An evaluation is made of the progress made to date on the application of radar polarimetry to the remote sensing of microphysical parameters via particle shapes, orientations, and compositions. Even without measuring the complete covariance or Mueller matrix, it is found possible to glean important quantitative and qualitative information, provided that the polarimetric estimators can be related to microphysical parameters. The complete covariance matrix approach is needed for particle identification at low temperatures.

The Advanced Detection Technology Program has as one objective the application of fully polarimetric, high-resolution radar data to the detection, discrimination, and classification of stationary targets. In support of this program, the Advanced Detection Technology Sensor (ADTS), a fully polarimetric, 35-GHz SAR with 1 ft by 1 ft resolution was developed. In April of 1989, the ADTS gathered target and clutter data near Stockbridge, NY. Data from this collection is being used to investigate optimal polarimetric processing techniques. This paper summarizes the results of a recent study of an optimal polarimetric method for reducing speckle in SAR imagery.

A problem originating in radar polarimetry is considered for which the radar target is to be characterized by its coherent polarization state properties, given complete coherent backscattering scattering matrix data sets at one frequency and for one target aspect angle. For the monostatic reciprocal case, it is shown that there exist in total five pairs of characteristic polarization states. These include the orthogonal cross-polarization null and copolarization maximum state pairs and the orthogonal cross polarization maximum state pairs, as well as a newly identified pair, the orthogonal cross-polarization saddle point extrema which are normal to the plane spanned by the other four pairs. With this complete and unique mathematical description of Huynen's polarization fork concept, it is now possible to study the polarimetric radar target optimization problem more rigorously.

Different experiments on remote sensing the environment by methods of active radiolocation showed that there arises a necessity for optimal processing received reflected signals. To illustrate this idea we shall give a part of experimental resuits received in sensing different meteorological objects by electromagnetic waves of mm length range of waves. The experiments were aimed at determining statistic characteristics of radiowaves reflected from different meteorological objects. At the same time changes in different polarization parameters of reflected electromagnetic wave (EMW) were investigated. The geometric parameters of reflected EMW polarization ellipsis, i.e. orientation angle of EMW polarization plane and 2 elliptic angle were chosen as polarization parameters to be determined. However, within the framework it was stated that elliptical angle of reflected EMW as well as its variations are close to zero.

At DLR (formerly DFVLR) in Oberpfaffenhofen, West Germany, a coherent polarimetric C-Band radar is in operation since three years for applications in microwave propagation (satellite communications) and cloud physics research. Its functional aspects are briefly described. The radar permits the characterization of scatter volumes consisting of statistically distributed targets by a time series of relative scattering matrices, the Stokes matrix and the covariance matrix. Radar scattering from extended meteorological targets is treated as a stationary ergodic stochastic process. The elements of the fundamental second moments (covariances) are discussed. Special consideration is given to the microphysical interpretation of polarimetric radar echoes using the scattering matrix and the covariance matrix approach. Polarization signatures for backscattered power and correlation observables are derived applying polarization transformations to the covariance matrix. Finally, the signal degradations on virtual Earth-satellite paths are studied with the aid of radar data for frequencies in the new bands (Ka-Band) for satellite communications.

An investigation is made of methods of averaging millimeter-wave (MMW) polarimetric data for subsequent target-identification/discrimination processing. The averaging is required because the radar cross-sections (RCSs) of the target vehicles, which rotate during the measurements, change rapidly with azimuth angle. The RCS in an individual polarimetric transmit-receive channel, the difference between the RCS in two channels, and the difference between the phase of the signals in two channels are subjected to averaging processes both in azimuth angle and radar-frequency spaces. The averages over the 640-MHz bandwidth of the radar are found to be significantly more effective in minimizing the mean spread in these quantities in a 1-deg scan than the averages of single-frequency data over a number of smaller-angle scans. Frequency averaging typically reduces the standard deviation in the measurements during a 1-deg scan of a complex target.

The discrimination of stationary ground targets from ground clutter with an airborne millimeter-wave radar sensor is discussed. Specifically, discrimination figures of merit, target and clutter statistics, radar resolution cell sizes, the use of polarization, and the probability of false-target density are discussed.

The problem of resolution improvement of various formations on the Earth surface, arising in remote sensing presupposes solving two main tasks. First, an increase in radar contrast of these formations on the Earth background. Second, the possibility of improving the selection of a useful signal out of its mixture with interference signal and the noise of the receiver in a radio receiving device. This paper is aimed at solving these two problems.

Applications of optical polarimetry are reviewed, and limitations of polarimetry are discussed in the context of information content and error management of polarimetry applications. Polarimetric measurement techniques are categorized as spectral or nonspectral, active or passive, monostatic or bistatic, and whether or not measurements of the complete Stokes vector or Mueller matrix are made. Specific examples from the literature of passive polarimetry, laser polarimetry, and spectropolarimetry are discussed.

An analysis of polarimetric measurements in the presence of noise is presented with emphasis on rotating polarizer polarimeters. A method is presented for estimating the error bars associated with a polarization measurement of an object whose intensity is rapidly fluctuating in a noisy manner. The method requires knowledge of the power spectral density of the intensity fluctuation, and uses this information combined with the sequence of polarizing element settings and measurement times to determine the error bar estimate. The rotating polarizer polarimeter is used as an example of this technique. A relationship is established between the power spectrum of the source intensity fluctuations and the error bars associated with measurements of the Stokes vector elements. The impulse response and transfer function of a polarimetric measurement and data reduction sequence are introduced. The analytical tools presented here will assist in designing polarimeters with specified accuracies for demanding roles, such as measuring polarization signatures in rapidly moving or fluctuating scenarios. It is recommended that for rotating polarizer polarimeters, the polarizer rotates through a minimum of 360o because image nutation from wedged polarizers and linear drift couple strongly into the fundamental frequency of the discrete Fourier transform. The optimum rotating polarizer angle settings for many applications is the eight angle sequence: 0°,45°, 90°, 135°, 180°, 225°, 270°, 315°.

The potential of infrared polarized signature components for suppressing background clutter and enhancing the detection of dim targets has been investigated. An imaging infrared polarimeter has been built, field experiments have been conducted, and measured results have been compared with theoretical predictions. A description of the measuring instrument, samples of empirical findings, and comparisons with simple theoretical predictions are presented.

Photopolarimetric observations are reported for three types of satellites with simple structures. The variation of the measured percentage of linear polarization as a function of the observed phase angle for each of the three types is shown and compared with model predictions. The models are based on specular reflections off surfaces represented by randomly oriented macroscopic facets. In the examples considered, the polarization verus phase angle curves imply a unique value for the effective complex index of refraction of the surface.

A technique to measure linear diattenuation and retardance spectra of IR samples in transmission is developed. A sample is rotated between two fixed linear polarizers in an IR spectrometer. The intensity modulation resulting from the rotation of the sample is Fourier analyzed and the diattenuation and retardation of the sample are calculated. For single-plate devices whose thickness is known, the birefringence spectrum can also be determined. This technique is utilized in the sample compartment of a Fourier transform IR spectropolarimeter to determine the linear diattenuation and retardance spectra from 2.5 to 16.5 microns of an IR multiple-order retarder, two achromatic retarders, and a liquid crystal sample.

Two configurations of imaging polarimeters are described that are designed for polaimetric optical metrology. The first is a Stokes imaging polarimeter which measures the polarization response of optical systems to spherical or planar waves of known polarization. The output is images of the degree of polarization, orientation, and eccentricity of polarization ellipses, or Stokes parameters displayed as a function of either the exit pupil or image coordinate of the optical system. The second configuration is a Mueller imaging polarimeter which measures the Mueller matrix of an optical system on a ray-by-ray basis. Calibration issues involved in building the instruments are addressed along with a brief discussion on polarization aberration mechanisms.

An automatic effipsometer that uses a four-detector photopolarimeter (FDP) to measure the state of polarization of light reflected from an optically isotropic or aniotropic, depolarizing or nondepolarizing, sample surface is described. For conventional effipsometry on specular isotropic surfaces, the incident light is polarized (at least partially) in one stable state (e.g., using a linear polarizer which need not be perfect), and the reflectance and ellipsometric parameters of the surface are encoded onto, hence can be retrieved from, the Stokes parameters of the reflected light. The latter are measured, virtually instantaneously, by the FDP. The FDP also greatly simplifies generalized Mueller-matrix ellipsometry on amisotropic or nonspecular surfaces. In this case, the polarization of the incident light is controlled by a linear polarizer followed by a quarter-wave retarder (QWR) with rotationally adjustable fast axis azimuth C. Fourier analysis of the output current vector of the FDP as a function of C yields a series of five terms whose vectorial coefficients determine the Mueller matrix column-by-column. In such an analysis, the small inevitable imperfections of the QWR are accounted for. Results are presented for measurements at 633 nm wavelength of the ellipsometric parameters of a Au surface and the zeroth-order Mueller matrix of a 1200 G/mm Al-coated holographic grating for three different orientations of the grooves relative to the plane of incidence.

This paper describes how FOV errors have affected the polarization measurements of the NASA Marshall vector magnetograph and how a polarimeter has been modified to minimize these errors. The design of a large-FOV (LFOV) KD*P modulator is discussed. The Poincare sphere is used to visualize the function of the different optical elements that make up the LFOV modulator.

Three stress effects on the state of polarization in high birefringent polarization-maintaining fibers are reviewed. The stresses induced in PM fibers by the different thermal expansion coefficients between the stress-applying parts and the surrounding glass make stress-induced-birefringent fibers very sensitive to temperature variations. Mechanical perturbations can change the polarization state of PM fibers. Temperature changes can enhance any random microbends in wound optical fiber and cause mode coupling. The effect of periodic bends, small-radius bends, and twist and fiber individual crossing in the presence of transverse pressures on the degradation of the state of polarization rotation is investigated. A procedure for evaluating the performance of wound PM fibers in low temperature environments is also discussed.

The on- and off-axis polarizing properties of asymmetric retroreflectors are studied in detail for various angles of incidence and for various incident linear polarization states. An analytic model is developed by applying Fresnel law to the incident and reflecting radiation on each facet of the retroreflector. It is shown that the polarization state of retroreflected radiation is a sensitive function of incident angle, incident polarization rate, and retroreflector material. These characteristics may be applicable to the determination of the relative angular position between the retroreflector and the analyzer.

Longitudinal KD*P modulators are used in ground-based solar magnetographs to eliminate seeing effects. Although the modulators can be used as variable retarders, the optical properties when zero voltage is applied influences the performance on instruments requiring very accurate polarization measurements. Measurements at the Marshall Space Flight Center are discussed in terms of the optical properties of KD*P modulators when zero voltage is applied. The measurements can be used to predict the modulation characteristics of the devices and to determine the polarization accuracy that can be expected from the vector magnetograph.

The Marshall Space Flight Center (MSFC) vector magnetograph records polarization images of absorption lines that are sensitive to magnetic fields. A method is presented for analyzing the Stokes spectral-line profiles of a photospheric Fe I absorption line (5250.2 A) which is influenced by the Zeeman effect. Using nonlinear least-square optimization, the observed Stokes profiles are compared with those generated from the theoretical solution of the polarized radiative transfer equations. The optimization process accounts for the spectral convolution of the source and the MSFC vector magnetograph. The resulting physical properties of the active region producing the polarized light are discussed.

X-ray scattering from a lithium disk and Bragg reflection from a mosaic graphite crystal can be exploited to measure the linear polarization of radiation emitted from cosmic X-ray sources. The sensitivity is enhanced if the polarimeters are placed at the focus of an X-ray telescope. Such devices form two of the components of the Stellar X-ray Polarimeter experiment scheduled to fly on the Spectrum-X-Gamma mission. The expected on-axis performance of the two components is described based on detailed Monte Carlo simulations. The polarimetry experiment is expected to provide sensitive measurements of linear polarization for many cosmic X-ray sources. The nature and utility of such observations is described for pulsing X-ray sources such as the Crab pulsar and Her X-1.

A polarimeter for characterization of the instrumental polarization of an imaging UV spectrometer has been designed. In the course of calibrating the polarimeter, the quarter-wave retarders were observed to possess polarization properties other than pure linear retardance. This forced the development of a complex procedure for the calibration of the retarders and a new generalized approach to polarimetric analysis.

The Advanced X-ray Astrophysical Facility (AXAF) telescope consists of six concentric paraboloid-hyperboloid pairs of mirrors operating near grazing incidence. Because of the substantial polarization effects at large angles of incidence, there is concern about the feasibility of polarimetry near the focal plane. The primary mirror acts as a tangentially directed half-wave linear retarder and nearly completely depolarizes the linearly polarized component of the light. The secondary mirror introduces an additional half wave of linear retardance. The tangentially directed one-wave linear retarder leaves the transmitted beam in the incident polarization state. The net instrumental polarization effects are small, and polarimetry is feasible with the AXAF.

A rotating analyzer ellipsometer for the far-infrared wavelength range from 30 to 1200 microns has been developed. The instrument uses well-collimated radiation from a step-tunable CW gas laser source. Wire-grid polarizers are employed which give a polarization purity of 10,000:1. To test the instrument, the birefringence of crystal quartz was measured in transmission. Applications are in the characterization of semiconductors and of high-Tc superconductors.

A system of photopolarimeters to measure the Mueller scattering matrices for either reflective or transparent scattering medium between 0.14 to 12 micrometers will be discussed. The system is based on the photopolarimeters suggested by Azzam and the phase modulation of the incident light by photoelastic or Pockel cell modulators. In the vacuum uv the system is complicated for it must be designed to operate in a vacuum. The technique will allow the analyzation of the data by Fourier techniques.

A comprehensive overview of the basic principles of radar polarimetry is presented. The relevant fundamental field equations are first provided in order to introduce the polarization state formulations of electromagnetic waves in the frequency domain, including the Jones and the Stokes vector formalism and its presentation on the Poincare sphere and on relevant map projections. In the next step, the scattering matrices (S) and (M) are given together with change of polarization bases transformation operators, and the optimal polarization states are determined for the coherent and partially coherent cases.

We are developing an interactive computer program which includes many of the possible options available for the design of a polarimeter, for the analysis of the output signal, and for the computation of the polarization of the source under investigation. The software is being implemented on a microcomputer and is operated interactively. It utilizes pull-down menus, dialog boxes, graphics and mouse-driven point and click techniques to ease user interaction. We have addressed the issue of how to represent the polarimetric model in software by manipulating optical element transformation matrices (Mueller matrices) to produce the resultant polarimeter matrix. To keep the polarimetric model general and to facilitate analytic evaluations of polarimeter designs we represent these matrices in symbolic form in the software. The software tool will have wide ranging applications in areas of engineering, research and education dealing with polarimetry. It can be used as a design tool for polarimeters in laboratories, remote sensing, or both ground- and space-based astronomy. It can also be used as a simulation tool for polarimetric measurements which will be useful for evaluating polarimeter designs or for educational purposes. The software contains a data reduction tool which can be used to evaluate intensity measurements made by a specified polarimeter.

Photographs of corner cube retroreflectors using a polarimeter and collimated laser illumination have verified their polarization properties.

The bireflectance effect is observed when a light beam is internally reflected in an anisotropic material. Generally, an unpolarized light beam splits into four beams on reflection. A new method for simultaneous measurement of the three refractive indices, in a single piece of a sample, via bireflectance, is demonstrated in the biaxial CdWO₄ crystal. The three principal refractive indices of CdWO₄ were obtained in the visible range. At λ = 633 nm, n_X = 2.09±0.01, n_Y - n_X = 0.052±0.001, and n_Z - n_X = 0.114±0.002.

There is a vast difference between the use of the target-scattering matrix SM and the Stokes (or Mueller) matrix MM. The SM contains essentially target information in terms of EM-field components, whereas the MM consists of terms which are correlations of these. The SM, although useful to provide the input data, has the disadvantage of being absolutely phase-dependent, whereas the MM has only parameter interphase (relative phase) dependencies. This makes it possible to interpret independent 'objects' which are added incoherently, just as is the case with objects of perception. The object approach is further enhanced through the interpretation of MM parameters in terms of unique physical properties of radar target classes.