Search

Deep learning architectures based on convolutional neural networks tend to rely on continuous, smooth features. While this characteristics provides significant robustness and proves useful in many real-world tasks, it is strikingly incompatible with the physical characteristic of the world, which, at the scale in which humans operate, comprises crisp objects, typically representing well-defined categories. This study proposes a class of neurosymbolic systems that learn by reconstructing the observed images in terms of visual primitives and are thus forced to form high-level, structural explanations of them. When applied to the task of diagnosing abnormalities in histological imaging, the method proved superior to a conventional deep learning architecture in terms of classification accuracy, while being more transparent., Comment: 16 pages, 9 figures

We investigated the size dependence of the ground state energy in self-assembled InAs quantum dots embedded in resonant tunneling diodes. Individual current steps observed in the current-voltage characteristics are attributed to resonant single-electron tunneling via the ground state of individual InAs quantum dots. The onset voltage of the first step observed is shown to decrease systematically from 200 mV to 0 with increasing InAs coverage. We relate this to a coverage-dependent size of InAs dots grown on AlAs. The results are confirmed by atomic force micrographs and photoluminescence experiments on reference samples., Comment: 3 pages, 3 figures

We investigate the noise properties of a GaAs-AlAs-GaAs tunneling structure with embedded self-assembled InAs quantum dots in the single-electron tunneling regime. We analyze the dependence of the relative noise amplitude of the shot noise on bias voltage. We observe a non-monotonic behaviour of the Fano-factor $\alpha$ with an average value of $\alpha \approx 0.8$ consistent with the asymmetry of the tunneling barriers. Reproducible fluctuations observed in $\alpha$ can be attributed to the successive participation of more and more InAs quantum dots in the tunneling current.

We have determined the Land\'e factor g* in self-organized InAs quantum dots using resonant-tunnelling experiments. With the magnetic field applied parallel to the growth direction z we find g*_\parallel = 0.75 for the specific dot investigated. When the magnetic field is tilted away by theta from the growth axis, g* gradually increases up to a value g*_\perp = 0.92 when B \perp z. Its angular dependence is found to follow the phenomenological behaviour g* (theta) = \sqrt{(g*_\parallel cos(theta)^2 + (g*_\perp sin(theta)^2}., Comment: 4 pages, 3 figures, International Conference on Semiconductor Quantum Dots July 31 - August 3, 2000, Munich, Germany

Current steps attributed to resonant tunneling through individual InAs quantum dots embedded in a GaAs-AlAs-GaAs tunneling device are investigated experimentally in magnetic fields up to 28 T. The steps evolve into strongly enhanced current peaks in high fields. This can be understood as a field-induced Fermi-edge singularity due to the Coulomb interaction between the tunneling electron on the quantum dot and the partly spin polarized Fermi sea in the Landau quantized three-dimensional emitter., Comment: 5 pages, 4 figures

A short overview of the QCD program at LEP~2 is given. Studies of final state interactions in e+e- -> WW -> qqqq decays are discussed., Comment: 6 pages, 2 figures, to be published in the proceedings of the 34th Rencontres de Moriond, Les Arrcs 1800, France, 13-20 March 1999

We study surface effects in amorphous polymer systems by means of computer simulation. In the framework of molecular dynamics, we present two different methods to prepare such surfaces. {\em Free} surfaces are stabilized solely by van--der--Waals interactions whereas {\em confined} surfaces emerge in the presence of repelling plates. The two models are compared in various computer simulations. For free surfaces, we analyze the migration of end--monomers to the surface. The buildup of density and pressure profiles from zero to their bulk values depends on the surface preparation method. In the case of confined surfaces, we find density and pressure oszillations next to the repelling plates. We investigate the influence of surfaces on the coordination number, on the orientation of single bonds, and on polymer end--to--end vectors. Furthermore, different statistical methods to determine location and width of the surface region for systems of various chain lengths are discussed and applied. We introduce a ``height function'' and show that this method allows to determine average surface profiles only by scanning the outermost layer of monomers., Comment: 23 pages, 12 figures

We study the deformation of nano--scale polymer films which are subject to external bending forces by means of computer simulation. The polymer is represented by a generalized bead--spring--model, intended to reproduce characteristic features of n--alkanes. The film is loaded by the action of a prismatic blade which is pressed into the polymer bulk from above and a pair of columns which support the film from below. The interaction between blade and support columns and the polymer is modelled by the repulsive part of a Lennard-Jones potential. For different system sizes as well as for different chainlengths, this nano--scale experiment is simulated by molecular dynamics methods. Our results allow us to give a first characterization of deformed states for such films. We resolve the kinetic and the dynamic stage of the deformation process in time and access the length scale between discrete particle and continuum mechanics behaviour. For the chainlengths considered here, we find that the deformation process is dominated by shear. We observe strangling effects for the film and deformation fluctuations in the steady state., Comment: 15 pages, 8 figures

Space weathering is the cumulative effect of physical and chemical changes that occur to substances exposed on the exterior of body void of an atmosphere [1], in this case the regolith on asteroid Eros. It is only recently that the scientific community has accepted the theory first developed in the mid- 1970s by Hapke and his colleagues of how space weathering occurs. The theory [1] asserts that optical and magnetic effects, first studied on moon rocks and lunar regolith, are caused by submicroscopic metallic iron (SMFe), smaller than the wavelength of light in vapor deposit coatings, on regolith grains, and in agglutinates. This vapor is generated by solar wind and micrometeorite impacts and does not require additional heating, melting, or a reducing environment to produce space weathering. One of the major finds of the first detailed reconnaissance of an asteroid by the NEAR Shoemaker mission was that the surface of Eros was essentially chondritic yet showed major depletions in sulfur [2, 3]. Here we propose space weathering sputtering experiments that may contribute to the explanation of sulfur depletion on asteroid Eros.

We present preliminary evidence that suggests a hemispheric albedo dichotomy on Titan, the largest planetary satellite in the Solar System. We have also studied the photometric properties of several dark circular features on Titan's surface to test if they might be of impact origin. The evidence is derived from photometric analysis of selected surface regions taken at different Titanian longitudes and solar phase angles using images from the Cassini Saturn Orbiter Visual and Infrared Mapping Spectrometer (VIMS). The VIMS instrument is able to image Titan's surface at spectral windows (e.g. 2.02 microns) in its atmosphere where methane, the principal atmospheric absorber is transparent. Additional information is included in the original extended abstract.

The Opposition Effect (OE) is the non-linear increase in the intensity of light scattered from a surface as phase angle approaches 0 deg. It is seen in laboratory experiments and in remote sensing observations of planetary surfaces. Understanding the OE is a requirement for fitting photometric models which produce meaningful results about regolith texture. Previously we have reported measurements from the JPL long arm goniometer and we have shown that this instrument enables us to distinguish between two distinct processes which create the opposition surges, Shadow Hiding (SHOE) and Coherent Backscattering (CBOE). SHOE arises because, as phase angle approaches zero, shadows cast by regolith grains on other grains become invisible to the observer. CBOE results from constructive interference between rays traveling the same path but in opposite directions. Additional information is included in the original extended abstract.

The Opposition Effect (OE) is the non-linear increase in the intensity of light scattered from a surface as phase angle approaches 0 deg. It is seen in laboratory experiments and in remote sensing observations of planetary surfaces. Understanding the OE is a requirement to fitting photometric models which will produce meaningful results about regolith texture. Our previous laboratory studies are consistent with the hypothesis that the OE in particulate materials is due to two processes, Shadow Hiding (SHOE) and Coherent Backscattering (CBOE). SHOE arises because, as phase angle approaches zero, shadows cast by regolith grains on other grains become invisible to the observer. CBOE results from constructive interference between rays traveling the same path but in opposite directions. In this study we measured the angular scattering properties of 9 mixtures of Aluminum Oxide and Boron Carbide powders of the same particle diameter (25 microns). The reflectance of the materials ranged from 7% (pure B4C) to 91% (pure Al2O3). Along with the reflectance phase curve we measured the circular polarization ratio (CPR) - the ratio of the intensity of the light returned with the same helicity as the incident light to that with the opposite helicity. An increase in CPR with decreasing phase angle indicates increased multiple scattering and is consistent with CBOE (Hapke, 1993). Popular conceptions of CBOE (Belskaya et al, 2003) hold that materials of higher albedo would exhibit increased multiple scattering and that the contribution of CBOE to the OE would increase as albedo increases. Remarkably, we find the highest albedo samples did not have the strongest CBOE opposition peaks. Instead, the maximum CBOE contribution is observed in samples with reflectance between 15 and 40%.

The Opposition Effect, the pronounced non-linear intensity increase in the reflectance phase curve with decreasing phase angle,theta , has long been observed in solar system bodies and in laboratory investigations of the angular scattering properties of particulate media. The size and shape of the phase curve, and the change in linear polarization with theta, have been related to the physical properties of planetary regolith scattering materials. Near zero degrees the increase in reflectance with decreasing phase angle has been attributed to two distinct processes. The first is the elimination of shadows cast between the regolith grains as the phase angle decreases. This is called the shadow hiding opposition effect (SHOE). The second is coherent constructive interference between rays of light traveling along identical but opposite paths in multiply scattering media. This is called the coherent backscattering opposition effect (CBOE).

The nature of the scattering of light is thought to be well understood when the medium is made up of independent scatterers that are much larger than the wavelength of that light. This is not the case when the size of the scattering objects is similar to or smaller than the wavelength or the scatterers are not independent. In an attempt to examine the applicability of independent particle scattering models, to planetary regoliths, a dataset of experimental results were compared with theoretical predictions.

We have studied the opposition surge of well sorted diamond powders. All particle sizes show a pronounced opposition surge and evidence for coherent backscatter, in disagreement with theoretical predictions. Additional information is contained in the original extended abstract.

Samples of well-sorted materials have been studied to determine how their scattering properties change with changing particle size. Additional information is contained in the original extended abstract.

Laboratory measurements of the reflectance phase curve, a measurement of great importance in understanding the texture of planetary regoliths, are not consistent with widely accepted theoretical models. Additional information is contained in the original extended abstract.

The continuity and timely implementation of the Mars exploration strategy relies heavily on the ability of the planetary community to interpret infrared spectral data. However, the increasing mission rate, data volume, and data variety, combined with the small number of spectroscopists within the planetary community, will require a coordinated community effort for effective and timely interpretation of the newly acquired and planned data sets. Relevant spectroscopic instruments include the 1996 TES, 2001 THEMIS, 2003 Pancam, 2003 Mini-TES, 2003 Mars Express OMEGA, 2003 Mars Express PFS, and 2005 CFUSM. In light of that, leaders of the Mars spectral community met June 4-6 to address the question: What terrestrial theoretical, laboratory, and field studies are most needed to best support timely interpretations of current and planned visible infrared spectrometer data sets, in light of the Mars Program goals? A primary goal of the spectral community is to provide a reservoir of information to enhance and expand the exploration of Mars. Spectroscopy has a long history of providing the fundamental compositional discoveries in the solar system, from atmospheric constituents to surface mineralogy, from earth-based to spacecraft-based observations. However, such spectroscopic compositional discoveries, especially surface mineralogies, have usually come after long periods of detailed integration of remote observations, laboratory analyses, and field measurements. Spectroscopic information of surfaces is particularly complex and often is confounded by interference of broad, overlapping absorption features as well as confusing issues of mixtures, coatings, and grain size effects. Thus some spectroscopic compositional discoveries have come only after many years of research. However, we are entering an era of Mars exploration with missions carrying sophisticated spectrometers launching about every 2 years. It is critical that each mission provide answers to relevant questions to optimize the success of the next mission. That will not occur effectively unless the spectroscopic remote sensing data can be processed and understood on an approximate 2-year rate. Our current knowledge of spectral properties of materials and confounding effects of the natural environment are note well enough understood for the accurate interpretations needed for such ambitious and time critical exploration objectives. This workshop focused on identify ing critical gaps in moving the field towards the goal of rapid and accurate analysis and interpretation.

Particles in a planetary regolith can be treated as independent scatterers when their size is large compared with the wavelength of light. The nature of this scattering, however, is poorly understood as the particle size approaches and becomes smaller than the wavelength. In order to understand this interaction better, the reflectances of well-sorted particulate samples of known composition were measured. This reflectance data was then analyzed to determine how scattering properties change with particle size.

The thermal region of the spectrum is one of special interest in planetary science as it is the only region where planetary emission is significant. Studying how planetary surfaces emit in the thermal infrared can tell us about their physical makeup and chemical composition, as well as their temperature profile with depth. This abstract will discuss a model of thermal energy transfer in planetary regoliths on airless bodies which includes both conductive and radiative processes while including the time dependence of the solar input function.

The laboratory investigation of the scattering properties of candidate planetary regolith materials is an important technique for understanding the physical properties of a planetary regolith. Additional information is contained in the original extended abstract.

UNKNOWN

We have searched for wavelength dependent effects in phase curves of candidate planetary regolith materials where coherent backscattering contributes to the opposition phase curve.

The thermal regime in planetary regoliths involves three processes: propagation of visible radiation, propagation of thermal radiation, and thermal conduction. The equations of radiative transfer and heat conduction are formulated for particulate media composed of anisotropically scattering particles. Although the equations are time dependent, only steady state problems are considered in this paper. Using the two-stream approximation, solutions are obtained for two cases: a layer of powder heated from below and an infinitely thick regolith illuminated by visible radiation. Radiative conductivity, subsurface temperature gradients, and the solid state greenhouse effect all appear intrinsically in the solutions without ad hoc additions. Although the equations are nonlinear, approximate analytic solutions that are accurate to a few percent are obtained. Analytic expressions are given for the temperature distribution, the optical and thermal radiance distributions, the hemispherical albedo, the hemispherical emissivity, and the directional emissivity. Additional applications of the new model to three problems of interest in planetary regoliths are presented by Hapke.

Several problems of interest in planetary infrared remote sensing are investigated using a new radiative-conductive model of energy transfer in regoliths: the solid-state greenhouse effect, thermal beaming, and reststrahlen spectra. The results of the analysis are as follows: (1) The solid-state greenhouse effect is self-limiting to a rise of a few tens of degrees in bodies of the outer solar system. (2) Non-Lambertian directional emissivity can account for only about 20% of the observed thermal beaming factor. The remainder must have another cause, presumably surface roughness effects. (3) The maximum in a reststrahlen emissivity spectrum does not occur exactly at the Christiansen wavelength where, by definition, the real part of the refractive index equals one, but rather at the first transition minimum in reflectance associated with the transition from particle scattering being dominated by volume scattering to that dominated by strong surface scattering. The transparency feature is at the second transition minimum and does not require the presence of a second band at longer wavelength for its occurance. Subsurface temperature gradients have only a small effect on emissivity bands.

Two different mechanisms, shadow-hiding and coherent backscatter, can cause a hot spot, or opposition effect, in the bidirectional reflectance of vegetation and soils. Because the two mechanisms sample different properties, it is important to know which one is primarily responsible in a given medium. This question can be answered by measuring the bidirectional reflectance in circularly polarized light. If the results of the limited experiments reported here can be extrapolated to a wider range of materials, it appears that the primary cause of the hot spot in most vegetation canopies and in moist, clumpy soils is shadow-hiding. However, in vegetation with large numbers of wavelength-sized structures, such as mosses, and in dry, fine-grained soils, the hot spot is dominated by coherent backscatter.

In a recent paper Mishchenko asserts that soil particles are strongly forward scattering, whereas particles on the surfaces of objects in the solar system have been inferred to be back scattering. Mishchenko suggests that this apparent discrepancy is an artifact caused by using an approximate light scattering model to analyse the data, and that planetary regolith particles are actually strong forward scatterers. The purpose of the present paper is to point out the errors in Mishchenko's paper and to show from both theoretical arguments and experimental data that inhomogencous composite particles which are large compared to the wavelength of visible light, such as rock fragments and agglutinates, can be strongly back scattering and are the fundamental scatterers in media composed of them. Such particles appear to be abundant in planetary regoliths and can account for the back scattering character of the surfaces of many bodies in the solar system. If the range of phase angles covered by a data set is insufficient, serious errors in retrieving the particle scattering properties can result whether an exact or approximate scattering model is used. However, if the data set includes both large and small phase angles, approximate regolith scattering models can correctly retrieve the sign of the particle scattering asymmetry.

The spectral evidence for the presence of iron on the surface of Mercury is equivocal. Estimates of strength of the 0.9 to 1.2 micron coordination band range from weak to non-existent. This band provides a measure of iron bound in pyroxenes, feldspars and olivine. This band is ubiquitous for spectra of Earth, the Moon, and Mars. It is reasonable to believe that Mercury would have similar total iron.

The intensity and polarization of light scattered by a variety of types of artificial partices large compared to the wavelength were measured as a function of phase angle. Shape, surface roughness, absorption coefficient, and internal scattering coefficient were varied systematically and their effects studied. Scattering by clear, smooth-surfaced spheres is in quantitative agreement with the predictions of the geometrical optics (ray theory) approximation to physical optics (Mie theory). The phase functions of almost all of the particles measured have both forward and backward scattering lobes. A two-parameter, double Henyey-Greenstein function generally provides reasonably good descriptions of the data, while keeping the number of free parameters to the minimum necessary. On a double Henyey- Greenstein parameter plot all of the particles fall into an L-shaped area of restricted size in which the location is characteristic of the particle type. Formalisms based on the equivalent slab model are also given for estimating the scattering efficiency of a large, irregular particle. For most dielectric particles the transmitted, forward scattered light is partially negatively polarized. It is this component that is respopnsible for the well-known maximum in the polarization curves of planetary regoliths at phase angles around 100 deg. For phase angles between about 30 deg and 70 deg the internally scattered light is found to be randomly polarized in the particles studied here, so that the only contribution to the second component of the Stokes vector is by Fresnel reflection from the particle surface. If this empirical result is general, measurement of the second Stokes vector of the light scattered from a regolith at these angles may provide a method of remotely measuring the mean refractive index.

The theory in which either or both reflected sunlight and thermally emitted radiation contribute to the power received by a detector viewing a particulate medium, such as a powder in the laboratory or a planetary regolith, is considered theoretically. This theory is of considerable interest for the interpretation of data from field or spacecraft instruments that are sensitive to the near-infrared region of the spectrum, such as NIMS (near-infrared mapping spectrometer) and VIMS (visual and infrared mapping spectrometer), as well as thermal infrared detectors.

Since the Apollo missions, we have emphasized the following points, which are based on theoretical calculations and on laboratory studies of the properties of evaporated silicate deposits and of lunar samples. The mass of vapor generated by impacts on the lunar surface is comparable in magnitude to the mass of impact melt glasses; the physics of impact into a porous regolith requires that much of this vapor be retained in the soil rather than lost to space (as is widely believed); experimental coatings made from vaporized or sputtered lunar basalt contain abundant inclusions of submicroscopic, super paramagnetic metallic Fe; and this Fe may explain the magnetic signature, low albedo, reddened spectrum, and subdued absorption bands of lunar regolith. Our conclusions have been generally rejected by the lunar geochemical community for two reasons: there seemed to be no direct evidence for vapor deposits in Apollo samples, and it seemed that the lunar optical properties could be explained by the presence of impact melt glasses alone. However, advances in our understanding of the optical properties of glasses and of light scattering by planetary regoliths, and now the direct detection of vapor deposits, show that these objections are not valid. Vapor phase transport is a major process on the lunar surface, and unless its effects are taken into account, the chemical, magnetic, and optical properties of the regolith cannot be understood.

This oral presentation explains that measurements of the opposition surge (the nonlinear increase in reflectance seen in particulate materials when observed at small phase angles) are the first ever made using the JPL long-arm goniometer, which permits very small phase angle measuremnets to be made.

The opposition effect, the sharp surge in brightness of an astronomical object observed near zero phase angle, which has been known for more than a century, has generally been explained by shadow hiding. The reflectances of several Apollo lunar soil samples have been measured as a function of phase angle in linearly and circularly polarized light. All samples exhibited a decrease in the linear polarization ratio and an increase in the circular polarization ratio in the opposition peak. This provides unequivocal proof that most of the lunar opposition effect is caused by coherent backscatter, not shadow hiding. This result has major implications for the interpretation of photometric observations of bodies in the solar system, including the Earth.

We observed slected mutual occultations of Jupiter's satellites in 1990-1991 and we used the occultation profiles to constrain the areal distribution of selected spectrally active compounds on Io's surface.

It was clear long before the Apollo missions that a darkening process occurs on the moon. However, its nature remains controversial and elusive. Current evidence implies that the darkening is associated with, and is probably caused by, submicroscopic metallic iron in the regolith. Questions discussed at the workshop include: (1) under what conditions will impact vitrification produce a dark glass; (2) what is the role of the submicroscopic metallic Fe (SMFe) in the lunar darkening process; (3) how is the SMFe produced; (4) is there a significant component of the regolith that has been deposited from a vapor, if so, what form is it in, and how can it be recognized, what are its effects on the chemistry of the regolith; (5) how do the processes of impact vitrification, vaporization, sputtering, and SMFe production vary as a function of distance from the sun and location in planetary magnetospheres; and (6) what other processes might affect optical properties. Ices have lower melting and boiling temperatures and sputtering yields several orders of magnitude larger than silicates. Hence, analogous processes will occur to an even greater extent on satellites of the outer planets, and these questions are relevant to those bodies as well.

We have measured the opposition effect, the nonlinear surge in reflectance seen in particulate materials as phase angle approaches zero degrees, in a suite of materials of varying particle size and reflectance. These samples were illuminated by linearly and circularly polarized monochromatic radiation at two wavelengths, 0.442 and 0.633 microns. By measuring the linear and circular polarization ratios for each sample, we have found that in highly reflective materials the behavior of the reflected radiation is consistent with the coherent backscattering process which has recently been proposed to explain the opposition surge that is seen in such media. The size and width of the coherent backscattering opposition peak vary as a function of reflectance of the sample. The opposition effect has been observed in particulate materials studied in the laboratory and it is also observed in the radiation reflected from solar system bodies which present a regolith to the earth based observer. The traditional explanation of the opposition effect, the shadow-hiding hypothesis, is that it was caused by the elimination of mutual shadows cast between the regolith grains as the phase angle of the observation became smaller. This shadow-hiding hypothesis, however, is unable to explain the opposition effect seen in highly reflective materials such as magnesium oxide and barium sulfate powders. This is because highly reflective media will multiply scatter the incident radiation between the regolith grains. This causes the shadows to be eliminated. We have measured the angular scattering properties of a suite of materials of different reflectivity. We have observed polarization ratios in reflective particulates that are consistent with coherent backscattering as the principal process which causes the opposition surge.

This paper reports the results of attempts to model the spectral properties of the lunar regolith as consisting of crystalline rocks, glass and submicroscopic metallic iron (SMFe), produced by a process involving vapor phase differentiation. The models differ in the location of the SMFe.

Reflectance spectrophotometry and polarimetry are major tools in remote sensing studies of surfaces of solar system bodies. The interpretations of such measurements are often based on laboratory studies of meteoritic, lunar, and terrestrial materials. However, the optical properties of regoliths are known to be affected by the space environment. Thus, some of the major questions addressed in the workshop include identity of the soil component responsible for alteration of the optical properties, the process that produced this component, and how reliably the effects of these processes could be extrapolated to other bodies of the solar system.

The photometric textures of Europa geologic units possessing the same small mean photometric slope angle as found by Dominique et al. (1991) are presently examined on the basis of Hapke's (1981) photometric model. With only one exception, the particles composing the surfaces of most of the units exhibit similar scattering properties. Attention is given to the single-particle scattering functions of the two bright plains regions examined; these are noted to be no less different from each other than both are from the other units. An aging process that affects the albedos and single-particle phase functions is hypothesized for Europa.

Ultraviolet-visible-infrared mapping digital array scanned interferometers for lunar compositional surveys was developed. The research has defined a no-moving-parts, low-weight and low-power, high-throughput, and electronically adaptable digital array scanned interferometer that achieves measurement objectives encompassing and improving upon all the requirements defined by the LEXSWIG for lunar mineralogical investigation. In addition, LUMIS provides a new, important, ultraviolet spectral mapping, high-spatial-resolution line scan camera, and multispectral camera capabilities. An instrument configuration optimized for spectral mapping and imaging of the lunar surface and provide spectral results in support of the instrument design are described.

The coherent backscatter model is investigated experimentally by means of an analog to examine the nature of the radar reflectivity of icy satellites. The laboratory analog involves the examination of He-Ne laser light with a wavelength of 0.633 microns reflected from 0.497-micron polystyrene beads suspended in water. A photomultiplier and a polarizing filter are employed to observe the radar at phase angles greater than zero which are extrapolated to infer observation at zero phase angle. The polarization angles measured at optical frequencies are found to be equivalent to those observed in the radar data. The results suggest that the regoliths of the icy satellites consist of matrices of small complex refractive indices which contain scatterers separated by distances of the order of a wavelength. The experiment explains the high reflectivity of Jupiter's icy satellites which results in the observation of narrow opposition-effects peaks at optical frequencies.

The coherent backscatter model of radar scattering from icy satellites argues that most of the received power comes from radio waves that have been multiply scattered from numerous refractive index discontinuities randomly located under the surfaces of the regoliths. In a laboratory analog experiment the circular and linear polarizations were measured in light scattered from a suspension of polystyrene spheres in water illuminated by laser. Results confirm that coherent backscatter from particle media reproduces the large intensity and unusual polarizations observed in the radar data. The model implies that the regoliths of icy satellites consist of matrices of small reflective index containing imbedded scatterers separated by distances of the order of a wavelength.

The surface of the Jovian satellite Europa is characterized on the basis of an analysis of ground photoelectric photometry at 470 and 550 nm and Voyager images. The data are presented in extensive tables and graphs and discussed in detail. At 550 nm, Europa has single-scattering albedo 0.964, opposition-effect amplitude 0.5, opposition-effect width 0.0016, double-lobed Henyey-Greenstein factors b = -0.429 and c = 0.113, and mean roughness angle 10 deg (much lower than on other solar-system objects). From the small roughness and the 96-percent porosity implied by the narrow opposition peak, it is concluded that the surface was formed mainly by endogenic processes. It is also noted that only one of three observational criteria for preferential ion bombardment of the trailing hemisphere are met in Europa.

It is presently suggested that the diffuse component-dominated high radar reflectivities and large polarization of Europa, Ganymede, and Callisto could be due to the illumination by a collimated source of a weakly absorbing particulate medium in which wavelength-scaled scatterers are separated by distances somewhat larger than the wavelength. A plausible medium could be a regolith composed of voids and/or silicate rocks imbedded in an icy matrix; multiply-scattered parts of the wavefront traversing the same path in opposite directions combine coherently in the backscatter direction to generate increased intensity, and the enhancement is different for the two components of polarized reflected radiation.