ALBERT

Description: Forest canopies over different sedimentary lithologies of valleys and ridges are composed of different dominant species and have significantly different reflectance and emittance. In a botanical survey of eighty-seven forest sites, sedimentary lithologies were found to differ in the species which dominate the forest canopy. Sandstone sites had abundant chestnut oak (Quercus prinus), black oak (Q. velutina), and northern red oak (Q. rubra). On shale sites chestnut oak, white oak (Q. alba), northern red oak and red pine (Pinus resinosa) were dominant. Limestone sites had a variable species composition with the most common species, northern red oak, white oak, and black locust (Robinia pseudocacia) accounting for only 30 percent of the total trees. Thematic Mapper Simulator (TMS) data obtained during the growing season were analyzed to determine if sandstones, shales, and limestones could be distinguished on the basis of forest-canopy reflectance. The observations compared in the analysis were means of the eight TMS bands for 10 x 10-pixel test sites selected from areas with complete canopy closure. In August imagery the three lithologies were separable based on differences in TMS band 3 (630-690 nm) and band 8 (10.4-12.5 microns).

Description: The use of remote-sensing data on forest leaf flush to identify areas with anomalously high soil heavy-metal concentrations is demonstrated using airborne Thematic Mapper Simulator (TMS) leaf-area-index data obtained over two sites in Virginia in spring 1983 and 1984. Mean-reflectance differences, especially in the 760-900-nm and 630-690-nm bands, corresponding to delayed leaf flush are found to be good indicators of higher heavy-metal concentration. Airborne and ground-based canopy-temperature measurements are also shown to be significantly higher in high-heavy-metal areas than in control areas.

Description: This paper continues a study on the accuracy of geological mapping using Landsat Thematic Mapper data (Short, 1984). In June 1976, both the White Mountain alteration zone and the Waterpocket Fold sedimentary rock sites in Utah were surveyed by the Bendix 24-band scanner on a NASA NC-130B aircraft. Mid-June 1984 TM data for these two sites have been processed like the 1976 data to test the quality of simulation of TM data. Principal-components (PC) color composite images for White Mountain show close correspondence to the Bendix PC images. At this site carbonate strata are uniquely discriminated in both Bendix and TM composites that use an inverted PC 3 image. Alunite/kaolinite and hematite/limonite alteration zones developed on volcanic flows are also sharply separated, but iron oxide and silicified zones are less so. The accuracy of rock-units mapping at the Waterpocket Fold site by supervised classification of the June TM data is significantly better, reaching 70 percent in the best case, than for January 1983 data for that site.

Description: A geobotanical investigation based on the detection of premature leaf senescence was conducted in an area of predominantly chalcocite mineralization of the Keweenaw Peninsula in Michigan's Upper Peninsula. Spectrophotometric measurements indicated that the region from 600 to 700 nm captures the rise in red reflectance characteristic of senescent leaves. Observations at other wavelengths do not distinguish between senescent and green leaves as clearly and unequivocably as observations at these wavelengths. Small format black and white aerial photographs filtered for the red band (600 to 700 nm) and Thematic Mapper Simulator imagery were collected during the period of fall senescence in the study area. Soil samples were collected from two areas identified by leaf senescence and from two additional sites where the leaf canopy was still green. Geochemical analysis revealed that the sites characterized by premature leaf senescence had a significantly higher median soil copper concentration than the other two areas.

Description: Airborne Imaging Spectrometer (AIS) data were analyzed to deduce plant density and species composition in three semi-arid shrub-dominated communities of Owens Valley, CA, occurring on either a sand, granite alluvium, or basalt substrate. The high-spectral resolution AIS data were related to spectra obtained with field portable spectrometers, which in turn were related to plant and soil characteristics of the communities. Many of the dominant species have unique spectral features which permit their identification in AIS pixel images. The canopy-induced shadow may be a major factor influencing substrate spectral properties during fall and winter, because of low sun angles. Moreover, changes in spectral signatures following dormancy and leaf senescence tend to decrease contrasts between the plant community and the geologic substrate, also suggesting that fall and winter are a difficult time of year for spectral analyses.

Description: A preliminary interpretation of structure and lithology from selected Shuttle Imaging Radar-B (SIR-B) images of Borneo, collected in October 1984, is presented. The SIR-B images, obtained at depression angles that ranged from 40 to 50 deg, were interpreted by using the approaches suggested by Sabins (1983). On the basis of radar signatures, six terrain categories; coastal and alluvial plains, and carbonate, clastic, volcanic, and melange, rocks, were defined in east, central, and south Kalimantan, and in the Malaysian state of Sarawak.

Description: Landsat Thematic Mapper (TM) data collected in the late summer, fall, and winter of 1982 over forested bedrocks in southeastern Missouri were used in conjunction with forest surveys, field work, aerial photographs, and laboratory analyses to evaluate multispectral and seasonal information from visible and reflected IR data. The forested bedrock included granites, rhyolites, carbonates, and sandstones. High reflectance in band 4 (760-900 nm) in the summer scene corresponds to regions of xeric forest type. The fact that the xeric regions tend to develop flat-topped canopies, as opposed to irregular canopy surfaces of the wetter mesic areas, may partially control the TM response in bands 4, 5 (155-175 nm) and 7 (208-235 nm). The xeric regions correlated with soils having poor water retention capabilities, such as rhyolites and certain carbonate rocks with nonporous residum layers. An opposite relationship between xeric and mesic forest biomass was noted, if the commonly used TM band ratio 4/3 was used as a surrogate biomass measure. The high band 4 response over xeric forests gives anomalously high biomass estimates.

Description: State of the art applications of remote sensing in geological exploration programs are discussed along with research and development activities aimed at increasing the future capabilities of this technology for exploration geology. The topics considered include: technical issues in the state of the art; regional exploration models; remote sensing applications for hydrocarbon exploration; commercialization of remote sensing satellites; and data integration. Also addressed are: remote sensing applications for mineral exploration; geobotanical and environmental remote sensing; image processing techniques and applications; advanced sensors, radar, and airborne systems; and engineering, logistics, and marine applications.

Description: Examples of photogrammetric measurements made with video cameras uncorrected for electronic and optical lens distortions are presented. The measurement and correction of electronic distortions of video cameras using both bilinear and polynomial interpolation are discussed. Examples showing the relative stability of electronic distortions over long periods of time are presented. Having corrected for electronic distortion, the data are further corrected for lens distortion using the plumb line method. Examples of close-range photogrammetric data taken with video cameras corrected for both electronic and optical lens distortion are presented.

Description: Multiple incidence angle SIR-B data of the Cordon la Grasa region of the Chubut Province of Argentina are used to discriminate various forest types by their relative brightness versus incidence angle signatures. The region consists of several species of Nothofagas which change in canopy structure with elevation, slope, and exposure. In general, the factors that appear to impact the radar response most are canopy structure, density, and ground cover (presence or absence of dead trunks and branches in particular). The results of this work indicate that (1) different forest species, and structures of a single species, may be discriminated using multiple incidence angle radar imagery and (2) it is essential to consider the variation in backscatter due to incidence angle when analyzing the comparing data collected at varying frequencies and polarizations.

Description: The images obtained by the Shuttle Imaging Radar (SIR)-A and -B systems over the southwestern Egypt and northwestern Sudan were coregistered with the Landsat images and the existing maps to aid in extrapolations of the buried paleodrainages ('radar rivers'), first discovered by SIR-A. Field observations explain the radar responses of three types of radar rivers, RR-1 (broad, aggraded valleys filled with alluvium), RR-2 (braided channels inset in the RR-1 valleys), and RR-3 (narrow, long, bedrock-incised channels). A generalized model of the radar rivers, based on field studies and regional geologic relations, shows inferred changes in river regimen since the large valleys were established during the later Paleogene-early Neogene. It is suggested that a former Trans-African master stream system may have flowed from headwaters in the Red Sea Hills southwestward across North Africa, discharging into the Atlantic at the Paleo-Niger delta, prior to the Neogene domal uplifts and building of volcanic edifices across the paths of these ancient watercourses.

Description: Synthetic Aperture Radar (SAR) images from the National Aeronautics and Space Administration's Shuttle Imaging Radar-B mission were used to analyze the effects of radar incidence angle on information content and vegetation penetration. Three SAR data sets using incidence angles of 26, 46, and 58 deg were acquired over the mangrove jungles of Southern Bangladesh. The data sets were digitally processed using 3 x 3, 7 x 7, and 11 x 11 spatial filters and geometrically registered to a multisource-multilevel-corraborative data base consisting of Landsat data, forest map data, and in situ acquired forest enumeration and topographic information. Analyses revealed that significant vegetation 'penetration' was found at all angles, and that tree and canopy structural morphology may exert an influence on this phenomenon.

Description: An experiment was conducted from an L-band SAR aboard Space Shuttle Challenger in October 1984 to study the microwave backscatter dependence on soil moisture, surface roughness, and vegetation cover. The results based on the analyses of an image obtained at 21-deg incidence angle show a positive correlatlion between scattering coefficient and soil moisture content, with a sensitivity comparable to that derived from the ground radar measurements reported by Ulaby et al. (1978). The surface roughness strongly affects the microwave backscatter. A factor of two change in the standard deviation of surface roughness height gives a corresponding change of about 8 dB in the scattering coefficient. The microwave backscatter also depends on the vegetation types. Under the dry soil conditions, the scattering coefficient is observed to change from about -24 dB for an alfalfa or lettuce field to about -17 dB for a mature corn field. These results suggest that observations with a SAR system of multiple frequencies and polarizations are required to unravel the effects of soil moisture, surface roughness, and vegetation cover.

Description: During the SIR-B mission in October 1984, a significant number of overlapping synthetic aperture radar (SAR) images of various ground areas was collected. This has offered the first opportunity to perform stereo analyses on images from space that cover large ground areas to determine elevation information. This paper presents the preliminary results of an investigation to obtain elevation data from stereo pairs of SIR-B images. First, the accuracy with which elevation information can be derived from SIR-B image pairs is evaluated theoretically. It is shown that elevation accuracy is a function of the slant range resolution, the incidence angles with which the stereo pair is obtained, the accuracies in spacecraft state estimation, and determination of corresponding pixels in the stereo pair. Next, a hierarchical method is developed to match the corresponding pixels. This method involves iterative removal of local distortions and correlations of pairs of local neighborhoods in the two images. Since it is necessary to perform the matching at every pixel in the image, it is very computationally intensive. Therefore, it has been implemented on the Massively Parallel Processor (MPP) at the Goddard Space Flight Center (GSFC). The MPP's speed permits two iterations of this technique to operate on a pair of 512 x 512 images within 7 s. Results of applying this algorithm of SIR-B images of Mount Shasta, CA, are shown. The matching algorithm performs well in regions of the image with significant features. An approximate elevation image derived from the matching process corresponds to published topographic map data, except for certain obvious discontinuities.

Description: On October 5, 1984, the second Shuttle Imaging Radar (SIR-B) was launched into orbit aboard the Space Shuttle Challenger. SIR-B is part of an evolutionary radar program designed to progressively develop a multifrequency, multipolarization synthetic aperture radar with a variable earth-imaging geometry. The SIR-B instrument is an upgraded version of SIR-A, with the additional capability of tilting the antenna mechanically to acquire imagery at variable incidence angles ranging from 15 to 60 deg. The variable look angle capability provided a means of acquiring multiple incidence angle imagery over specific targets on successive days of the mission. These data are being used to classify surface features by their backscatter signatures as a function of incidence angle and for topographic mapping. In addition to the antenna tilt capability, a digital data-handling system was added to increase the dynamic range, the resolution was improved by a factor of two over SIR-A, and a calibration subsystem was added to improve the radiometric accuracy of the data. The mission had a number of problems, including loss of the primary digital data path between the Shuttle and the ground. In spite of these problems, approximately 20 percent of the planned digital data were collected over the 8-day Shuttle mission corresponding to an areal coverage of about 6.4 million sq km.

Description: Sample images obtained with the Shuttle Imaging Radar A (SIR-A) are presented, along with design and performance features of the SIR-A, Seasat and Landsat images of the same scenes for comparison purposes. The SIR-A functions at the L-band 25 cm at a frequency of 1278 GHz with a spectral bandwidth of 6 MHz. The images were taken at an angle of 47 deg and furnished a resolution of 40 m from an altitude of 259 km. The images covered a ground swath 50 km wide. The images are provided to assist in the development of effective techniques for interpreting radar imagery. The SIR-A instrument is a precursor of another imaging device which will be flown around Venus. The images provided include sections of France, Sardinia and Algeria.

Description: The use of the Thermal Infrared Multispectral Scanner (TIMS) data in agricultural landscapes is discussed. The TIMS allows for narrow-band analysis in the 8.2-11.6 micron range at spatial resolutions down to 5 meters in cell size. A coastal plain region in SE Alabama was studied using the TIMS. The crop/plant vigor, canopy density, and thermal response changes for soils obtained from thermal imagery are examined. The application of TIMS data to hydrologic and topographic issues, inventory and conservation monitoring, and the enhancement and extraction of cartographic features is described.

Description: Preliminary results of analyzing digital radar imagery data obtained by the SIR-B aboard the Space Shuttle Challenger STS 41-G are presented. The data cover 5 million square kilometers of the earth surface between 57 deg north and south latitudes. Radar imagery of the same target at different incidence angles was used to classify surfaces by their backscatter response as a function of incidence angle. The SIR-B proved to be useful for collecting multiple incidence angle data sets over a broad range of targets, providing information in the areas of geology, archeology, forestry, agriculture, oceanography, geography, and hydrology. The analysis is also used to optimize radar parameters such as look angle for future missions.

Description: This article describes a separability measure for class discrimination. This measure is based on the Fisher information measure for estimating the mixing proportion of two classes. The Fisher information measure not only provides a means to assess quantitatively the information content in the features for separating classes, but also gives the lower bound for the variance of any unbiased estimate of the mixing proportion based on observations of the features. Unlike most commonly used separability measures, this measure is not dependent on the form of the probability distribution of the features and does not imply a specific estimation procedure. This is important because the probability distribution function that describes the data for a given class does not have simple analytic forms, such as a Gaussian. Results of applying this measure to compare the information content provided by three Landsat-derived feature vectors for the purpose of separating small grains from other crops are presented.

Description: The use of radar polarization diversity for discriminating forest canopy variables on airborne synthetic-aperture radar (SAR) images is evaluated. SAR images were acquired at L-Band (24.6 cm) simultaneously in four linear polarization states (HH, HV, VH, and VV) in South Carolina on March 1, 1984. In order to relate the polarization signatures to biophysical properties, false-color composite images were compared to maps of forest stands in the timber compartment. In decreasing order, the most useful correlative forest data are stand basal area, forest age, site condition index, and forest management type. It is found that multipolarization images discriminate variation in tree density and difference in the amount of understory, but do not discriminate between evergreen and deciduous forest types.

Description: Recent advances in digital data acquisition and signal processing technology permit simultaneous measurement of the complex (amplitude and phase) radar backscatter from several polarization-diverse antennas. While absolute phase mesurements remain to be analyzed in detail. The differential phase of signals polarized parallel and perpendicular to the plane of incidence provide information on the scattering mechanisms that dominate the interaction of the radio waves with the terrain. Analysis of phase backscatter maps from a typical urban area yields a bimodal distribution with the two peaks separated by approximately 180 degrees, highly indicative of a dominant simple geometric one bounce-two bounce mechanism. Some maps of agricultural areas exhibit a similar distribution, however, other agricultural areas yield a distribution that, while still bimodal, consists of two peaks separated by about 110 deg. Still other agricultural areas exhibit a more complex distribution. All of the observed phase shifts appear to be independent of incidence angle from at least 20 deg to 55 deg, therefore the 110 degree shifts are inconsistent with both the geometric model used for the urban area and with common dielectric slab models.

Description: A simple way to build up a library of models which may be used to distinguish between the different types of vegetation and ground surfaces by means of their backscatter properties is presented. The curve of constant power received by the antenna (Gamma sphere) is calculated for the given Stokes Scattering Operator, and model parameters are adopted of the most similar library model Gamma sphere. Results calculated for a single scattering model resembling coniferous trees are compared with the Gamma spheres of a model resembling tropical region trees. The polarization which would minimize the effect of either the ground surface or the vegetation layer can be calculated and used to analyze the backscatter from the ground surface/vegetation layer combination, and enhance the power received from the desired part of the combination.

Description: The use of Suits' (1972a, b) digital radiative transfer model to simulate the effect of nonLambertian canopy reflectance on off-nadir observations of vegetation is discussed. Canopy reflectances of cord grass are calculated using the radiative transfer model, field radiometric measurements, and airborne multispectral scanner data. The effects of varying view angles on canopy reflectance are analyzed and compared. The comparison reveals that the model is effective in simulating the sense and magnitude of reflectance change due to variable angles of observations; however, the model does not reproduce the observed dependence of nadir canopy reflectance on solar zenith angle. It is concluded that the radiative transfer model is applicable for predicting the variation in canopy reflectance due to changing view zenith angles.

Description: Techniques have been developed for the sub-pixel location of control points in satellite images returned by the Voyager spacecraft. The procedure uses digital imaging data in the neighborhood of the point to form a multipicture model of a piece of the surface. Comparison of this model with the digital image in each picture determines the control point locations to about a tenth of a pixel. At this level of precision, previously insignificant effects must be considered, including chromatic aberration, high level imaging distortions, and systematic errors due to navigation uncertainties. Use of these methods in the study of Jupiter's satellite Io has proven very fruitful.

Description: The snow parameters affecting sensor responses at different wavelengths are discussed. The effects of snow depth and background radiation on gamma ray sensors and of crystal size, contaminants, snow depth, liquid water, and surface roughness on visible and near-infrared sensors are considered. The influence of temperature, crystal size, and liquid water on thermal infrared sensors and of liquid water, crystal size, water equivalent depth, stratification, snow surface roughness, density, temperature, and soil condition on microwave sensors are addressed.

Description: Structural features of ecosystems, such as leaf area index, phytomass and canopy chemical contents, are beginning to be estimated from remotely sensed data. This development, in combination with ecological modeling, is permitting the estimation of functional features of ecosystems including primary productivity and nutrient cycling. Such techniques are also being applied to the problem of monitoring the effects of air or water pollutants on biota. Sensors that obtain data at a coarse spatial scale (1 km2 or more) are also permitting the observation of biospheric patterns at a large regional or global scale for the first time. When coupled with atmospheric measurements, field data and simulation models, such data may serve to address ecological processes, including pollution effects, at large regional or global scales.

Description: Normalized 18-GHz microwave brightness temperatures, T(B), and a vegetation index determined from satellite radiometer data are combined with climatically modeled surface moisture estimates to constrain a simple physically based soil moisture model. It is found that the normalized T(B) values correlated well with soil moisture when the data were segregated by vegetation index range, but less so when all the data were combined. By using the vegetation index parameter, the model is shown to account for about 70 percent of the variability in modeled surface soil moisture.

Description: A simple equation relating the Microwave Polarization Difference Index (MPDI) and the Normalized Difference Vegetation Index (NDVI) is proposed which represents well data obtained from Nimbus 7/SMMR at 37 GHz and NOAA/AVHRR Channels 1 and 2. It is found that there is a limit which is characteristic of a particular type of cover for which both indices are equally sensitive to the variation of vegetation, and below which MPDI is more efficient than NDVI. The results provide insight into the relationship between water content and chlorophyll absorption at pixel size scales.

Description: This study evaluates the potential of measuring/mapping forest damage in spruce-fir forests in the Green Mountains of Vermont and White Mountains of New Hampshire using Landsat Thematic Mapper (TM) data. The TM 1.65/0.83-micron (TM5/4) and 2.22/0.83-micron (TM7/4) band ratios were found to correlate well with ground-based measurements of forest damage (a measure of percentage foliar loss) at 11 spruce-fir stands located on Camels Hump, a mountain in northern Vermont. Images using 0.56 and 1.65-micron bands with 1.65/0.83-micron band ratios indicated locations of heavy conifer forest damage. Both 1.65/0.83 and 2.22/0.83-micron band ratios were used to quantify levels of conifer forest damage among individual mountains throughout many of the Green and White Mountains. Damage was found to be consistently higher for the Green than the White Mountains.

Description: An airborne pulsed laser system was used to obtain canopy height data over a southern pine forest in Georgia in order to predict ground-measured forest biomass and timber volume. Although biomass and volume estimates obtained from the laser data were variable when compared with the corresponding ground measurements site by site, the present models are found to predict mean total tree volume within 2.6 percent of the ground value, and mean biomass within 2.0 percent. The results indicate that species stratification did not consistently improve regression relationships for four southern pine species.

Description: The use of remote sensing data to monitor and analyze the arctic environment is examined. Landsat MSS, TM simulated, NS001, Seasat, and airborne radar are employed to investigate the Strand and Dune areas on the Arctic Coastal Plain in Alaska. The Strand area contains landforms associated with permafrost and the Dune area is dominated by eolian deposits consisting of large longitudinal dunes. The remote sensing data are compared to baseline geomorphic maps derived from aerial photography. It is observed that the multispectral data are better than the radar data for the detection and recognition of arctic landforms, and the NS001 data provided the highest spatial resolution and correlated well with the high-altitude aerial photography.

Description: Over large areas in the tropics, satellite imagery is the principal source of data on the area, current stature, and extent of disturbance of the forests. The information from imagery that covers large areas at low resolution is greatly enhanced when different types of imagery can be compared. The paper presents a comparison of data from Landsat MSS and from the Shuttle Imaging Radar (SIR-A) L band HH polarization data for sites in the Amazon Basin. Results indicate that SIR-A backscatter from the undisturbed forest was lower than that from some disturbed areas and from flooded forests and that SIR-A brightness, increases nonlinearly with the Landsat normalized difference vegetation index. It is hypothesized that the brightest radar returns in southern Amazonia are from newly cleared forests that are littered with standing and fallen tree boles that function as corner reflectors; and that backscatter will diminish from disturbed areas over time as fields are burned repeatedly.

Description: This study evaluated the utility of data collected by the high-spectral resolution airborne imaging spectrometer (AIS-2, tree mode, spectral range 0.8-2.2 microns) and the broad-band Daedalus airborne thematic mapper (ATM, spectral range 0.42-13.0 micron) in assessing forest decline damage at a predominantly Scotch pine forest in the FRG. Analysis of spectral radiance values from the ATM and raw digital number values from AIS-2 showed that higher reflectance in the near infrared was characteristic of high damage (heavy chlorosis, limited needle loss) in Scotch pine canopies. A classification image of a portion of the AIS-2 flight line agreed very well with a damage assessment map produced by standard aerial photointerpretation techniques.

Description: Visible IR Intelligent Spectrometer (VIRIS) reflectance data have been found to have similar features that are related to air-pollution-induced forest decline and visible damage in both the red spruce of Vermont and the Norway spruce of Baden-Wuerttemberg; the similarity suggests a common source of damage. Spectra of both species include a 5-nm blueshifting of the red-edge inflection point, while pigment data for both species indicate a loss of total chlorophylls. The blue shift of the chlorophyll absorption maximum, as well as the increased red radiance and decreased near-IR radiance of the damaged spruce, may be used to delineate and map damage areas.

Description: Recent advances in imaging spectroscopy for remote sensing applications are discussed, reviewing the results of recent investigations. The advantages offered by the higher spectral resolution of imaging spectroscopy relative to scanners such as Landsat MSS and TM are explained; the design and performance of the Airborne Imaging Spectrometer (Vane et al., 1984) are described and illustrated with drawings, photographs, and sample images; data processing and analysis techniques are outlined; and applications to geological and botanical research are considered.

Description: The results of the NASA Landsat Image Data Quality Analysis (LIDQA) program are reviewed. Landsat-4 and Landsat-5 TM data quality with regard to image geometry and radiometry are discussed. The results indicate that the TM provides excellent imagery that can be used in the form of satellite image maps meeting cartographic standards at scales of 1:100,000 or smaller. These data can be used to locate features or guide the revision or updating of maps for scales up to 1:24,000. The TM sensor is also providing data of good radiometric quality and stability, with radiometric uncertainties of 1 percent or smaller. The temperature dependence in the absolute radiometry is on the order of 1 to 5 percent of full scale. In terms of bidirectional reflectance estimated at the satellite, the error is estimated at under 6 percent and commonly 3 percent. Preliminary results also corroborate the utility of the TM data for geological or geographical studies.

Description: Because of the considerable cost of an interactive multispectral image processing system, an evaluation of a prospective system should be performed to ascertain if it will be acceptable to the anticipated users. Evaluation of a developmental system indicated that the important system elements include documentation, user friendliness, image processing capabilities, and system services. The criteria and evaluation procedures for these elements are described herein. The following factors contributed to the success of the evaluation of the developmental system: (1) careful review of documentation prior to program development, (2) construction and testing of macromodules representing typical processing scenarios, (3) availability of other image processing systems for referral and verification, and (4) use of testing personnel with an applications perspective and experience with other systems. This evaluation was done in addition to and independently of program testing by the software developers of the system.

Description: The elevation gradients affecting tropical forest stand characteristics are presently studied in light of multipolarization airborne SAR data. A 'rubber sheeting' computer code was used to georeference the SAR data sets to the digital elevation data. The TOPO code from NASA's NSTL generated the terrain slope and aspect angle data from the terrain elevation data set; computed local incidence angles were used to delete those data areas that were shadowed, and to produce local incidence angle data that were not shadowed. The results obtained demonstrate that the SAR data are related to the elevation gradient.

Description: An X-band FM-CW very fine range resolution scatterometer was used to acquire backscattering data for a number of tree species. Using a model to describe the scattering source function and an experimental procedure for selected removal of plant parts allows the estimation of the volume backscatter coefficient and the volume extinction coefficient. It is found that: (1) leaves are strong attenuators as well as scatterers; (2) the albedo at a given angle of incidence is nearly independent of the tree type; (3) the tree limbs are good attenuators but rather poor scatterers; and (4) the albedo changes as a function of the angle of incidence and for deciduous trees is also a function of the season.

Description: A series of field experiments was carried out to extend the data base available for verifying agricultural row effect models of emissivity. The row effects model was used to simulate a data base from which an algorithm could be developed to account for row effects when the scene dielectric constant and small-scale roughness are unknown. One objective of the study was to quantify the significance of row structure and to develop a practical procedure for removing the effects of periodic row structure on the microwave emissivity of a soil in order to use the emissivity values to estimate the soil moisture. A second objective was to expand the data set available for model verification through field observations using a truck-mounted 1.4-GHz microwave radiometer.

Description: Landsat and Shuttle Imaging Radar-A L band (23.5 cm wavelength) data from 1981 were used to analyze areas of intact tropical forest and areas recently cleared from forest for agriculture and pasture in Mato Grosso, Brazil. Portions of SIR-A Data Takes #24C and #31 film were digitized using a microdensitometer. Landsat MSS data of July 1981 were also examined. The digital values from SIR-A DT 31 were compared with the normalized difference vegetation index values (NDVI) from the Landsat data for the same sites. Contrary to expectations some cleared areas had brighter radar responses than surrounding forest. The explanation seems to be that a recently cleared forest (cut and burned during the dry season) is texturally very rough as the exposed standing and fallen boles and woody litter may function as effective corner or dihedral reflectors. Combining radar data with NDVI data may help to assess the relative age of forest clearings and determine differences in both woody and green leaf biomass of primary and secondary tropical forests.

Description: Forest canopy chemistry and biomass indicators of ecosystem photosynthesis and decomposition processes are presently studied in view of Airborne Imaging Spectrometer data, which generated spectra from averaged 3 x 3 pixel areas for each of 20 sites for mutual qualitative comparison. Vegetation spectra were strongly differentiated from other cover types by an apparent absorption feature at 1500-1700 nm. Preliminary work with stepwise regression suggests that lignin may play a role in canopy reflectance, and that there is potential for remote detection of forest canopy lignin.

Description: The conceptual design of a Scanning Radar Altimeter system capable of collecting less than 300-m spatial and less than 3-m height resolution digital topography data for the entire globe, from an orbital platform, is presented. A 37-GHz frequency SRA system is used to achieve the requisite resolution while reducing antenna length in the along-track dimension. Near-global coverage in a short time period is obtained by scanning the antenna beam cross-track, in a swath of about 100 km. Attention is given to the algorithm that will be used to retrieve pixel height from the return waveform.

Description: A program is presented to perform coordinated global experiments designed to use the unique features of synthetic aperture radar (SAR) sensors such as the ones on ERS-1, SIR-C and EOS to characterize the physical nature of forest stands as input to global ecosystem and climatology models. Details about the objectives, program and expected results are presented.

Description: The effectiveness of EOS data as a basis for the study of desert surfaces' evolution is presently evaluated for both long and short term geomorphic evolution. Attention is given to the usefulness of such sensor systems planned for EOS as MODIS for regional vegetation distribution/variability monitoring, HIRIS for visible-near IR observations, TIMS for lithological identification, HMMR and SSMI for soil characteristics, LASA for atmospheric profiles, SAR for surface roughness, ALT for two-dimensional topography, ACR for the calibration of imaging sensors, and ERBE for climate modeling and regional surface albedo variation determinations.

Description: Four flights were conducted by the NASA C-130 aircraft sensor platform bearing the 'pushbroom' microwave radiometer (PBMR) over the Konza Prairie in central Kansas in 1985, in order to monitor soil surface variations. When the brightness temperature maps thus obtained were analyzed, a striking difference was noted between burned and unburned watersheds; the latter had a very high emissivity despite having saturated soils, while the former had low values that increased with the gradual drying of the soils. The lack of sensitivity for the unburned watershed is tentatively attributed to the build-up of a thatch layer by the decaying vegetation, which serves as a good microwave absorber when wet.

Description: A simple model is presented for the prediction of the full polarization signature of vegetation resembling tall grass. This polarization signature can be used to detect the presence of vegetation even in those cases in which the vegetation layers are comparatively thin. Also presented is a model which predicts the polarization dependence of different tree types. Attention is given to the cases of pine and deciduous forest model predictions; both types of forest can be expected to contain terms representing the scatter from the ground, as well as forward, double reflections from the ground and limbs/trunk.

Description: Five coding techniques in the spatial and transform domains have been evaluated for SAR image compression: linear three-point predictor (LTPP), block truncation coding (BTC), microadaptive picture sequencing (MAPS), adaptive discrete cosine transform (ADCT), and adaptive Hadamard transform (AHT). These techniques have been tested with Seasat data. Both LTPP and BTC spatial domain coding techniques provide very good performance at rates of 1-2 bits/pixel. The two transform techniques, ADCT and AHT, demonstrate the capability to compress the SAR imagery to less than 0.5 bits/pixel without visible artifacts. Tradeoffs such as the rate distortion performance, the computational complexity, the algorithm flexibility, and the controllability of compression ratios are also discussed.

Description: Two data reduction algorithms developed using the scattering and phase matrix approaches are described. In the scattering matrix approach, the scattering matrices of four consecutive along-track pixels are averaged and in the phase matrix approach, the phase matrices of four consecutive along-track pixels are averaged. The basic procedures necessary to generate a synthetic polarization image from original data sets are discussed. The two algorithms are evaluated in terms of data volume reduction and the number of errors introduced in the synthesized images. It is observed that the reduced data set produced by the scattering matrix algorithm is smaller than that generated by the phase matrix algorithm; however, greater errors are introduced into the data set by the scattering matrix algorithm than the phase algorithm. Flowcharts for the scattering and phase matrix approaches and for synthesis of uncompressible data are presented.

Description: A leaf is modeled as a deterministic two-dimensional structure consisting of a network of circular arcs designed to represent the internal morphology of major species. The path of an individual ray through the leaf is computed using geometric optics. At each intersection of the ray with an arc, the specular reflected and transmitted rays are calculated according to the Snell and Fresnel equations. Diffuse scattering is treated according to Lambert's law. Absorption is also permitted but requires a detailed knowledge of the spectral attenuation coefficients. An ensemble of initial rays are chosen for each incident direction with the initial intersection points on the leaf surface selected randomly. The final equilibrium state after all interactions then yields the leaf bidirectional reflectance and transmittance distributions. The model also yields the internal two dimensional light gradient profile of the leaf.

Description: The development of imaging radar polarimetry at JPL is considered. Advances in the correlation of polarimetric SAR and in the synthesis of arbitrary polarization are discussed. The capabilities of the polarization workstation are described. Various applications for polarimetric SAR data include: the identification of scatter, the evaluation of scattering theories, the measurement of the roughness of geologic targets, the classification of forest stands, and the characterization of trees within a particular stand. The development and testing of the DC-8 SAR and the SIR-C system are examined.

Description: In this paper software for simulation of the response of the SIR-C sensor to a point target is described. Synthetic SAR data is generated by passing successive chirps through a simulation of the transmitter electronics, propagation path and receiver electronics. This result is then processed with a digital correlator to yield the point target response of the system. This allows an accurate assessment of the effect of the radar design on the final image product.

Description: A squint mode processor for the processing of JPL aircraft SAR data collected in L-band and C-band is described which meets the criteria of optimum SNR, polarization isolation, and resolution. A steerable prefilter maximizes SNR while reducing the amount of data for processing efficiency. The resolution of images produced from data recorded at nonzero drift angle is improved by range migration compensation. The standard output products include high resolution images (with a range resolution of 7.5 m and an azimuth resolution of 2.74 m), four-look images (produced by averaging every four pixels in the azimuth direction for speckle reduction), and magnitude/phase-difference images (produced by correlating images of similar polarizations).

Description: A multiparameter synthetic aperture imaging radar is planned as a facility instrument for the Earth Orbiting System (EOS). This sensor will operate at L, C, and X and at all possible polarizations (HH, VV, HV, VH), thus allowing the acquisition of detailed information about the surface physical and electrical properties. When combined with the visible and IR imaging spectrometry data and the surface topography, a full description of the surface structure, composition, thermal properties and physical properties could then be extracted.

Description: The landscape-scale community dynamics of a boreal forest ecosystem was investigated using the Landsat MSS data record form 1973 to 1983 to generate a stochastic description of the key life cycle states of the community landscape elements. Such descriptions can provide input and verification for models of community development and landscape dynamics. It is anticipated that the proposed approach may be extended to measure, monitor, and model ecosystems at continental and planetary scales.

Description: High precision laser ranging measurements from an Earth Observing System (EOS) platform would provide data for conducting a variety of basic geoscience investigations. For example, range data to retroreflectors arranged in grids located at tectonic plate boundaries could be used to study the spatial and temporal distribution of the crustal movements associated with stress accumulation and release in an earthquake cycle. The Geodynamics Laser Ranging System (GLRS) is a proposed facility instrument for EOS that can perform the ranging function as well as simple height-measurement altimetry over ice sheets. This instrument is the derivative of several generations of ground-based laser trackers and prototype laboratory systems. The Geoscience Laser Altimetry/Ranging System (GLARS) is an advanced laser ranging system which would provide higher accuracy ranging observations and detailed altimetric mapping of surface height distributions over land and ice sheets.

Description: The HIRIS is designed to acquire images in 196 spectral bands simultaneously in the 0.4-2.5 micron wavelength region. HIRIS is a targeting rather than a continuous acquisition instrument and obtains high spatial and spectral resolution images in a 23 km swath with a 30 m GIFOV in vertical viewing. Gimbal pointing is proposed which will allow image acquisition at -30 + 60 deg down-track and + or - 25 deg cross-track. The raw data rate of the instrument is 393 Mbs. The high spectral resolution will make it possible to directly identify surficial materials such as rocks, soils, and suspended matter in water, and HIRIS opens up the possibility of studying biogeochemical processes in vegetation canopies. HIRIS will be used in conjunction with MODIS as a multistage sampling system.

Description: The Moderate Resolution Imaging Spectrometer (MODIS), a key part of the Earth Observing System planned for the 1990's, is described. The complementary MODIS-T (64 channels) and MODIS-N (40 channels) instruments provide a multispectral observing capability that has application to land, ocean, and atmospheric research. The modules have a 500-1000 meter spatial resolution to accompany a swath width sufficient to provide two-day repeat coverage from a polar-orbiting, sun-synchronous, space-station serviceable platform. High signal-to-noise capability (500/1 or better) and 10-12 bit quantization over the dynamic ranges of the various spectral bands will be provided by the two modules.

Description: An automated post-processing system has been developed for rectification and geocoding of SAR (Synthetic Aperture Radar) imagery. The system uses as input a raw uncorrected image from the operational SAR correlator, and produces as a standard output a rectified and geocoded product. The accurate geolocation of SAR image pixels is provided by a spatial transformation model which maps the slant range-azimuth SAR image pixels into their location on a prespecified map grid. This model predicts the geodetic location of each pixel by utilizing: the sensor platform position; a geoid model; the parameters of the data collection system and the processing parameters used in the SAR correlator. Based on their geodetic locations, the pixels are mapped by using the desired cartographic projection equations. This rectification and geocoding technique has been tested with Seasat and SIR-B images. The test results demonstrate absolute location uncertainty of less than 50 m and relative distortion (scale factor and skew) of less than 0.1 percent relative to local variations from the assumed geoid.

Description: Multipolarization SAR data were collected over the SIR-B target area near Fresno, California, a few days before the Shuttle flight. Attempts to use these data to evaluate roughness models were unsuccessful. It is speculated that the processing of those data to give good total swath and target contrast resulted in very little variation among the bare fields. Special processing to provide adequate contrast among the targets of interest and to account for a large incidence angle variation may be necessary to make these data useful for this type of study.

Description: The use of dual-frequency data of backscattering coefficients at a fixed angle to estimate surface roughness parameters is evaluated. Radar backscattering coefficients at 1.5 and 4.25 GHz are calculated using a model based on Kirchhoff approximation of electromagnetic wave scatttering from a rough soil surface. Plots of the calculated backscattering coefficients for Kansas soil moisture contents at the C- and L-band frequencies and HH polarization are analyzed. The effects of changes in correlation length on the backscattering coefficients are investigated. The calculated backscattering coefficients are compared with scatterometer data collected at 1.5 and 4.25 GHz, and it is detected that the model and field data correlate well. The data reveal that it is possible to retrieve the surface roughness parameters from measured radar data.

Description: A technique is presented to monitor vegetation changes for a selected study area in Costa Rica. A normalized difference vegetation index was computed for three dates of Landsat satellite data and a modified parallelipiped classifier was employed to generate a multitemporal greenness image representing all three dates. A second-generation image was created by partitioning the intensity levels at each date into high, medium, and low and thereby reducing the number of classes to 21. A sampling technique was applied to describe forest and other land cover change occurring between time periods based on interpretation of aerial photography that closely matched the dates of satellite acquisition. Comparison of the Landsat-derived classes with the photo-interpreted sample areas can provide a basis for evaluating the satellite monitoring technique and the accuracy of estimating forest clearing and regrowth rates and trends.

Description: In order to understand what areas can be used for spaceborne scatterometer calibration and to aid in the design of future land radar systems, the Seasat-1 Scatterometer System (SASS) data on the backscattering coefficient sigma0 have been studied. Global images of land and ice backscatter are presented and compared to the terrain type. The global statistics of sigma0 for incidence angles between 0 and 70 deg are presented and compared with the Skylab and ground spectrometer results. The brightest land regions are the rain forests. For two such regions, the Amazon Basin and the Congo Basin, detailed images and the variation of sigma0 with polarization, angle of incidence, and terrain type are presented. For comparison, a highly nonhomogeneous region, part of the Sahara desert, is studied.

Description: Two techniques for enhancing the color of multispectral images are described; both involve ratioing of data from different image channels. In the first technique, the ratioed data are assigned the primary color for display as color ratio pictures, and in the second method, image data are transformed to RGB chromaticity coordinates by ratioing the data acquired in three channels to the sum of their intensities. The two techniques are applied to a NASA Thermal-IR Multispectral Scanner (TIMS) image of Death Valley and to a Landsat MSS image of the Mojave Desert. The basic principles of ratioing are discussed, and the effects of atmospheric path radiances on the interpretation of ratioed images are investigated. It is observed that the color pictures produced using these two enhancement techniques are similar to the pictures enhanced by decorrelation and hue-saturation-intensity methods.

Description: A method for estimating the cloud effect on remote sensing is described, and it is applied to cloudiness in several climatological conditions. The algorithm is based on the Haurwitz (1948) measurements of the cloud layer transmission of solar radiation for an overcast sky and on an empirical interpolation of data for broken cloudiness by Pochop et al. (1968). Radiances for a sunny area observed directly from space and through a cloud, and for a shady area observed from space and through a cloud are computed. Methods for detecting the cloud effect from satellite images are discussed. The relation between cloud reflectance and cloud size is studied. It is observed that the subpixel clouds affect the detected radiance and vegetation index, and the effect depends on the cloud types and the dependence of the cloud transmissivity on cloud fraction. Procedures for decreasing or eliminating cloud effect are examined.

Description: The effects of varying atmospheric aerosol optical depth on the bidirectional reflectance distribution of vegetation canopies is investigated. The reflectance distributions of two pasture grass canopies and one soya bean canopy under different sky irradiance distributions were measured, and the data were analyzed in the visible and IR spectral bands. It is observed that, for the pasture grass canopies, the change in reflectance is due to the percentage of shadowed area viewed by the sensor, and for the soya bean, the specular reflection effect and increased diffuse irradiance penetration into the canopy cause reflectance changes. It is detected that the reflectivity for the soya bean canopy on a hazy day is lower than on a clear day; however, the opposite change is observed for the pasture grass. It is also detected that the normalized difference vegetation index values differ under clear and hazy conditions for the same vegetation canopy conditions.

Description: The missions of the German Modular Optoelectronic Multispectral Scanner (MOMS) aboard two STS flights demonstrated the feasibility of a novel concept with regard to both technical and scientific objectives. On account of the successful missions, a cooperation was agreed between the German Federal Minister for Research nad Technology and NASA for comparing MOMS observations with the more familiar operational Landsat-TM data over selected test sites, as a means of obtaining some relative measure of performance. This paper summarizes the results obtained and presents the MOMS-02, a further experimental representative of the MOMS program aiming at the realization of an operational system for the mid-nineties.

Description: Airborne radar scatterometer data on sand dunes, acquired at multiple frequencies and polarizations, are reported. Radar backscatter from sand dunes is very sensitive to the imaging geometry. At small incidence angles the radar return is mainly due to quasi-specular reflection from dune slopes favorably oriented toward the radar. A peak return usually occurs at the incidence angle equal to the angle of repose for the dunes. The peak angle is the same at all frequencies as computed from specular reflection theory. At larger angles the return is significantly weaker. The scatterometer measurements verified observations made with airborne and spaceborne radar images acquired over a number of dune fields in the U.S., central Africa, and the Arabian peninsula. The imaging geometry constraints indicate that possible dunes on other planets, such as Venus, will probably not be detected in radar images unless the incidence angle is less than the angles of repose of such dunes and the radar look direction is approximately orthogonal to the dune trends.

Description: The unique directional scattering properties of a deciduous hardwood forest without leaves during the winter period was measured in a visible and near-infrared band. A radiative transfer model was used to explore the scattering properties of such a forest. The reflectance distributions look similar to sparse homogeneous vegetation canopies. The overall reflectance distribution is a combination of the extreme azimuthal scattering behavior of tree limbs and the more typical scattering behavior of understory litter.

Description: The ability of C-band radar (4.75 GHz) to discriminate features of forest structure, including biomass, is tested using a truck-mounted scatterometer for field tests on a 1.5-3.0 m pygmy forest of cypress (Cupressus pygmaea) and pine (Pinus contorta ssp, Bolanderi) near Mendocino, CA. In all, 31 structural variables of the forest are quantified at seven sites. Also measured was the backscatter from a life-sized physical model of the pygmy forest, composed of nine wooden trees with 'leafy branches' of sponge-wrapped dowels. This model enabled independent testing of the effects of stem, branch, and leafy branch biomass, branch angle, and moisture content on radar backscatter. Field results suggested that surface area of leaves played a greater role in leaf scattering properties than leaf biomass per se. Tree leaf area index was strongly correlated with vertically polarized power backscatter (r = 0.94; P less than 0.01). Field results suggested that the scattering role of leaf water is enhanced as leaf surface area per unit leaf mass increases; i.e., as the moist scattering surfaces become more dispersed. Fog condensate caused a measurable rise in forest backscatter, both from surface and internal rises in water content. Tree branch mass per unit area was highly correlated with cross-polarized backscatter in the field (r = 0.93; P less than 0.01), a result also seen in the physical model.

Description: A two-stream approximation to the radiative-transfer equation is used to calculate the vegetation indices (simple ratio and normalized difference), the fraction of incident photosynthetically active radiation (PAR) absorbed by the canopy, and the daily mean canopy net photosynthesis under clear-sky conditions. The model calculations are tested against field observations over wheat, cotton, corn, and soybean. The relationships between the vegetation indices and radiation absorption or net photosynthesis are generally found to be curvilinear, and changes in the soil reflectance affected these relationships. The curvilinearity of the relationship between normalized differences and PAR absorption decreases as the magnitude of soil reflectance increases. The vegetation indices might provide the fractional radiation absorption with some a priori knowledge about soil reflectance. The relationship between the vegetation indices and net photosynthesis must be distinguished for C3 and C4 crops. Effects of spatial heterogeneity are discussed.

Description: This paper provides practical advice for investigators new to the field of high accuracy radiometric experiments. Sources typically used for calibrating land reflectance measurement instruments are identified, and some special considerations for their use are highlighted. The physical nature of the processes determining the output of these sources is emphasized, and an estimate of the typical calibration accuracies achieved with these sources is described. Several techniques for the validation and verification of instrument characterization are reviewed. Experiment design of the Backscattered Ultraviolet Spectrometer (BUV) and the Earth Radiation Budget Experiment (ERBE) are introduced as successful mission designs for accomplishing high accuracy radiation measurements from space. The goal of this paper is to introduce reasoning developed in other missions to experiments devoted to sensing the earth's surface and near surface.

Description: An airborne pointable imaging multispectral linear array sensor has been developed for the multidirectional observation of surface reflectance anisotropy. The sensor design permits observations up to 45 deg off-nadir in three spectral bands (green, red, and near-infrared). Calibration permits the conversion of sensor data to radiance units with an absolute uncertainty of 6 percent. Observations of five field plots from seven view directions are discussed. Calibration and atmospheric corrections are used to derive hemispherical-directional reflectance factors. A three-term reflectance model is fit to the reflectance factors for each plot to represent the continuous distribution of reflectance factors with view direction. The reflectance model is integrated over all view directions to calculate bihemispherical reflectance factors. The calculated bihemispherical factors differed by 1 to 25 percent from values based on an assumption of isotropic reflectance depending on spectral band and field plot. These calculations demonstrate the technologic and scientific capabilities required for the remote characterization of surface reflectance anisotropy. Remote multidirectional observations are both feasible and needed to fully evaluate land reflectance characteristics.

Description: This paper describes a technique to generate geocoded synthetic aperture radar (SAR) imagery corrected for terrain induced geometric distortions. This algorithm transforms the raw slant range image, generated by the signal processor, into a map registered product, resampled to either Universal Transverse Mercator (UTM) or Polar Stereographic projections, and corrected for foreshortening. The technique utilizes the space platform trajectory information in conjunction with a digital elevation map (DEM) of the target area to generate an ortho-radar map with near-autonomous operation. The current procedure requires only two to three tie-points to compensate for the platform position uncertainty that results in translational error between the image and the DEM. This approach is unique in that it does not require generation of a simulated radar image from the DEM or a grid of tie-points to characterize the image-to-map distortions. Rather, it models the inherent distortions based on knowledge of the radar data collection characteristics, the signal Doppler parameters, and the local terrain height to automatically predict the registration transformation. This algorithm has been implemented on a minicomputer system equipped with an array processor and a large random-access memory to optimize the throughput.

Description: This study evaluates the potential of measuring/mapping forest decline in spruce-fir forests using airborne NS-001 TMS data. Using field instruments, it was found that ratios of 1.65/1.23 and 1.65/0.83-micron reflectance discriminated between spruce samples of low and high-damage sites. Using TMS data, band ratios were found to be strongly correlated with ground-based measurements of forest damage. Ratio colo-density slice images using these band ratios, and images using 0.56 and 1.65-micron bands with either of these band ratios in a false-color composite, provide accurate means of detecting, quantifying and mapping levels of forest decline.

Description: Reflectance variations of a stressed cotton canopy were conducted in the presence of a fan-generated air stream to investigate the effects of air movement and the resulting temperature changes on remotely-sensed data. The initial drop in reflectance after application of the air stream was found to be greatest in the morning because leaf turgor was at a maximum, enabling leaves on the windward side of the canopy to assume surprisingly stable vertical positions. By afternoon, a reduction in leaf turgor was responsible for less stem displacement and consequently a reduction in light-trapping capability. However, reflectance oscillations were greater because the leaves had become sufficiently limp to flutter at the edges and about the petioles exposing both adaxial and abaxial surfaces to the incident light.

Description: Bidirectional reflectance to nadir in the reflective TM bands and the 1.15-1.3-micron band was measured in the laboratory as moisture content was varied in ten soils. Stronger absorption by water in TM5 and TM7 was expected to cause ratios of other bands to TM5 and TM7 to increase with water content, but in most cases these ratios were constant or decreased at low to intermediate water content and increased only at high moisture levels. Because these ratios were found to decrease as illumination elevation angle decreased, it was suggested that increased roughness resulting from the methods of moistening and mixing the soil may have tended to counteract the expected ratio increases.

Description: The weekly global vegetation index (GVI) derived from the NOAA AVHRR instrument has been analyzed for the 1982-1985 period over a wide range of vegetation formations of Asia. Temporal development curves of the index are presented for environments ranging from the desert of central Asia to the tropical forest of Borneo. The paper shows that, despite the coarse resolution of the GVI product, a large set of useful information on ecosystem dynamics and cropping practices can be consistently derived from time series of such data. In addition, it is shown that the impact of the 1982-1983 El Nino Southern Oscillation-related drought can be detected in the GVI data through an analysis of anomalies in the development of selected vegetation formations. The relevance of such analysis for global vegetation monitoring and change detection is then underlined.

Description: A November 1982 Landsat-4 TM scene and March and September 1984 airborne L-band radar data for a brackish-wetland area of the Blackwater National Wildlife Refuge (near Chesapeake Bay) are analyzed to monitor changes in vegetation and water area. The accuracy of level-I classification of the TM image is found to be 81 percent, but that of the few level-II/III classes for which ground truth was available is only 53 percent. The value of radar images for discriminating water areas obscured by vegetation and estimating plant heights is indicated.

Description: Preliminary results are discussed from an investigation into the effects of the solar zenith angle on the surface reflectances of cloud types being categorized in the International Satellite Cloud Climatology Project. The analysis is based on AVHRR channel one data at 0.6 micron over the period July 1983-July 1984. Subsatellite scenes of eight geographically separated targets distinguished by their predominant vegetation type were examined pixel-by-pixel. Attempts were made to discover any spatial variability index in each pixel to classify pixels representing clouds by the recorded brightness temperature.

Description: The application of TM data to rock discrimination is discussed. Sixteen specific terrains derived from geologic maps are examined on TM images of the Arabian shield obtained on Apr. 14, 1984; visual enhancement procedures are applied to the images. The rock types observed in the test site are described; the major sedimentary formations in the test area are laterite and sandstone. The data reveal that the layered rocks in the outcrop consist of a variety of metamorphosed volcanics, metamorphosed sediments, and amphibolite, and the intrusive complex is composed of several classes of mafic and acidic rocks.

Description: A procedure is presented to discriminate and characterize regions of uniform image texture. The procedure utilizes textural features consisting of pixel-by-pixel estimates of the relative emphases of annular regions of the Fourier transform. The utility and derivation of the features are described through presentation of a theoretical justification of the concept followed by a heuristic extension to a real environment. Two examples are provided that validate the technique on synthetic images and demonstrate its applicability to the discrimination of geologic texture in a radar image of a tropical vegetated area.

Description: This paper discusses the effects of non-Lambertian reflection from a homogeneous surface on remote sensing of the surface reflectance and vegetation index from a satellite. Remote measurement of the surface characteristics is perturbed by atmospheric scattering of sun light. This scattering tends to smooth the angular dependence of non-Lambertian surface reflectances, an effect that is not present in the case of Lambertian surfaces. This effect is calculated to test the validity of a Lambertian assumption used in remote sensing. For the three types of vegetations considered in this study, the assumption of Lambertian surface can be used satisfactorily in the derivation of surface reflectance from remotely measured radiance for a view angle outside the backscattering region. Within the backscattering region, however, the use of the assumption can result in a considerable error in the derived surface reflectance. Accuracy also deteriorates with increasing solar zenith angle. The angular distribution of the surface reflectance derived from remote measurements is smoother than that at the surface. The effect of surface non-Lambertianity on remote sensing of vegetation index is very weak. Since the effect is similiar in the visible and near infrared part of the solar spectrum for the vegetations treated in this study, it is canceled in deriving the vegetation index. The effect of the diffuse skylight on surface reflectance measurements at ground level is also discussed.

Description: Similarities and/or dissimilarities in radiance characteristics were studied among barley (Hordeum vulgare L.), oats (Avena fatua L.), spring and winter wheat (Triticum aestivum L.), and short-grass prairie vegetation. The site was a Williams loam soil (fine-loamy mixed, Typic Argiborolls) near Sidney, Montana. Radiances were measured with a truck-mounted radiometer. The radiometer was equipped with four wavelength bands: 0.45 to 0.52, 0.52 to 0.60, 0.63 to 0.69, and 0.76 to 0.90 micron. Airborne scanner measurements were made at an altitude of 600 m four times during the season under clear sky conditions. The airborne scanner was equipped with the same four bands as the truck-mounted radiometer plus the following: 1.00 to 1.30, 1.55 to 1.75, 2.08 to 2.35, and 10.4 to 12.5 microns. Comparisons using individual wave bands, the near IR/red, (0.76 to 0.90 micron)/(0.63 to 0.69 micron) ratio and the normalized difference vegetation index, ND = (IR - red)/(IR + red), showed that only during limited times during the growing season were some of the small grains distinguishable from one another and from native rangeland vegetation. There was a common relation for all small grains between leaf area index and green leaf phytomass and between leaf area index or green leaf phytomass and the IR/red ratio.

Description: The eruptions of Mount St. Helens created new surfaces by stripping and implacing large volumes of eroded material and depositing tephra in the blast area and on the flanks of the mountain. Areas of major disturbance are those in the blast zone that were subject to debris avalanche, pyroclastic flows, mudflows, and blowdown and scorched timber; and those outside the blast zone that received extensive tephra deposits. These zones represent a spectrum of disturbance types and intensities that can be indexed by temperature, impact force, and depth of subsequent deposition. This paper describes an application of NASA's Thermal Infrared Multispectral Scanner (TIMS) in monitoring vegetation recovery patterns in disturbed areas. Preliminary study results indicate a significant correlation between measured effective radiant temperature and vegetated/nonvegetated areas, percent vegetation cover, and vegetation type.

Description: A World Data Base of potential thematic mapper (TM) scenes was developed to aid in acquisition planning. The World Data Base contains geopolitical, geographic and economic regions along with a format that enables users to find the satellite day, sun angle and cloud cover probability for any month of the year. Scenes that have been acquired by TM and have an average cloud cover of 30 percent of less from July 1982 when TM was launched until the Landsat system was taken over by NOAA in September 1984 are also in the World Data Base. Processed data are referenced in maps and data bases at EROS Data Center; however, a large number of acquistions have never been processed and therefore are not accessible. The World Data Base enables the rapid location of scenes and areas with the least effort making it invaluable in TM scheduling. Users of TM data can use the World Data Base to determine if scenes of interest have been acquired, the acquisition date, and if scenes have been processed to computer-compatible tape (CCT). These uses of the World Data Base make it a valuable tool in the acquisition and location of TM scenes.

Description: A canopy-reflectance (CR) model for row-planted vegetation is presented. Its use of an estimation of important biophysical variables like leaf-area index (LAI) and average leaf angle (ALA) from bidirectional CR data is discussed. Using field-measured CR data for a partially covered soybean canopy, it is shown that one can accurately estimate LAI, ALA and extent of percentage of ground cover from CR data.

Description: Analysis of Thermal Infrared Multispectral Scanner data collected over H. J. Andrews experimental forest in western Oregon indicated that aspect and slope gradient had a greater effect on the thermal emission of younger reforested clearcuts than of older stands. Older forest stands (older than 25 years) with greater amounts of green biomass and closed canopies, had lower effective radiant temperatures than younger, less dense stands. Aspect and slope had little effect on the effective radiant temperature of these older stands. Canopy temperature recorded at approximately 1:30 pm local time July 29, 1983 were nearly equal to maximum daily air temperature recorded at eight reference stands. The investigation provided some insights into the utility of the thermal sensor for detecting surface temperature differences related to forest composition and green biomass amounts in mountain terrain.

Description: Earlier encouraging test results of a contextual classifier that combines spatial and spectral information employing a general statistical approach are expanded. The earlier results were of limited meaning because they were produced from small (50-by-50 pixel) data sets. An implementation of the contextual classifier on NASA Goddard's Massively Parallel Processor (MPP) is presented; for the first time the MPP makes feasible the testing of the classifier on large data sets (a 12-hour test on a VAX-11/780 minicomputer now takes 5 minutes on the MPP). The MPP is a Single-Instruction, Multiple Data Stream computer, consisting of 16,384 bit serial microprocessors connected in a 128-by-128 mesh array with each element having data transfer connections with its four nearest neighbors so that the MPP is capable of billions of operations per second. Preliminary results are given (with more expected for the conference) and plans are mentioned for extended testing of the contextual classifier on Thematic Mapper data sets.

Description: Useful quantitative information about soil properties may be obtained by calibrating energy and moisture balance models with remotely sensed data. A soil physics model solves heat and moisture flux equations in the soil profile and is driven by the surface energy balance. Model generated surface temperature and soil moisture and temperature profiles are then used in a microwave emission model to predict the soil brightness temperature. The model hydraulic parameters are varied until the predicted temperatures agree with the remotely sensed values. This method is used to estimate values for saturated hydraulic conductivity, saturated matrix potential, and a soil texture parameter. The conductivity agreed well with a value measured with an infiltration ring and the other parameters agreed with values in the literature.

Description: Texture analysis was performed as part of an investigation of the information content of Thematic Mapper (TM) imagery. High altitude aircraft scanner imagery from the Airborne Thematic Mapper (ATM) instrument was acquired over central California and used to simulate TM data. Edge density texture images were constructed by computation of proportions of edge pixels in a 31 x 31 moving window on a near infrared ATM band. A training technique was employed to select computational parameters to maximize the difference between edge density measurements in urban and in rural areas. The results of classification of the texture images showed that urban and rural areas could be distinguished with texture alone, indicating that inclusion of texture in automated classification procedures could significantly improve their accuracy.

Description: This paper presents the techniques and the utility of multipolarization Synthetic Aperture Radar (SAR) data for surface feature delineation. Three channels of ratioed data (VV/HH, VH/HH, and VH/VV) are generated from the HH, VV, and VH polarization data (V = vertical, H = horizontal). The technique assumes redundancy of the VH and HV polarization and only VH polarization is used. The ratioed data are linearly stretched to yield a digital number within a range of 0 to 255. Based on the separability measure for two-class delineation, it was found that (1) the ratioed data resulted in a better delineation of surface features with high like (HH or VV) polarization digital number, and (2) the use of ratioed data provided further information not available from the original three-polarization data. The results suggest an advantage in using the ratioed data and the original three-polarization data for surface feature delineation.