clouds convect, advect, as well as diffuse both horizontally and vertically with a fully interactive cloud-microphysics throughout the life-cycle of the cloud, while the optical properties of clouds are derived from the statistical distribution of hydrometeors and idealized cloud geometry. An evaluation of McRAS in a single column model (SCM) with the GATE Phase III and 5-ARN CART datasets has shown that together with the rest of the model physics, McRAS can simulate the observed temperature, humidity, and precipitation without many systematic errors. The time history and time mean incloud water and ice distribution, fractional cloudiness, cloud optical thickness, origin of precipitation in the convective anvil and towers, and the convective updraft and downdraft velocities and mass fluxes all show a realistic behavior. Performance of McRAS in GEOS 11 GCM shows several satisfactory features but some of the remaining deficiencies suggest need for additional research involving convective triggers and inhibitors, provision for continuously detraining updraft, a realistic scheme for cumulus gravity wave drag, and refinements to physical conditions for ascertaining cloud detrainment level.

Sud, Yogesh C.↗

NASA NTRS1990

Coarsening in binary solid-liquid mixtures

A theory of Ostwald ripening has been developed for a solid-liquid mixture cosisting of a low volume fraction array of spherical solid particles in a liquid wherein the coarsening process proceeds via the transport of both heat and mass. It is found that the simultaneous transport of heat and mass during ripening does not alter the exponents of the temporal power laws governing the ripening process from their classical values but does alter the amplitudes of these power laws. The growth rate of the cube of the average particle radius, the rate constant, is found to depend both on the alloy solute concentration and the ratio of the thermal to solutal diffusivities. In most metallic systems, a large decrease in the rate constant can be expected with small additions of solute to a pure metal. Possible extensions of this theory to the analogous problem of ripening in isothermal ternary alloys are also discussed.

Voorhees, P. W.↗

NASA NTRS2007

Runtime and Pressurization Analyses of Propellant Tanks

Multi-element unstructured CFD has been utilized at NASA SSC to carry out analyses of propellant tank systems in different modes of operation. The three regimes of interest at SSC include (a) tank chill down (b) tank pressurization and (c) runtime propellant draw-down and purge. While tank chill down is an important event that is best addressed with long time-scale heat transfer calculations, CFD can play a critical role in the tank pressurization and runtime modes of operation. In these situations, problems with contamination of the propellant by inclusion of the pressurant gas from the ullage causes a deterioration of the quality of the propellant delivered to the test article. CFD can be used to help quantify the mixing and propellant degradation. During tank pressurization under some circumstances, rapid mixing of relatively warm pressurant gas with cryogenic propellant can lead to rapid densification of the gas and loss of pressure in the tank. This phenomenon can cause serious problems during testing because of the resulting decrease in propellant flow rate. With proper physical models implemented, CFD can model the coupling between the propellant and pressurant including heat transfer and phase change effects and accurately capture the complex physics in the evolving flowfields. This holds the promise of allowing the specification of operational conditions and procedures that could minimize the undesirable mixing and heat transfer inherent in propellant tank operation. It should be noted that traditional CFD modeling is inadequate for such simulations because the fluids in the tank are in a range of different sub-critical and supercritical states and elaborate phase change and mixing rules have to be developed to accurately model the interaction between the ullage gas and the propellant. We show a typical run-time simulation of a spherical propellant tank, containing RP-1 in this case, being pressurized with room-temperature nitrogen at 540 R. Nitrogen, shown in blue on the right-hand side of the figures, enters the tank from the diffuser at the top of the figures and impinges on the RP-1, shown in red, while the propellant is being continuously drained at the rate of 1050 lbs/sec through a pipe at the bottom of the tank. The sequence of frames in Figure 1 shows the resultant velocity fields and mixing between nitrogen and RP-1 in a cross-section of the tank at different times. A vortex is seen to form in the incoming nitrogen stream that tends to entrain propellant, mixing it with the pressurant gas. The RP-1 mass fraction contours in Figure 1 are also indicative of the level of mixing and contamination of the propellant. The simulation is used to track the propagation of the pure propellant front as it is drawn toward the exit with the evolution of the mixing processes in the tank. The CFD simulation modeled a total of 10 seconds of run time. As is seen from Figure 1d, after 5.65 seconds the propellant front is nearing the drain pipe, especially near the center of the tank. Behind this pure propellant front is a mixed fluid of compromised quality that would require the test to end when it reaches the exit pipe. Such unsteady simulations provide an estimate of the time that a high-quality propellant supply to the test article can be guaranteed at the modeled mass flow rate. In the final paper, we will discuss simulations of the LOX and propellant tanks at NASA SSC being pressurized by an inert ullage. Detailed comparisons will be made between the CFD simulations and lower order models as well as with test data. Conditions leading to cryo collapse in the tank will also be identified.

Field, Robert E.↗

NASA NTRS1983

An analytical comparison of two wing structures for Mach 5 cruise airplanes

Mach 5 cruise research conducted by NASA is related to aerodynamics, propulsion, and structures. The study of structures includes the propulsion system, fuselage, and wings. Various studies have shown that the achievement of adequate range is largely dependent on a low structural mass fraction. The present investigation is concerned with a study of two wing structure configurations for Mach 5 aircraft. An uprated version (Ti-6242 replacing B-120 titanium) of the YF-12/SR-71 wing structure is considered. The B-120 titanium structure represents the current art of high speed aircraft wing structures. The YF-12 wing structure was designed about 20 years ago when the analytical methods for calculating thermal stresses were limited. The second wing structural configuration studied in the present investigation also used Ti-6242 materials but replaced the corrugated-beaded panels with diffusion bonded honeycomb-core sandwich panels, and replaced the z-stiffened shear webs with sine-wave stiffened shear webs.

Taylor, A. H.↗

NASA NTRS1999

Buoyant Low Stretch Diffusion Flames Beneath Cylindrical PMMA Samples

A unique new way to study low gravity flames in normal gravity has been developed. To study flame structure and extinction characteristics in low stretch environments, a normal gravity low-stretch diffusion flame is generated using a cylindrical PMMA sample of varying large radii. Burning rates, visible flame thickness, visible flame standoff distance, temperature profiles in the solid and gas, and radiative loss from the system were measured. A transition from the blowoff side of the flammability map to the quenching side of the flammability map is observed at approximately 6-7/ sec, as determined by curvefits to the non-monotonic trends in peak temperatures, solid and gas-phase temperature gradients, and non-dimensional standoff distances. A surface energy balance reveals that the fraction of heat transfer from the flame that is lost to in-depth conduction and surface radiation increases with decreasing stretch until quenching extinction is observed. This is primarily due to decreased heat transfer from the flame, while the magnitude of the losses remains the same. A unique local extinction flamelet phenomena and associated pre-extinction oscillations are observed at very low stretch. An ultimate quenching extinction limit is found at low stretch with sufficiently high induced heat losses.

Olson, S. L.↗

NASA NTRS1986

Opposed jet burner studies of silane-methane, silane-hydrogen and hydrogen diffusion flames with air

An atmospheric pressure tubular opposed jet burner technique was used to characterize certain diffusion-flame transitions and associated burning rates for N2-diluted mixtures of highly-reactive fuels. Presented are: (1) details of the technique, with emphasis on features permitting the study of flames involving pyrophoric gases and particle-forming combustion reactions: (2) discoveries on the properties of these flames which correspond to physically and chemically distinct stages of silane and hydrogen combustion; and (3) unburnt gas velocity data obtained from flames based on SiH4-CH4-N2, SiH4-H2-N2, and H2-N2 fuel mixtures, and plotted as functions of combustible-fuel mole fraction and fuel/oxygen molar input flow ratios. In addition, these burning velocity results are analyzed and interpreted.

Pellett, G. L.↗

NASA NTRS1992

Is a pyrene-like molecular ion the cause of the 4,430-A diffuse interstellar absorption band?

The diffuse interstellar band (DIB) absorption features of astronomical spectra are suggested by recent results to be separable from the grains that cause visual extinction. Attention is presently given to laboratory measurements of the optical spectrum of the pyrene cation C16H10(+), which is one of the polycyclic aromatic hydrocarbon (PAH) molecular candidates proposed as carriers for DIBs. This ion exhibits an intense but strangely broad continuum similar to that of the naphthalene cation, so that this may be a common feature of all PAH cations and the basis of an explanation for PAHs' converting of an interstellar radiation fraction as large as that from the UV and visible range down to the IR.

Salama, F.↗

NASA NTRS1993

Evaporation in equilibrium, in vacuum, and in hydrogen gas

Evaporation experiments were conducted for SiO2 in three different conditions: in equilibrium, in vacuum, and in hydrogen gas. Evaporation rate in vacuum is about two orders of magnitude smaller than that in equilibrium, which is consistent with previous works. The rate in hydrogen gas changes depending on hydrogen pressure. The rate at 10 exp -7 bar of hydrogen pressure is as small as that of free evaporation, but at 10 exp -5 bar of hydrogen pressure it is larger than that in equilibrium. In equilibrium and in vacuum, the evaporation rate is limited by decomposition of SiO2 on the crystal surface, but it is limited by a diffusion process for evaporation in hydrogen gas. Therefore, evaporation rate of minerals in the solar nebula can be shown neither by that in equilibrium nor by that in vacuum. The maximum temperature of the solar nebula at the midplane at 2-3 AU where chondrites are believed to have originated is calculated to be as low as 150 K, 1500 K, or in between them. The temperature is, in any case, not high enough for total evaporation of the interstellar materials. Therefore, evaporation of interstellar materials is one of the most important processes for the origin and fractionation of solid materials. The fundamental process of evaporation of minerals has been intensively studied for these several years. Those experiments were carried out either in equilibrium or in vacuum; however, evaporation in the solar nebula is in hydrogen (and much smaller amount of helium) gas. In order to investigate evaporation rate and compositional (including isotopic) fractionation during evaporation, vaporization experiments for various minerals in various conditions are conducted. At first, SiO2 was adopted for a starting material, because thermochemical data and its nature of congruent vaporization are well known. Experiments were carried out in a vacuum furnace system.

Nagahara, Hiroko↗

NASA NTRS2015

Mantle Water Contents Beneath the Rio Grande Rift (NM, USA): FTIR Analysis of Rio Puerco and Kilbourne Hole Peridotite Xenoliths

Peridotite xenoliths from the Rio Grande Rift (RGR) are being analyzed for H (sub 2) O contents by FTIR (Fourier Transform Infrared) as well as for major and trace element compositions. Nine samples are from the Rio Puerco Volcanic Field (RP) which overlaps the central RGR and southeastern Colorado Plateau; seventeen samples are from Kilbourne Hole (KH) in the southern RGR. Spinel Cr# (Cr/(Cr+Al)) (0.08-0.46) and olivine Mg# (Mg/(Mg plus Fe)) (0.883-0.911) of all RGR samples fall within the olivine-spinel mantle array from [1], an indicator that peridotites are residues of partial melting. Pyroxene H (sub 2) O in KH correlate with bulk rock and pyroxene Al (sub 2) O (sub 3).The KH clinopyroxene rare earth element (REE) variations fit models of 0-13 percent fractional melting of a primitive upper mantle. Most KH peridotites have bulk-rock light REE depleted patterns, but five are enriched in light REEs consistent with metasomatism. Variation in H (sub 2) O content is unrelated to REE enrichment. Metasomatism is seen in RP pyroxenite xenoliths [2] and will be examined in the peridotites studied here. Olivine H (sub 2) O contents are low (less than or equal to 15 parts per million), and decrease from core to rim within grains. This is likely due to H loss during xenolith transport by the host magma [3]. Diffusion models of H suggest that mantle H (sub 2) O contents are still preserved in cores of KH olivine, but not RP olivine. The average H (sub 2) O content of Colorado Plateau clinopyroxene (670 parts per million) [4] is approximately 300 parts per million higher than RGR clinopyroxene (350 parts per million). This upholds the hypothesis that hydration-induced lithospheric melting occurred during flat-slab subduction of the Farallon plate [5]. Numerical models indicate hydration via slab fluids is possible beneath the plateau, approximately 600 kilometers from the paleo-trench, but less likely approximately 850 kilometers away beneath the rift [6].

Schaffer, L. A.↗

NASA NTRS1986

Opposed jet burner studies of silane-methane, silane-hydrogen, and hydrogen diffusion flames with air

An atmospheric pressure tubular opposed jet burner technique was used to characterize certain diffusion-flame transitions and associated burning rates for N2-diluted mixtures of highly-reactive fuels. The paper presents: (1) details of the technique, with emphasis on features permitting the study of flames involving pyrophoric gases and particle-forming combustion reactions; (2) discoveries on the properties of these flames which correspond to physically and chemically distinct stages of silane and hydrogen combustion; and (3) unburnt gas velocity data obtained from flames based on SiH4-CH4-N2, SiH4-H2-N2, and H2-N2 fuel mixtures, and plotted as functions of combustible-fuel mole fraction and fuel/oxygen molar input flow ratios. In addition, these burning velocity results are analyzed and interpreted.

Pellett, G. L.↗

NASA NTRS1995

ROSAT detection of diffuse hot gas in the edge-on galaxy NGC 4631

We present our ROSAT observation of the edge-on spiral galaxy NGC 4631, a nearby Sc/SBd galaxy best known for its extended radio halo. Because of the low foreground Galactic X-ray-absorbing gas column density, N(sub HI) approximately = 1.4 x 10(exp 20)/sq cm, this observation is sensitive to gas of temperature greater than or approximately = a few x 10(exp 5) K. We detected soft (approximately 0.25 keV) X-ray radiation out to more than 8 kpc above the midplane of the galaxy. The stongest X-ray emission in the halo is above the central disk, a region of about 3 kpc radius which shows high star formation activity. The X-ray emission in the halo is bordered by two extended filaments of radio continuum emission. We also found diffuse X-ray emission from hot gas in the galaxy's disk. The spectrum of the radiation can be characterized by a thermal plasma with a temperature of 3 x 10(exp 6) K and a radiative cooling rate of approximately 8 x 10(exp 39) ergs/sec. This rate is only a few percent of the estimated supernova energy release in the interstellar medium of the galaxy. Analysis of the X-ray spectrum shows evidence for the presence of a cooler (several times 10(exp 5) K) halo gas component that could consume a much larger fraction of the supernova energy. We found strong evidence for disk/halo interaction. Hot gas apparently blows out from supershells in the galaxy's disk at a rate of approximately 1 solar mass/yr. This outflow of hot gas drags magnetic field lines up in the halo and forms a magnetized gaseous halo. If the magnetic field lines are still anchored to the disk gas at large disk radii, the outflowing gas may be confined high above the disk by magnetic pressure. We have identified a strong X-ray source which coincides spatially with an H I supershell. However, the source is likely an extremely luminous X-ray binary with L(sub x)(0.1-2 keV) approximately = 5 x 10(exp 39) ergs/sec, which makes it a stellar mass black hole candidate.

Wang, Q. Daniel↗

NASA NTRS2005

Numerical Analysis of Flow Evolution in a Helium Jet Injected into Ambient Air

A computational model to study the stability characteristics of an evolving buoyant helium gas jet in ambient air environment is presented. Numerical formulation incorporates a segregated approach to solve for the transport equations of helium mass fraction coupled with the conservation equations of mixture mass and momentum using a staggered grid method. The operating parameters correspond to the Reynolds number varying from 30 to 300 to demarcate the flow dynamics in oscillating and non-oscillating regimes. Computed velocity and concentration fields were used to analyze the flow structure in the evolving jet. For Re=300 case, results showed that an instability mode that sets in during the evolution process in Earth gravity is absent in zero gravity, signifying the importance of buoyancy. Though buoyancy initiates the instability, below a certain jet exit velocity, diffusion dominates the entrainment process to make the jet non-oscillatory as observed for the Re=30 case. Initiation of the instability was found to be dependent on the interaction of buoyancy and momentum forces along the jet shear layer.

Satti, Rajani P.↗

NASA NTRS1979

Analysis of turbulent free-jet hydrogen-air diffusion flames with finite chemical reaction rates

A numerical analysis is presented of the nonequilibrium flow field resulting from the turbulent mixing and combustion of an axisymmetric hydrogen jet in a supersonic parallel ambient air stream. The effective turbulent transport properties are determined by means of a two-equation model of turbulence. The finite-rate chemistry model considers eight elementary reactions among six chemical species: H, O, H2O, OH, O2 and H2. The governing set of nonlinear partial differential equations was solved by using an implicit finite-difference procedure. Radial distributions were obtained at two downstream locations for some important variables affecting the flow development, such as the turbulent kinetic energy and its dissipation rate. The results show that these variables attain their peak values on the axis of symmetry. The computed distribution of velocity, temperature, and mass fractions of the chemical species gives a complete description of the flow field. The numerical predictions were compared with two sets of experimental data. Good qualitative agreement was obtained.

Sislian, J. P.↗

NASA NTRS1986

A brief description of the simple biosphere model (SiB)

A biosphere model for calculating the transfer of energy, mass, and momentum between the atmosphere and the vegetated surface of the Earth was designed for atmospheric general circulation models. An upper vegetation layer represents the perennial canopy of trees or shrubs, a lower layer represents the annual ground cover of grasses and other herbacious species. The local coverage of each vegetation layer may be fractional or complete but as the individual vegetation elements are considered to be evenly spaced, their root systems are assumed to extend uniformly throughout the entire grid-area. The biosphere has seven prognostic physical-state variables: two temperatures (one for the canopy and one for the ground cover and soil surface); two interception water stores (one for the canopy and one for the ground cover); and three soil moisture stores (two of which can be reached by the vegetation root systems and one underlying recharge layer into and out of which moisture is transferred only by hydraulic diffusion).

Sellers, P. J.↗