Engineering topics
Publications and source records attributed to Dubois, Pascale C..
None given. Paper covers earlier models for determining soil moisture and surface roughness from radar data. It covers problems with other models, how the data can be thrown off, and possible solutions to remedy those calculations.
In an earlier study, an empirical model was developed to infer soil moisture and surface roughness from radar data. The inversion technique was extensively tested over bare surfaces by comparing the estimated soil moisture to in situ measurements. The overall root mean square (RMS) error in the soil moisture estimate was found to be about 3.5% and the RMS error in the RMS height estimate was less than 0.35 cm absolute for bare or slightly vegetated surfaces. However, inversion results indicate that significant amounts of vegetation cause the algorithm to underestimate soil moisture and overestimate RMS height. Among the areas over which the inversion cannot be applied, the areas with intermediate vegetation cover are of particular interest as both the vegetation and the underlying bare surface affect the backscatter. This paper concentrates mostly on these areas. Using the full polarimetric information and the Cloude target decomposition approach, three different components of the target backscattering can be isolated. One of these three components can be identified as the surface component in the case of intermediate vegetation cover. Once the surface component of the scattering is isolated, the bare surface inversion can then be applied.
In an earlier study, an empirical model was developed to infer soil moisture and surface roughness from radar data. The inversion technique was extensively tested over bare surfaces by comparing the estimated soil moisture to in situ measurements. The overall RMS error in the soil moisture estimate was found to be 3.5% and the RMS error in the RMS height estimate was less than 0.35 cm absolute for bare or slightly vegetated surfaces. However, inversion results indicate that significant amounts of vegetation cause the algorithm to underestimate soil moisture and overestimate RMS height. Among the areas over which the inversion cannot be applied, the areas with intermediate vegetation cover are of particular interest as both the vegetation and the underlying bare surface affect the backscatter. This paper concentrates mostly on these areas. Using the full polarimetric information and the Cloude target decomposition approach. Three different components of the target backscattering can be isolated. One of these three components can be identified as the surface component in the case of intermediate vegetation cover. Once the surface component of the scattering is isolated, the bare surface inversion can then be applied.
An empirical algorithm for the retrieval of soil moisture content and surface Room Mean Square
Technique for processing synthetic-aperture-radar (SAR) signals exploits fact that center frequency of frequency channel of data being processed linked directly to squint angle and to azimuth angle. Also exploits fact in multilook SAR data processing, total bandwidth of each SAR polarization channel divided into N (typically, N = 16) equal frequency channels, data in each of which processed separately. Technique has yielded data on directionality of radar back scatter from agricultural fields.
An empirical model was developed to infer soil moisture and surface roughness from radar data. The accuracy of the inversion technique is assessed by comparing soil moisture obtained with the inversion technique to in situ measurements. The effect of vegetation on the inversion is studied and a method to eliminate the areas where vegetation impairs the algorithm is described.
This paper presents a study of the direction angle sensitivity of radar backscatter from agricultural fields. The direction angle is defined as the angle between the incident plane and the perpendicular to the field row direction. Previous studies have concentrated on comparing the backscatter of agricultural fields imaged with 0, 45, and 90 degree direction angles. In contrast, this study concentrates on the backscatter changes occurring when the direction angle is modified by a few degrees or even by fractions of a degree. This is possible by using the output of the NASA/JPL AIRSAR processor, in which sixteen independent frames are formed, each one corresponding to the same radar scene imaged with a slightly different squint angle. The studied data set is an agricultural area in La Mancha, Spain acquired in June and July 1991 during the EFEDA experiment. This paper describes the observed backscatter variations of the agricultural fields with direction angle measured at P, L, and C bands. As expected, the backscatter is maximum for a 0 degree direction angle. For several fields, the backscatter at P and L bands drops by more than 10 dB for a 5 degree change in direction angle. Furthermore, the sensitivity to the direction angle decreases with increasing vegetation. The variations in backscatter are compared with model predictions. One model, which agrees with scatterometer data, underestimates the observed backscatter variations with direction angle by more than 10 dB. It does not take into account the possible coherent component of the radar signal. We believe the strong direction sensitivity of agricultural field backscatter obtained with SAR data is due to a Bragg resonant effect, resulting in a strong coherent return when the direction angle is zero. The observations are then projected to the case of spaceborne SAR data.
Many of the experiments proposed for the forthcoming SIR-C mission require calibrated data, for example those which emphasize (1) deriving quantitative geophysical information (e.g., surface roughness and dielectric constant), (2) monitoring daily and seasonal changes in the Earth's surface (e.g., soil moisture), (3) extending local case studies to regional and worldwide scales, and (4) using SIR-C data with other spaceborne sensors (e.g., ERS-1, JERS-1, and Radarsat). There are three different aspects to the SIR-C calibration problem: radiometric and geometric calibration, which have been previously reported, and polarimetric calibration. The study described in this paper is an attempt at determining the science requirements for polarimetric calibration for SIR-C. A model describing the effect of miscalibration is presented first, followed by an example describing how to assess the calibration requirements specific to an experiment. The effects of miscalibration on some commonly used polarimetric parameters are also discussed. It is shown that polarimetric calibration requirements are strongly application dependent. In consequence, the SIR-C investigators are advised to assess the calibration requirements of their own experiment. A set of numbers summarizing SIR-C polarimetric calibration goals concludes this paper.
A method for unsupervised segmentation of polarimetric synthetic aperture radar (SAR) data into classes of homogeneous microwave polarimetric backscatter characteristics is presented. Classes of polarimetric backscatter are selected on the basis of a multidimensional fuzzy clustering of the logarithm of the parameters composing the polarimetric covariance matrix. The clustering procedure uses both polarimetric amplitude and phase information, is adapted to the presence of image speckle, and does not require an arbitrary weighting of the different polarimetric channels; it also provides a partitioning of each data sample used for clustering into multiple clusters. Given the classes of polarimetric backscatter, the entire image is classified using a maximum a posteriori polarimetric classifier. Four-look polarimetric SAR complex data of lava flows and of sea ice acquired by the NASA/JPL airborne polarimetric radar (AIRSAR) are segmented using this technique. The results are discussed and compared with those obtained using supervised techniques.
Multifrequency polarimetric synthetic aperture radar (SAR) images of the Walnut Gulch watershed near Tombstone, Arizona were acquired on 28 Mar. 1990 and on 1 Aug. 1990. Trihedral corner reflectors were deployed prior to both overflights to allow calibration of the two SAR data sets. During both overflights, gravimetric soil moisture and dielectric constant measurements were made. Detailed vegetation height, density, and water content measurements were made as part of the Monsoon 1990 Experiment. Preliminary results based on analysis of the multitemporal polarimetric SAR data are presented. Only the C-band data (5.7-cm wavelength) radar images show significant difference between Mar. and Aug., with the strongest difference observed in the HV images. Based on the radar data analysis and the in situ measurements, we conclude that these differences are mainly due to changes in the vegetation and not due to the soil moisture changes.
The direction angle sensitivity of agricultural field backscatter is studied. The direction angle is defined as the angle between the incident plane and the perpendicular to the row direction. Maximum backscatter power from an angricultural field is expected to occur when the furrow induced slopes are oriented towards the radar, i.e., for a 0 deg direction angle. This effect is known as the cardinal effect. Because of the way the looks are formed in the NASA/JPL airborne synthetic aperture radar (AIRSAR) processor, each look corresponds to a slightly different squint angle. This provides a unique data set to analyze the cardinal effect, as it allows simultaneous observations of the backscatter of a field for sixteen different direction angles. The backscatter variations of the agricultural fields with direction angle at P-, L-, and C-bands is described. The observed variations in backscatter are compared with model predictions. The model predicts that the maximum backscatter occurs for a 0 deg direction angle, but underestimates the backscatter variations with direction angle by more than 10 dB.
A polarization filter can maximize the signal-to-noise ratio of a polarimetric SAR and help discriminate between targets or enhance image features, e.g., enhance contract between different types of target. The method disclosed is based on the Stokes matrix/Stokes vector representation, so the targets of interest can be extended targets, and the method can also be applied to the case of bistatic polarimetric radars.
Theoretical analysis of polarization filtering of synthetic-aperture-radar (SAR) returns provide hybrid method applied to either (1) maximize signal-to-noise ratio of return from given target or (2) enhance contrast between targets of two different types (that have different polarization properties). Method valid for both point and extended targets and for both monostatic and bistatic radars as well as SAR. Polarization information in return signals provides more complete description of radar-scattering properties of targets and used to obtain additional information about targets for use in classifying them, discriminating between them, or enhancing features of radar images.
A polarization filter can maximize the signal-to-noise ratio of a polarimetric synthetic aperture radar (SAR) and help discriminate between targets or enhance image features, e.g., enhance contrast between different types of target. The method disclosed is based on the Stokes matrix/ Stokes vector representation, so the targets of interest can be extended targets, and the method can also be applied to the case of bistatic polarimetric radars.
Two alternative methods are disclosed for digital reduction of synthetic aperture multipolarized radar data using scattering matrices, or using Stokes matrices, of four consecutive along-track pixels to produce averaged data for generating a synthetic polarization image.
A theoretical analysis of polarization filtering for the bistatic case is developed for optimum discrimination between two types of targets. The resulting method is half analytical and half numerical. Because it is based on the Stokes matrix representation, the targets of interest can be extended targets. The scattered field from such targets is partially polarized. This method is then applied to the monostatic case with numerical examples relying on the JPL (Jet Propulsion Laboratory) full-polarimetric L-band radar data. A matched filter to maximize the power ratio between urban and natural targets is developed. The results show that the same filter is optimal for both ocean and forest targets as natural targets.
Two alternative methods are presented for digital reduction of synthetic aperture multipolarized radar data using scattering matrices, or using Stokes matrices, of four consecutive along-track pixels to produce averaged data for generating a synthetic polarization image.
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.