Studies of the role of chirality in the scattering of electromagnetic (EM) waves in a discrete random chiral composite are presented, using a vector multiple scattering formalism in the microwave region. For the first time the effect of concentration, chirality, and size of the individual scatterers are related to the chirality and wavenumber of the effective medium. The chiral composite is modeled as an infinite nonchiral medium containing a random distribution of identical finite scatterers made of a chiral material. The concept of a T-matrix for a single scatterer is used to relate the scattered EM fields to the exciting plane-polarized fields. Ensemble averaging over the position of the scatterers, along with the quasi-crystalline approximation (QCA), results in a frequency dependent dispersion equations. Under long wavelength approximation, dispersion relations relating wavenumbers K(L), K(R) of the scattered left- and right-circularly polarized (LCP and RCP) waves are obtained and it is verified that they are circularly polarized. From a knowledge of these wavenumbers, the effective wavenumber and effective chirality of the chiral composite medium can be obtained.

A model is developed for simulating the behavior of electromagnetic waves in a composite material consisting of a lossy dielectric medium with slendered inclusions of conductors. The model is used for solving the problem of homogenization for classical Maxwell equation system. The results are extended to composite media with conducting fibers, to a case where conducting inclusions are not simply connected, and to a composite medium made of two chiral materials.

Considered here are the effective medium properties of a chiral composite made of a dilute concentration of small non-interacting chiral spheres, randomly suspended in a chiral host medium. Volume integral equations to determine the scattering characteristics of an inhomogeneous chiral scatterer are given. These equations are utilized to derive the electromagnetic polarizabilities of a small, homogeneous, chiral sphere embedded in a homogeneous chiral host medium. Finally, from the polarizabilities, the effective properties of the chiral-chiral composite are obtained.

The procedure for the preparation of CdSe/CdS/ZnS photoconductors is described, and the results of characterization of photoconductors are presented. It was found that the spectral response peak of the photoconductor can be shifted to the shorter wavelength region by changing the relative content of ZnS. However, this can be done only at the cost of decreasing photosensitivity.

A simplified (discrete angle, DA) radiative transfer theory is presented as a computationally and conceptually advantageous alternative to standard (continuous angle) theory. After briefly reviewing the basic ideas of random fractal geometry and multifractal cascade theory, we present some of our recent two-dimensional DA numerical simulations of transfer through a specific log-normal multifractal cloud model where the radiation fields are spatially resolved on a 1024 x 1024 point grid. Using this data base, we demonstrate (1) how in inhomogeneous transfer problems horizontal fluxes work in quite subtle ways to create dramatic overall differences with homogeneous predictions for the same amount of scattering material, and (2) how strongly multiple scattering can smooth extremely singular density fields. Furthermore, both of these effects are enhanced by increasing optical thickness which can be viewed as a measure of the strength of the nonlinear coupling between the density and radiance fields. Finally, we discuss some basic inequalities that arise between the various ways of computing overall (spatially averaged) response to illumination.

The extreme variability of geophysical fields can be characterized by scale invariant (sensor resolution independent) 'codimension' functions, which are exponents characterizing the probability distribution. These codimension functions form a three parameter universality class. The parameter H measures the degree of nonstationarity of the process, C1 characterizes the sparseness/inhomogeneity of the mean of the process, (alpha) characterizes the degree of multifractality; (alpha) equals 0 is monofractal, (alpha) equals 2 is the maximum. We review the properties of these multifractal processes and describe the 'double trace moment' technique that is the first data analysis technique specifically designed to estimate these parameters. The technique is then applied to digital elevation maps of Deadman's Butte, to the topography of France, to a pair of aircraft photos of the ocean surface, and to a visible satellite image of a cloud field.

Pertinent issues concerning cloud-radiation interactions that are relevant to studies of climate are discussed in terms of cloud optical properties. These optical properties are classified either inherent or apparent; the former are functions of cloud microphysics, the latter come about from the illumination of the cloud by radiation. The connection between the two sets of optical properties is discussed under the format of radiative transfer. The state of our lack of understanding of this connection is illustrated using examples derived from recent observational studies. Further evidence is presented that questions the validity of one dimensional radiative transfer theory as applied to the earth's atmosphere.

The wavefront of an acoustic, low-frequency source propagation a long distance in the ocean can be recovered by a holographic technique which removes the effects of ocean variability within measurements of that source at an acoustic receiving array. The method assumes that the ocean varies adiabatically in range, and that any variability along the path between a reference and the unknown source can be modeled. The phase distortion along a particular path can be corrected by combining reference and unknown signals so that their relative phases nearly cancel. Since sound propagates to long distances in the ocean over many paths and since each of these paths are distorted in a different way, a different phase correction must be provided for each path. The method that we will present has its origins in an optical technique for wavefront reconstruction in which a distorted image is corrected by placing a reference source sufficiently near the object that their fields propagate through the same media. Recording the reference and object fields together with a hologram provides the phase correction. A simple theory based on the propagation of normal modes in the ocean will be presented. Ocean variability is introduced through a range dependent sound speed profile which is incorporated into the normal modes by means of the adiabatic approximation valid for weakly, range dependent oceans. The wavefront reconstruction technique will be applied to the problem of locating an unknown source.

An efficient simple numerical procedure is used to produce the motion of shockfront with small amplitude through inhomogeneous media. The mathematical model uses an intrinsic coordinate system and is based on Whitham's (1974) theory of geometrical shock dynamics. By using this model, the motion of the wavefront can be determined without explicitly calculating the flowfield quantities behind the wavefront. The model is given by a system of four partial differential equations (PDEs) which for small amplitude wavefront does not produce caustic in the wavefront. The small variations in sound speed and the corresponding distortion of the rays due to nonlinearity are included in the system of PDEs. Rays are given as the orthogonal trajectories to the wavefront and, because of the nonlinearity effects of the media, rays are not straight lines. We represent the wavefront by a discrete set of points, and then we propagate each point along orthogonal trajectories with sound speed determined by a discretized set of two PDEs relating Mach number and area of the ray tube. The numerical results obtained using the above procedure is compared with exact numerical solutions and experimental data given by other authors.

Many natural structures possess self-similar multiscales which can be characterized by power law spectra. Under appropriate conditions, knowledge of the strength of these scale sizes provides information on the physical processes which form these objects. In this paper, we investigate wave interactions with continuous fractal layers which model geological and variegated structures. Since fractal characteristics of the layers are embedded in the scattered field, they can be retrieved under appropriate conditions. This inversion can be performed in either the frequency or the time domain as desired.

Reflection by and transmission through a chiral slab are explored for normally incident plane waves, it being assumed that the constitutive parameters of the slab vary along the direction of incidence. It is shown that the Faraday and the Ampere-Maxwell equations decompose into two autonomous matrix differential equations. The effect of the chiral slab is accounted for by the matrices S+/- that can be numerically evaluated for a linearly inhomogeneous chiral slab using two methods.

Extensive literature is available for wave propagation and scattering in random media and rough surfaces. However, much less work on pulse propagation and scattering in random media has been reported. In this paper, we discuss some of the recent developments in pulse propagation in turbulence, discrete scatterers, and rough surfaces.

An optimization algorithm is developed and used in conjunction with a synthesis program for the determination of the number of layers, as well as the dielectric constant of each layer, in composite multilayer device designed for use in microwave frequency range. Numerical results are presented using a three-layer design, in which the first layer is made of dispersing 72 wt pct barium strontium titanate (BST) powder in eccogel, the second layer is a pure eccogel material, and the third layer is similar to the first, except the content of BST powder is reduced to 54 wt pct.

To obtain the effective refractive indices of three-phase coating materials, in which multiple scattering can also occur between the two different discrete phases, a multiple scattering formalism for two-phase systems is modified to include the third phase. Numerical results for different refractive indices of various arrangements are presented for a paint coating in which some of the TiO2 pigment particles are replaced by microbubbles.

A switching technique using the polarization sensitivity of photorefractive holograms in crystals is experimentally demonstrated. The design is capable of handling a large number of 2 X 2 switching channels with only four holograms. The concept is applicable to either synchronous or asynchronous switching and also to a variety of photorefractive materials.

This paper reports an improved laser-addressed LCLV that employs scattering structure of smectic liquid crystal as display background. Its LC layer will not be easily electrically broken down. Full erasure can be achieved within 0.5 - 1 second. We also present a new method utilizing an insulative layer (inert insulative film) that is sandwiched between the LC layer and transparent electrode-absorbing layer. Such structure makes it possible to avoid decreasing LCLVs input sensitivity, which is caused by an increase of LC layer thickness, to improve bearing ability to high voltage and large current, and to achieve fast full erasure. Its highest breakdown voltage reaches 690 Vrms.

Chiral, literally means handed and comes from the Greek work cheir for 'hand' since our hands are non-superimposable on their mirror images. As Lord Kelvin defined it in 1893: 'I call any geometrical figure, or any group of points, chiral, and say it has chirality, if its image in a plane mirror, ideally realized, cannot be brought to coincide with itself.' Characteristically, a substance with chirality must have neither a center nor a plane of symmetry for its molecular framework. Chirality is present in many organic substances, liquids, and crystals, but also in more common objects such as golf clubs, helices, human hands and feet, etc. Recently, chirality has aroused much interest due to its interesting effects on electromagnetic wave-material interaction resulting in drastically changed polarization, reflection, and absorption relative to a non-chiral material with the same (epsilon) and (mu) . There are two phenomena that can be observed and measured when an electromagnetic wave travels through a suspension of chiral inclusions. Optical rotatory dispersion (ORD), which is due to the different phase velocities for the left and right circularly polarized waves, rotates the plane of the polarization. Circular dichroism (CD), which is caused by the different absorption for the left and right circularly polarized waves, elliptically polarizes the transmitted fields.

In recent years, considerable interest has been focused on the study of scattering from very rough surfaces. Very rough surfaces are defined as those with an rms height variation of the order of a wavelength and an rms slope of the order of unity. These surfaces occur in several practical problems including underwater acoustics, microwave scattering by terrain, ultrasound scattering by tissues, and optical scattering by rough metallic surfaces. In addition, recent experimental and numerical studies show backscattering enhancement phenomena from very rough surfaces.16 In spite of their theoretical interest and practical importance, theories of scattering by very rough surfaces are scarce7 and outside the range of validity of conventional theories such as the perturbation method and the Kirchhoff approximation.811 Recently, we proposed a theory based on the modified Kirchhoff approximation (KA) with angular and propagation shadowing for one—dimensional Dirichlet rough surfaces.12 This paper extends our previous theory to include scattering by very rough metallic and dielectric surfaces. The range of validity of the theory is examined by comparing it with the Monte Carlo simulation. Some material in this paper is also in our recent papers.2022 Numerical studies on very rough surfaces show that the second—order KA, when the surface integral is limited within the distance of double bounces without being intercepted by the surface, agrees well with the exact Monte Carlo simulation.13 This is consistent with the observation made by Liszka and McCoy that any signal which intersects the rough surface, one or more times, will be canceled by some higher iteration.14 This cancellation is accomplished by the shadowing function. This also indicates that the first— and second— order KA with proper shadowing gives a good approximate solution to very rough surface scattering. Conventional shadowing functions are used for first—order KA scattering. For second—order KA, in addition to the conventional shadowing, the shadowing representing the probability that the wave scattered from a point on the surface arrives at the other point on the surface without being intercepted by the surface is included. This shadowing is given by the angular and the propagation distance probabilities. These two shadowing functions, angular and propagation, modify the second—order KA and give the proper energy conservation and enhanced backscattering. The shadowing functions for the second—order KA used in this paper are similar to that used by Jin.15 The effect of shadowing for double scattering by random surfaces is also studied by Pavel'yev.16 Our analytical method employs the positive and negative traveling waves for the second— order KA and this results in a clear physical interpretation of the processes for the ladder and the cross or cyclic terms. The cyclic terms represent two waves propagating over the surface in opposite directions, giving rise to the backscattering enhancement. We also use Fourier transform in the vertical direction to facilitate computation of the second moments. Figure 1 shows the approximate range where our theory is applicable in terms of an rms height o and a correlation distance 1. The ranges where the Kirchhoff approximation, field perturbation, and phase perturbation methods are valid are noted by KA, FP and PP, respectively. The range where backscattering enhancement takes place is noted by E, and this is the range where none of the conventional techniques is applicable. Those points marked with fl are where our theory agrees well with exact numerical simulations, and the energy is conserved within 5 % error. Our theory covers most of the range E where the enhancement takes place, and the theory also reduces to conventional KA in the range where KA is applicable.

The canonical problem in theoretical work on rough surface scattering concerns monochromatic plane waves incident on a single infinite irregular interface in a constant sound speed environment. Scattering amplitudes are calculated, and it is hoped that the differential cross-section per unit area (i.e., the scattering strength) formed from the square of the magnitude of the scattering amplitude will be adequate to describe experiments which use point sources in variable sound speed settings where there are multiple boundaries, e.g., reverberation in the ocean. Ray based reverberation models which allow computation of a local angle of incidence do use the differential cross-section. However, there are many workable wave propagation models for which the scattering strength can only be introduced in an ad hoc manner. In the present paper some variational principles will be examined as starting points for describing rough surface scattering by point sources. Dashen's exact, manifestly reciprocal formula for scattering amplitudes follows from a variational calculation and can be somewhat generalized to a point-source, variable sound-speed setting. In the special case of reverberation with a constant sound-speed with a pulsed source, the small-slope approximation, which follows simply from Dashen's formula, will be shown to have a simple form in both the near and far field.

We present computational results for scattering from rough surfaces. All the results are for the Dirichlet boundary value problem and one-dimensional surfaces s(x). Electromagnetically this corresponds to TE-polarization. The first set of results describes the reconstruction of rough- surface profiles from scattered field data. Two methods are presented. Both use the spectral- coordinate approach. The first is based on perturbation theory and is valid when kcos((theta) s)s(x)<<1, where k is wavenumber and (theta) s the scattering angle. Shallow profiles are reconstructed well. The second is based on the Kirchhoff approximation for the normal derivative of the total field on the surface. Using a combination of incident and scattered angles we develop a Fourier transform relation between the scattered data and the surface profile. The result is also good only for shallow surfaces. Both results are FFT-based. The second set of results is based on ensemble average results for homogeneous Gaussian distributed random surfaces. We illustrate two conclusions. They are (1) the coherent specular intensity is predominantly single scattering even when multiple scattering occurs, and (2) beyond a certain roughness the predominant field in the specular direction is incoherent rather than coherent.

The integral operator connecting the normal gradient of the scattered field to its boundary values can be represented by an expansion in roughness amplitude that is more rapidly convergent than conventional solution. Each term in the operator series consists of alternating applications of Fourier transforms and multiplications by functions of surface position and wavenumber. The procedure is consequently efficient enough to provide accurate scattering solutions for two-dimensional surfaces (zetz) (x,y) of high roughness amplitude and substantial detail. The appropriate expansion parameter for the scalar problem is the Fresnel roughness, which for a composite random surface scales like the rms slopes times the rms Rayleigh height. For the vector electromagnetic problem an additional, partially independent, parameter arises in the form of squared slope for those roughness scales shorter than the radiation wavelength.

An experimental study of light scattering from gold-coated ground glass and chemically etched surfaces is presented. A substantial amount of energy was detected in the cross-polarized component of the light scattered by all our samples and, with the chemically etched surfaces, the phenomenon of enhanced backscattering was observed. The surfaces were characterized with a mechanical profilometer but the estimated statistics are inconsistent with the evidence presented by the light scattering experiments. We believe that this is due to a lack of fidelity of the profilometric measurements. A computer simulation of the effects of the non-vanishing stylus tip has also been conducted to get some insight into the causes of this discrepancy.

It is generally believed that the enhanced backscattering of light from a highly reflecting, multiply-scattering, random surface is due to the coherent interference of each multiply- reflected optical path with its time-reversed partner, and is already present in the double- scattering approximation. If enhanced backscattering is indeed a multiple-scattering effect we should see it from any random surface that can multiply scatter light. The statistical properties of the surface, such as whether the surface profile function is a Gaussianly-distributed random variable or not, whether it is stationary or not, and the form of the surface height correlation function, should therefore be of secondary importance in determining whether enhanced backscattering occurs or not. In this paper we study the role played by the form of the surface height correlation function in the existence of enhanced backscattering and in the dependence on the scattering angle of the contribution to the mean differential reflection coefficient from the incoherent component of the scattered light. We consider the scattering of p- and s- polarized beams of light incident normally onto a random, one-dimensional metallic surface, when the plane of incidence is perpendicular to the generators of the surface. Four different forms of the surface height power spectrum g(Q) are considered: (a) g(Q) equals (pi) a exp(-Qa); (b) g(Q) equals (pi) 1/2a exp(-Q2a2/4); (c) g(Q) equals 2a[1 - (Qa/(pi) )](theta) ((pi) - Qa); and (d) g(Q) equals a(theta) ((pi) - Qa), where we have presented them in the order of increasing rate of decay to zero with increasing Q. In these expressions (theta) (Q) is the Heaviside unit step function. For each form of g(Q) we have calculated the mean value of , the distance between consecutive peaks and valleys on the surface and the variance of this quantity, (sigma) d equals [2> - 2]1/2. The values of these quantities are correlated with the rate of decay of g(Q) with increasing Q. For a fixed values of the wavelength of the incident light we then calculate the contribution to the mean differential reflection coefficient from the incoherent component of the scattered light when the value of a in case (b) above is varied in such a way that (lambda) / increases systematically, with the rms slope held constant. Enhanced backscattering is observed in each case. The width of the enhanced backscattering peak is related to the value of (lambda) /, as is the occurrence of first order subsidiary maxima. The latter disappear when (lambda) / has increased to about 0.6. This is interpreted as due to the inability of the incident light to resolve the structure of the surface responsible for these subsidiary maxima. Similar calculations are carried out for a band-limited fractal surface, characterized by g(Q) equals ((pi) a/tan-1Qoa)(theta) (Qo - Q)/(1 + a2Q2), and similar results are obtained. No evidence of second order subsidiary maxima is seen in the differential reflection coefficient. This result is believed to be due to the magnitude of (sigma) d/ for each of the forms of g(Q) considered. We conclude that while the detailed form of the mean differential reflection coefficient depends on the form of the surface height correlation function, the existence of enhanced backscattering does not, as long as the surface remains multiply reflecting.

The backscattering of an optical wave from arbitrary convex dielectric objects with rough surfaces is investigated and the formulas for calculating the backscattering cross section of both coherent and incoherent fields are obtained. In the infrared waveband, the influence of the geometry, permittivity, and statistic characteristic of the rough surface on backscattering cross section is analyzed, making use of rough ellipsoids as examples.

In a dense medium, the particles do not scatter independently. The effects of correlated scattering become important, and the spatial correlations of particles have to be included [1-5]. These have been verified in controlled laboratory experiments [3,4]. Propagation and scattering in dense media have been studied with the quasicrystalline approximation [2], and the quasicrystalline approximation with coherent potential for the first moment of the field [1] and the correlated ladder approximation for the second moment of the field [6]. The dense medium radiative transfer theory has also been developed from these approximations to study multiple scattering effects in dense media [6-8]. We have recently extended the results to medium to high frequencies and included the effects of Mie scattering from correlated scatterers of multiple sizes [9]. As a function of frequency, scattering first increases rapidly in the Rayleigh regime, then starts to level off at the Mie scattering regime. Comparisons have been made with experimental data of snow.

Reflection and transmission characteristics of chiral composites are obtained for a normally incident, linearly polarized plane wave by employing a free-space measurement system in the frequency range of 8-40 GHz. The artificial chiral composites are fashioned by embedding chiral inclusions (metallic helices) into an epoxy medium (Eccogel) (Guire et al. 1990). One reflection measurement is made for each sample since the reflected field is linearly polarized as the incident wave. Two transmission measurements at different polarization directions are needed in order to fully characterize the transmitted polarization ellipse. The rotation angle and ellipticity of the transmitted polarization ellipse are calculated from these two transmission measurements. By examining the characteristics of the ellipticity and rotation angle, the Cotton effect is observed in the frequency range where the maximum power absorption occurs. This ensures that the left- and right-handed chiral composites can be represented as left- and right-handed electromagnetically active media, respectively, in the frequency range of interest. From one reflection and two transmission measurement data, the electromagnetic properties of chiral composites are for the first time computed numerically using a suitable inverse algorithm.

We consider the scattering of light by a system of randomly distributed particles above an interface. It is shown that this system can produce backscattering enhancement even for a very dilute system such that multiple scattering is negligible. A physical mechanisms is proposed and the role of the Fresnel reflectivity of the interface is examined. Differences between the behavior of s and p-polarization are considered.

The scattering of electromagnetic radiation from a finite conducting cylinder with complex permittivity at an arbitrary orientation was analyzed using a first order approximation to the iteration technique for the integro-differential equation first developed by Shifrin and later modified by Acquista. The classical Kerker solution for a simple infinite dielectric cylinder was extended to a more physically realistic solution according for a finite length cylinder with complex permittivity by a modified Drude conductivity approach. The diameter of the cylinder is on the order of one wavelength of the incident radiation. The lowest order approximation to the internal field solution for the iteration process is a function of the effective polarized electric field inside the cylinder and the polarization matrix of the scattering medium. The polarization matrix of the cylinder is determined from the electrostatic solution for a finite cylinder in a constant electric field, and is a function of the length to diameter ratio (aspect ratio) and the permittivity of the cylinder. The electrostatic solution for a finite cylinder does not permit a closed solution; therefore the cylinder is approximated by an inscribed ellipsoid which provides a converging analytic expression. Results are compared to published data. The complex frequency dependent permittivity of the cylinder material was modeled using a modified Drude conductivity approach. The effects of typical variations in the length diameter, and bulk conductivity of the cylinder were analyzed for TE, TM, and TEM polarizations.

Temperature is a very important parameter in physic oceanography research and ocean forecast. The ocean scientific and technical community has been laying extreme stress on fast measuring sea temperature distribution over large areas. Recently, remote sensing of sea surface temperature measurement with an infrared channel of NOAA satellite became an advanced technical means. In China, a group at Ocean University of Qingdao has obtained excellent successful results in inversing sea surface temperature from satellite data. But, it is difficult to obtain sea temperature vertical distribution with fast measurement over large area from satellite remote sensing. The data of sea temperature vertical distribution is very important for the research of sea-air interaction and ocean frontier behavior. Recently, the method of measuring sea temperature profile with lidar techniques has made some progress. Laser Raman spectrum is a good method. Therefore, this paper mainly reports the results of finely measuring laser Raman spectrum of different temperature sea water in lab, and the data processing method of obtaining temperature parameter from Raman spectrum. The experiment proves that the temperature precision of the data processing method reaches +/- 0.30 degree(s)C.

Experimental results for autocorrelation function of scattered laser light from rotating ground glass are reported, which set aside the controversy regarding its speed and angular dependence.

Precipitation media consisting of a variety of scatterers such as raindrops, hail, graupel and snow cause random fluctuations and scattering of the incident wave. A random distribution of many discrete scatterers causes the precipitation structures to be modeled as random media. The physical properties of interfaces and dielectric properties of the scatterer and background medium are the major factors responsible for scattering behavior at microwave frequencies. Waves scattered by random media have the potential to remotely determine microphysical properties such as average size, shape, orientation and concentration of scatterers. Most clouds evolve with the initial growth of liquid cloud droplets. Their subsequent development can be via a warm- or cold-based process, or a mixed-phase process. Warm-based clouds are warmer than 0°C and free of ice. Precipitation formation is restricted to condensation and coalescence. Most tropical clouds are formed by a warm process. In a cold-based process, diffusional growth of ice crystals is the dominant growth mechanism. Cold processes are common in winter storms and anvil regions of thunderstorms. In the mixedphase process, the cloud development begins with the condensation and coalescence process, but subsequently involves an ice process as well. In general, the precipitation structure contains both liquid and ice with varying shape and size throughout. Multiparameter radar measurements based on dual-polarization and dual-frequency techniques (in addition to the conventional Doppler parameters reflectivity, velocity, and velocity spectrum width), can p1ai an important role in the remote sensing of precipitating clouds. These techniques are fairly well established and can yield significantly more microphysical information than the Doppler products. For example, that the differential reflectivity ZDR, specific attenuation rate at X-band (10 GHz) A, and specific differential propagation phase shift KDP are important measurables related to the microphysical evolution of the cloud. In particular, the vertical profiles of the above radar observables are closely related to precipitation content and water phase2' ' ' , while ZDR has been shown to be useful for differentiating between rain and ice . A coupled graupel melting and radar model together with aircraft 2D-PMS (Particle Measuring System) data has shown that the vertical profile of ZDR provides an indication of the onset and progression of melting ice into raindrops5'7. Hail detection using ZDR and dual-frequency ratio hail signal (DFR-HS) is well documented5. At long wavelengths, ZDR is an excellent estimator of the reflectivity- weighted mean axis ratio of the raindrops filling the radar resolution volume8'9. TheX-band specific attenuation is related approximately to the fourth moment of the raindrop size distribution . Also, S-band (3 GHz) KDp is nearly related to mass weighted axis ratio of the raindrop size distribution9. The above mentioned multiparameter radar observables show marked differences in the ice, melting ice, and rain regions because the shape, orientation and dielectric constant are distinctly different in each region. Active remote sensing techniques using radar are capable of mapping the vertical inhomogeneities in precipitation clouds, but is not effective for global precipitation retrieval, especially over ocean. Spaceborne radiometer measurements over both sea and land offer an obvious advantage. The uncertainties associated with top-of-atmosphere (TOA) brightness temperature (TB) measurements to infer cloud vertical structure can be reduced by multi-frequency radiometer techniques and/or simultaneous radar observations . In the microwave regime, rain and cloud drops are good absorbers (albedo 0.5), while ice is essentially nonabsorbing (albedo < 0.95). Passive observations of the precipitation media with a microwave radiometer can be broadly classified as emission-based, scattering-based, or a combination of both . Scattering-based observations are generally made at frequencies above 60 GHz, where thermal emission from the underlying rain layer is scattered away from radiometer field of view by the presence of high albedo ice or snow. Thus, the scattering-based methods are relatively insensitive to the background surface. However, since the ice is responsible for the cold TB, the rain rate estimate through a purely scattering-based method is indirect. Emission-based methods are generally made below 22 GHz, and rely upon increases in the thermal emission from rain and cloud over a radiometrically cool ocean in accordance with Kirchoff's Law. In the intermediate frequency range, both scattering and emission/aborption processes are equally important. However, it is difficult to separate ground effects such as reduced surface emissivity due to land surface wetting. Using the differences in polarization brightness temperature, Spencer et al. developed a method to estimate rainfall over the ocean. The upwelling brightness temperature is the net result of interactions within a radiometer's footprint including all of the particle types (rain, melting, ice, etc.) and the background medium (land or ocean). Hence, it is difficult to resolve the individual layers in a precipitation media by using just a single frequency and polarization state. Recently, there have been a number of radiative transfer models developed which incorporate multi-frequency, multi-polarization components'3'14"5"6'17"8 . These are vertically and angularly detailed plane-parallel radiative transfer models. In this review, we will discuss the multifrequency TB output of a plane-parallel radiative transfer model which uses both multiparameter radar and cloud model data as input. Multiparameter radar is capable of measuring both range resolved backscatter parameters (Z, ZDR, and linear depolarization ratio LDR), and range cumulative propagation parameters such as KDP and Ax. However, TB observations represent a vertically integrated effect through the cloud structure. Similarly, resolution in radiometer observations can be improved by using frequency diversity. Thus there is ample scope in improving the existing remote sensing techniques by combining both active and passive microwave methods. In this review, multiparameter radar observations and their interpretation for microphysical retrieval are discussed first, while radiative transfer model simulations using both radar measurements and cloud model results are presented in the following section. Both qualitative and quantitative aspects of the precipitation remote sensing is emphasized throughout the discussion.

A new technique--the mutual interaction method (MIM)--allows computation of the exact scattering characteristics of an object near a rough surface, or any pair of scattering objects, from their individual scattering functions. MIM is well suited to numerical computation; moreover, for a class of small scattering objects near plane surfaces, analytic solutions can be found. The application of MIM to the surface scatter problem is described and illustrated.

The main problem regarding free surface temperature measurement by infrared pyrometry is the lack of emissivity data of shocked materials. Moreover, the short rise time in an hydrodynamic experiment requires particular detectors and experimental measurement technique. This paper describes a three channel pyrometer adapted to the study of shocked tin sample melted in release. The radiative infrared emission is collected by ZnSe lens and carried from the firing chamber to the pyrometer by fluoride glass optical fiber. In order to deduce the temperature from the electrical signals amplified before recording on a numerical oscilloscope, we use different narrow filters associated with a previous static calibration of the detectors by means of a continuously heated black body. The spectral integration of Planck's formula and the comparison between the static calibration with the black body and the dynamic signals give, for each experiment, three temperature-emissivity couples in agreement with Planck's theory. The originality of this measurement technique is the use of a 15 meter infrared optical triple core fiber for transporting the radiation from the target to the detectors, which simultaneously permits the measurement of three temperature-emissivity couples in the case of very small and hermetic experimental set-up. The main characteristics of this pyrometer are: a rise time of about 20 nanoseconds; an analysis area with a diameter smaller than 5 millimeters; continuous free surface temperature measurement during more than 5 microseconds; and the study of the wavelengths between 2 and 5 micrometers. Experiments have been achieved with optical polished tin samples. The abacus voltages, temperature, and emissivity versus time for a 45 GPa shock pressure is discussed.

We discuss measurements of the infrared scattering properties of one- and two-dimensional conducting randomly rough surfaces. The surfaces are fabricated in photoresist and are checked with a stylus profilometer to verify that the surface statistics agree with the desired results. For surfaces that have steep slopes and lateral scale sizes comparable to the illumination wavelength, we observe strongly enhanced backscattering toward the source. These observations are shown to be strongly dependent on polarization. In the case of a one- dimensional surface, four distinct quantities appear in the Stokes scattering matrix, and examples of measurements of these quantities are presented. For the case of a two- dimensionally rough surface it is discussed that, even if the incident field is purely linearly polarized, the scattered light consists of both polarized and randomly polarized components. In the backscattering region, the polarized component contains linear, elliptical, and even nearly circular polarization states at various field angles. These data are interpreted and are consistent with the statistical isotropy of the surface.

Small angle scattering in and near the specular direction has recently been measured with a monostatic laser interferometric reflectometer for a laser beam incident normal to the surface. Employing the Ricean statistics of amplitude distribution for coherent glint with an additive scattered component of speckles, the scattered portion has been partitioned from the total reflected signal at 1.06 and 10.6 micrometers . Such measurements are basic to surface metrology and useful in assessing the performance of mirrors used in lasers and astronomical telescopes.

In this paper, a Nd:YAG Lidar system (YLS) is described. The backscattering problem of sea water is surveyed in the China Sea. A concurrent Raman scattering signal is used as a calibrating signal to reduce the influence of the variable random sea surface. The calibrated results show that this method is feasible and this experimental system can be used in a survey of scattering characteristics in the China Sea.

We report measurements of the statistical properties associated with the intensity fluctuations of a He-Ne laser propagated through a laboratory simulated atmospheric turbulence as well as through the open air atmospheric turbulence region. From these measurements the probability density function representative of the random process was identified by comparing the measured parameters with those of standard theoretical models. Photon counting techniques have been used in all our measurements.

An adaptive optical neuro network using inexpensive pocket size liquid crystal televisions (LCTVs) was recently developed by graduate students in the Electro-Optics Laboratory at The Pennsylvania State University. Although this neuro-computing has only 8 X 8 equals 64 neurons, it can be easily extended to 16 X 20 equals 320 neurons. The major advantages of this LCTV architecture as compared wither other ONCs, are low cost and flexibility to operate. To test the performance, several neural net models are used. These models are interpattern association, hetero-association, and unsupervised learning algorithms. The system design considerations and experimental demonstrations are also included.

This paper presents a large-scale static holographic optical neural network (HONN) based on volume holographic materials. A 1024 neuron HONN has been experimentally demonstrated for pattern recognition operations. It is possible to build a compact size and cost-effective HONN of 64,000 neurons which is capable of storing up to 20,000 training patterns and performing parallel processing in high speed (10 exp 13 interconnections per second).

This report presents an application of a learning network to the detection of cell membranes in human corneal endothelial micrograms. Our neural network model is a multilayered feed- forward network, and units in any single layer are divided into clusters. Every unit in the higher layer is connected with some of the units in each cluster of the lower layer. Units in the same layer have the same size of receptive field. In order to perform space-invariant processing in the same cluster, units in the same cluster have the same pattern of connectivity, but units in the different clusters have a different one. Such a network has been shown to be robust against distortions of input patterns and to match well with optical implementations. The neural network is trained by small parts of a microgram to extract the boundaries of the endothelial cells using the supervised learning algorithm. Desired output images are their cell membrane images that are traced by hand. After training, the network showed good performance with the whole microgram, which contained non-experienced parts. The final membrane image was obtained with the help of additional processing by a conventional digital filter based on mathematical morphology and linear filtering. The approach for shortcut learning and the internal representations of the network are studied.

The spatial light modulator (SLM) is a critical element in most optical processing systems. Different devices on the market today include the Hughes Liquid Crystal Light Valve (LCLV), the Ferroelectric LCLV, the GEC-Marconi LCLV, the Semetex MOSLM, Liquid Crystal Televisions, and the Deformable Mirror Device. The parameters of the above modulators have been evaluated at the Army Missile Command's Photonics and Optical Sciences Labs at Redstone Arsenal in an effort to determine the utility of these modulators as image transducers in optical correlator architectures. This paper will focus on another device perhaps applicable to optical correlators, the Microchannel Spatial Light Modulator (MSLM). The results of speed, maximum resolution, and visibility measurements will be presented.

Liquid crystal televisions have received extensive attention in literature for use as input and Fourier plane devices in joint transform correlators, and for use in other optical processing architectures requiring television rate inputs. The device has also been used in incoherent optical neural network systems at Penn State University. Recent research investigated the hybrid phase and amplitude modulating properties of the LCTV in a joint transform correlator. A new LCTV made by Epson is the focus of this paper. The hybrid modulation properties of this television will be evaluated and the results presented. Results from applications using the new LCTVs will also be presented.

Electron trapping materials are stimulable phosphors that can optically perform analog addition, subtraction, and multiplication. An optical, adaptive matrix-vector multiplier required for a neural network can be implemented using electron trapping materials.

A polychromatic optical neural network using cascaded liquid crystal televisions (LCTV) that uses a polychromatic interconnection weight matrix (IWM) for color pattern recognition is presented and simulated. Extension of the polychromatic neural net for multichannel operation is proposed.

Three adaptive version of the Ho-Kashyap perceptron training algorithm are derived based on gradient descent strategies. These adaptive Ho-Kashyap (AHK) training rules are comparable in their complexity to the LMS and perceptron training rules and are capable of adaptively forming linear discriminant surfaces which guarantee linear separability and of positioning such surfaces for maximal classification robustness. In particular, a derived version called AHK II is capable of adaptively identifying critical input vectors lying close to class boundaries in linearly separable problems. We extend this algorithm as AHK III, which adds the capability of fast convergence to linear discriminant surfaces that are 'good' approximations for nonlinearly separable problems.

The concept and implementation of an inner-product associative memory is here motivated largely by recognizing the excellent suitability of modern optical devices (such as magneto-optic spatial light modulators (SLMs) and self-scanning detectors) for realizing matrix algebra operations. Binary input vectors and reference vectors are loaded into SLM devices, and a correlation is produced in accordance with the effects on polarization and phase, including the summing function provided by a lens. In this simple way, the first half of an inner-product associative memory is constructed, which should then be followed by a nonlinearity at the correlation plane (as opposed to the output plane). A photorefractive element allows the associative memory system to be completed; this element is used in an unconventional manner.

A systematic procedure is presented for designing optimal correlation filters for implementation on deformable mirror devices (DMDs) exhibiting cross-coupled amplitude and phase characteristics. The utility of the algorithm for designing such filters is illustrated using five different device characteristics: phase-only filter, a binary phase-only filter, a diagonal line characteristic, a DMD zeroth-order characteristic, and a DMD first-order characteristic. Results are also presented regarding the signal-to-noise ratio and peak-to-correlation energy obtainable using these filters. The performance achievable using DMD type characteristics was found to be close to that of phase-only filter.

A new technique for implementing fully complex spatial filters with a phase mostly deformable mirror device (DMD) light modulator is described. The technique combines two or more phase-modulating flexure-beam mirror elements into a single macro-pixel. By manipulating the relative phases of the individual sub-pixels within the macro-pixel, the amplitude and the phase can be independently set for this filtering element. The combination of DMD sub-pixels into a macro-pixel is accomplished by adjusting the optical system resolution, thereby trading off system space bandwidth product for increased filtering flexibility. Volume in the larger dimensioned space, space bandwidth-complex axes count, is conserved. Experimental results are presented mapping out the coupled amplitude and phase characteristics of the individual flexure-beam DMD elements and demonstrating the independent control of amplitude and phase in a combined macro-pixel. This technique is generally applicable for implementation with any type of phase modulating light modulator.

Although the action of a spatial light modulator (SLM) is usually restricted to certain locations on the operating curve of the complex plane, NASA is planning to use architectures that allow two continuously variable SLMs to function jointly so as to access the full interior of a closed curve in the complex plane. This paper describes three fundamental methods for attaining full complex modulation. The mathematics for two of these methods is presented, and signal decomposition in their terms is outlined.

A feature-extraction-based optoelectronic neural network is introduced. The system implementation approach applies the principle of the neocognitron paradigm first introduced by Fukushima et al. (1983). A multichannel correlator is used as a building block of a generic single layer of the neocognitron for shift-invariant feature correlation. Multilayer processing is achieved by iteratively feeding back the output of the feature correlator to the input spatial light modulator. Successful pattern recognition with intraclass fault tolerance and interclass discrimination is achieved using this optoelectronic neocognitron. Detailed system analysis is described. Experimental demonstration of radar signature processing is also provided.

An experimental approach to establish a single channel holographic associative memory with bipolar neural states and bipolar interconnections is presented. The optical implementation of bipolar interconnections is fulfilled by biasing the interconnection weights and dynamically thresholding the output estimate. To give the optical neurons bipolar features, a proper distributed offset is added to the output. A holographic associative memory was established. To record the memory hologram, a pair of transparencies are prepared for each of the stored patterns. Each transparency bears an additional transparent part whose aperture size is determined by the difference of transparent and opaque parts of the corresponding stored pattern. Numerical simulations show that its performance is in perfect coincidence with an ideal bipolar memory. Preliminary experimental results are also shown.

This paper describes photorefractive self-pump phase conjugation in BaTiO3and multi-wave mixing in Bi12SiO20 spatial light modulators, and furtherly utilizes these systems to impletment optical image storage, wavelength conversion, image subtraction and optical image shifting. The operational principle and experimental results are also presented.