A review of the experimental and theoretical aspects of optical polarization is presented with definitions of the observed polarization characteristics and relationship to the Stokes parameters. A typical terrestrial soil polarization curve is characterized and related to the current theoretical knowledge. This polarization relationship is extended to cover planetary surfaces, such as the Moon, and Mars and terrestrial surfaces composed of farm areas and water surfaces. Instrumentation for imaging and non-imaging polarimetry are described including the use of focal plane arrays. Recent Space Shuttle polarimetric observations of the region around the Island of Hawaii and New Madrid, Missouri are described, as well as concurrent cloud and haze observations. Polarization is a sensitive indicator of cloud particle size distributions, soil texture, farm crops, sea state and atmospheric aerosols and haze. Cloud particle size distributions are uniquely characterized by polarization, and this cannot be achieved with photometry. An extensive bibliography of polarization in remote sensing is appended.

The measurement of temperature of the earth's surface from space is an important remote sensing parameter and depends upon the surface emissivity. Directional-hemispherical emissivities have been measured for several different soil samples with 10 micrometers CO₂ laser radiation. The CO₂ laser is a possible polarized source for active remote sensing. The angular variation of the directional-hemispherical emissivity has been calculated from directional reflectance measurements for horizontal and vertical polarized CO₂ radiation on different soil samples and they depended upon the polarization of the incident light. In this paper it is demonstrated that the directional-hemispherical reflectance, absorbance, and emissivity are 4 X 4 Mueller matrices. For uniform incident radiance of definite state of polarization incident on an area dA' within a projected solid angle d(Omega) ' equals Cos(Theta) 'd(omega) ' at angles ((Theta) ', (phi) ') Kirchhoff's formula relating the emissivity and reflectance involves definite sums and/or differences of Mueller matrix components of the reflectance and emissivity and depends on the polarization of the incident light. The directional-hemispherical emissivity of opaque soil samples are polarization dependent.

An improved surface reflectance model (SRM) for describing the polarimetric and photometric spectral polarized laser surface scattering is proposed. The model uses shadowing/obscuration and retroreflectance functions that are physically correlated to the surface. Results are compared to those obtained with the surface mode developed by the Environmental Research Institute of Michigan. The model under consideration is in an early stage of development.

Polarization and directionality of Earth reflectances (POLDER) is an optical imaging radiometer developed by CNES for flight on NASA's ADEOS mission (launch date 1995). It is a wide field of view (50 degree(s)) and moderate spatial resolution (6 km) instrument, giving an almost daily coverage of the Earth. It gives access to the bidirectional reflectance and polarization distribution functions (BRDFs and BPDFs) by acquiring, for a given target, measurements in variable viewing configurations along the track of the satellite. These measurements are acquired in nine spectral bands of the visible-near infrared spectrum, and at three polarization angles at 443, 665, and 865 nm. Its main scientific objectives are: (1) tropospheric aerosols -- determination of the characteristics of the aerosols; (2) ocean color -- accurate determination of the sea surface reflectances, due in particular to improved capabilities of atmospheric corrections; (3) land surface -- determination of surface BRDFs and improvement of the correction of surface bidirectional effects and atmospheric effects on vegetation indices; (4) Earth radiation budget and clouds -- determination of clouds BRDF and classification of clouds according to bidirectional radiative properties, but also on cloud altitude and phase. The objective of the paper is to present and discuss these mission objectives.

A numerical code has been developed to calculate Stokes parameters of the visible solar radiation, scattered in the atmosphere-ocean system. Mathematical modeling is used to examine spectral and angular (azimuth and zenith angle) variations of degree of polarization at sea level and at different heights in the atmosphere above the sea surface. On the basis of a developed computer code the efficiency of the polarization measurements for different optical passive remote sensing techniques of the ocean has been investigated. For the passive spectral measurements of the water bio-productivity (chlorophyll-a, dissolved organic matter, concentration of suspended particles) the polarizer can improve signal-to-background ratio. The magnitude of this effect and optimum direction of the polarizer depend upon height, viewing direction, and solar zenith angle. Within the framework of polarization remote sensing technique the influence of the observation height and viewing direction on the results of water turbidity measurements is investigated. Optimal viewing directions in such polarization passive remote sensing technique are discussed.

The Space Shuttle Polarization experiment began in 1984. The first data was collected by the crew of STS-51A during October of that year. Since that time polarization data has been acquired from six additional flights. The objective of the experiment was to test the feasibility of concept and the utility of polarized imagery of solar radiation reflected from Earth as acquired from low earth orbit. While this objective has been met, the preliminary results have been well documented in the open literature, and two Space Shuttle payload bay experiments, based in part on these findings are in the design phase, the data acquired have not been used to their full potential. The users of these data have been limited in number even though the images are inexpensive to acquire and are readily available in the public domain. To broaden their use, this paper summarizes the data available, scene content of the images, procedures used to acquire the experimental data, the attributes and shortcomings of the data and a description of uses to which they appear applicable. A catalog of images is now in preparation that should facilitate the application of these data.

During the four American Space Shuttle missions of 1985, the crewmembers took pictures of the Earth in polarized light. Selected images were digitalized and a first step was devoted to evaluate the quality of the data and to convert the numerical counts into radiances and degrees of polarization. For this last purpose, different techniques of in-flight calibration were analyzed and applied to suitable scenes. The analysis of the selected data first confirmed the previous observations over bright and homogeneous targets such as snow and sand. Indeed, we observe a predicted low polarization over these surfaces. Over land, the general situation is more confusing with equal contributions of the ground and of the atmosphere. Nevertheless, for the polarized reflectance, the atmosphere dominates outside of the specular direction while in this region the reflection on the surface is predominant around the Brewster angle. At last, during an atmospheric contamination by forest fires, the analysis of the degree of polarization provided a clear and accurate characterization of the aerosols.

This paper describes to relationship of polarimetric observations from orbital and aerial platforms and the determination optimum sun-target-sensor geometry. Polarimetric observations were evaluated for feature discrimination. The Space Shuttle experiment was performed using two boresighted Hasselblad 70 mm cameras with identical settings with linear polarizing filters aligned orthogonally about the optic axis. The aerial experiment was performed using a single 35 mm Nikon FE2 and rotating the linear polarizing filter 90 deg to acquire both minimum and maximum photographs. Characteristic curves were created by covertype and waveband for both aerial and Space Shuttle imagery. Though significant differences existed between the two datasets, the observed polarimetric signatures were unique and separable.

Polarimetry has become an indispensable analysis tool, in remote sensing as well as in radar metrology. In the same way, it is possible to characterize immersed targets illuminated by a laser beam, by using signal polarization. Here our purpose is to discriminate them in relation with their depolarizing and diffusing properties. Using the Stokes-Mueller formalism and Gil- Bernabeu depolarization criterion, we have synthesized two different types of images from which we build analysis methods. In this paper, we present different results for some characteristic samples.

Polarimetric images of the earth have been obtained for the first time by means of the Space Shuttle. These depict the environs of Hawaii in three visible bands of polarized light. A transect drawn on each image, crossing only cloud-free ocean areas, has both positive and negative polarization values along its path. These results are compared with those of several models, and with aircraft measurements of a haze-free, hydrosol-free ocean area, regarded as `air truth.' Based on the models and the air truth, the negative percent polarization values are attributed to haze, even though haze is not visible in a photograph taken from space. This haze obscures the effects of wind-generated surface roughness, and suspended hydrosols, so that they cannot be separately distinguished. The Hansen-Travis radiative transfer model predicts that the haze must have a large optical depth. The air truth shows that in the absence of haze, the wind velocity can be characterized by polarimetric measurements. The effect of hydrosols on polarization is not yet clear.

Tropospheric aerosols have potential climate forcing roles through both the direct effect on solar radiation scattered and absorbed and the indirect effect as condensation nuclei for cloud particles. Present capabilities for characterizing the global aerosol climatology are limited by the existing satellite remote sensing measurements, making a quantitative determination of the climate forcing effects of aerosols rather difficult and uncertain. The Earth Observing Scanning Polarimeter (EOSP) instrument under consideration for Earth Observing System (EOS) mission platforms and for a proposed Climsat Earth Probe mission is intended to provide global mapping with multispectral photopolarimetry, which can be exploited for the more pronounced aerosol signature in the polarization of the scattered light as compared to the radiance. We describe the essential features and performance specifications of the EOSP instrument and discuss approaches for retrieving tropospheric aerosol properties from multispectral photopolarimetry.

Ground-based measurements of the transmitted and scattered sunlight were performed during several campaign measurements, in desertic, oceanic, and continental sites. The measurements were mainly conducted to assess the capability of passive measurements to provide information about the aerosols and the capability of Mie theory to provide the convenient framework for the data analysis. Special attention was given to polarization measurements, as a known very sensitive test of scattering features. The results of different campaigns are presented and analyzed. The polarization capacity to provide comprehensive information about the atmospheric aerosol is discussed.

The POLDER instrument is designed to perform spectral and directional measurements of the polarized light reflected by the Earth ground-atmosphere system. The instrument is planned to be launched on the future Japanese ADEOS platform. An airborne version of POLDER has been built. As a preparation of the satellite mission, campaign measurements were conducted in 1990 and 1991, over land surfaces, in the south of France. During the flights of the POLDER instrument, ground-based measurements of the skylight were performed in order to derive the characteristics of the atmospheric aerosols. Aerosol remote sensing over land surfaces is known to be a difficult task, because ground reflectances are much larger than the aerosol ones. The POLDER instrument provides reflectance and also polarized reflectance. An interesting feature is that polarized reflectances prove to be mainly informative about atmospheric scattering. The POLDER measurements are compared with simulations of the measurements, based on radiative transfer calculations using as the input the aerosol model derived from the ground-based measurements. The good agreement between simulated and observed polarizations assesses the POLDER capability to detect aerosols over land surfaces.

This paper describes a recently completed electrooptical camera flying onboard the NASA ER-2 high altitude aircraft. The device includes a six-position filter wheel which can be fitted with a combination of polarizing and/or spectral filters. An alternate configuration will include a polarizing filter which can be rotated to any angle under computer control. The camera mount in the nose of the ER-2 can tilt forward or aft up to 40 degrees, both for bidirectional reflectance studies and for image motion compensation (the aircraft moves 34 meters between frame acquisitions). The ground resolution is nominally 5 meters from and altitude of 20 km. Spectral responsivity is that of the silicon imaging array (Kodak KAF-1400). Initial data sets were acquired in support of the International Satellite Cloud Climatology Program Regional Experiment of November, 1991, and will be used to study cirrus cloud properties.

The thermal activity of a seismically active area in Central Asia has been monitored remotely and with in situ sensors. The remote sensing in the 8 to 14 micron region was made using the NOAA series satellites AVHRR-2 and HCMM radiometers, with temperature sensitivity of 0.5 K. The surface thermal effects were of the order of 3 to 5 K. These temperature anomalies associated with the Kopet-Dag, Karatau and other deep faults which are the source of seismicity of adjacent crustal blocks. The thermal variations may precede, occur consistently or be subsequent to the seismic activity. Remote observations of thermal effects are dependent on surface ground layers as well as atmospheric effects. The ultimate objective of this research is to monitor seismic activity from space. Polarization of the radiation is proposed as a technique to enhance the remote sensing observation as has been accomplished in the visible region of the spectrum.

Sir David Brewster, Scottish physicist of the Nineteenth Century, was one of the pioneers in the investigation of the polarization of light. Every physics student is familiar with the Brewster angle of reflection, and the Brewster neutral point of skylight polarization is a well- known feature in atmospheric optics. He was at one time the most honored natural philosopher in Britain, having received numerous medals plus a knighthood for his work in the polarization of light. This paper, having arisen from my new biography of Brewster, traces his polarization work throughout his most productive period in the first half of the last century. It is of interest to science historians, as well as to those working in the field of polarization phenomena.

When unpolarized light is scattered by a rough surface, or by a cloud of gas, of particles or of crystals, it becomes partially linearly polarized, and the plane of the polarization is usually found to be either normal to the plane containing the incident and observation rays, or parallel to this plane. If the scattered light has intensities l and 12 polarized in planes normal and parallel to the ray plane, then the degree of polarization is defined as P = (Ii —12 )/(li +12 )' usually expressed as a percentage, or in parts per thousand. P is found to change with the angle V between the incident and observation rays, usually known in astronomy as the phase angle. A plot of P against V is found to give a curve characteristic of the surface. Typically, curves of polarization for a rough or particulate solid surface is as shown in Fig. 1 , which refers to telescopic observations of the Moon for several wavelengths. Many such curves have been obtained by telescopic observation of bodies in the Solar System. Extensive laboratory studies have also been developed and are available for interpretation of the observations. The bulk of these results can be found in the review papers and major publications listed in the references.

A remote minefield detection system (REMIDS) developed as part of the U.S. Army's Standoff Minefield System Research Program is presented. This helicopter-mounted system based on an active/passive line scanner, real-time processing, and display and navigational equipment obtains image data in three principal coregistered channels via line scanning. Two channels provide near-IR linear polarization reflectance vector information while the third channel provides passive thermal information. Numerous flight tests showed that the REMIDS system is capable of detecting mines during both day and night flight. Polarization is confirmed to be are a good discriminator between man-made and natural objects. Active polarization and reflectance information proved to be superior to thermal data in several natural scenarios including arid regions, overcast conditions, and diurnal thermal crossover periods.

A survey of retroreflection measurements is presented with specific emphasis on the opposition effect (or brightening of the moon) at full moon phase. This is followed by a discussion of laser retroreflectance on photometric standards and coatings, that reveals a peaking in the retrodirection even for accepted photometric standards. The retroreflectance property exists for all surfaces, and even for gross remotely sensed targets. The most recent retroreflectance measurements are presented showing that laser retroreflectance is a polarized single scatter-interference phenomena.

A compact, acousto-optic tunable filter (AOTF) imaging spectropolarimeter for ground based astronomy from 400-1100 nm has been constructed at NASA/GSFC. The key components of this instrument are a TeO₂ non-collinear AOTF, CCD camera, and an all-reflective optical relay assembly which uses a single elliptical mirror to produce side-by-side orthogonally polarized spectral images. The instrument was used at the Lowell Observatory 42-inch telescope for 'first light' planetary imaging and measurements of photometric standard stars. Narrow-band images of Saturn near 700 nm appear to show polarization effects which result from multiple scattering by aerosols. The instrument has recently been upgraded in order to integrate the RF drive electronics and eliminate contamination by scattered light. Design of the instrument and some initial results are presented.