Engineering topics
Publications and source records attributed to Freeman, Anthony.
Report describes recent improvements in calibration of data gathered by polarimetric synthetic-aperture-radar (SAR) system operated by NASA's Jet Propulsion Laboratory. Improvements are: relocation of calibration site; JPL POLCAL software; and ANTENCAL software. Presents analysis of calibration performance over number of sites during years 1989 and 1990.
As Synthetic Aperture Radar (SAR) system performance and experience in SAR signature evaluation increase, quantitative analysis becomes more and more important. Such analyses require an absolute radiometric calibration of the complete SAR system. To keep the expenditure on calibration of future multichannel and multisensor remote sensing systems (e.g., X-SAR/SIR-C) within a tolerable level, data from different tracks and different sensors (channels) must be cross calibrated. The 1989 joint E-SAR/DC-8 SAR calibration campaign gave a first opportunity for such an experiment, including cross sensor and cross track calibration. A basic requirement for successful cross calibration is the stability of the SAR systems. The calibration parameters derived from different tracks and the polarimetric properties of the uncalibrated data are used to describe this stability. Quality criteria for a successful cross calibration are the agreement of alpha degree values and the consistency of radar cross sections of equally sized corner reflectors. Channel imbalance and cross talk provide additional quality in case of the polarimetric DC-8 SAR.
Noise equivalent sigma(sup 0) is a fundamental parameter in any synthetic aperture radar (SAR) measurement. Knowledge of the noise equivalent sigma(sup 0) allows the radar application scientist to band the error contributions due to noise present in his/her data. In this paper, a novel technique for determining noise equivalent sigma(sup 0) is presented. It has been observed that a pair of images obtained by a SAR over the same area in a nearly repeating orbit can demonstrate a high degree of correlation, depending on the baseline. If noise independence and signal power invariance are assumed, an estimate for the noise equivalent sigma(sup 0) for each image can be determined without any known ground targets. We examine the application of this technique to several images of different backscatter obtained by ERS-1.
We seek to combine high-resolution remotely sensed data with models and ground truth measurements, in the context of a Geographical Information System (GIS), integrated with specialized image processing software. We will use this integrated system to analyze the data from two Case Studies, one at a boreal forest site, the other a tropical forest site. We will assess the information content of the different components of the data, determine the optimum data combinations to study biogeophysical changes in the forest, assess the best way to visualize the results, and validate the models for the forest response to different radar wavelengths/polarizations. During the 1990's, unprecedented amounts of high-resolution images from space of the Earth's surface will become available to the applications scientist from the LANDSAT/TM series, European and Japanese ERS-1 satellites, RADARSAT and SIR-C missions. When the Earth Observation Systems (EOS) program is operational, the amount of data available for a particular site can only increase. The interdisciplinary scientist, seeking to use data from various sensors to study his site of interest, may be faced with massive difficulties in manipulating such large data sets, assessing their information content, determining the optimum combinations of data to study a particular parameter, visualizing his results and validating his model of the surface. The techniques to deal with these problems are also needed to support the analysis of data from NASA's current program of Multi-sensor Airborne Campaigns, which will also generate large volumes of data. In the Case Studies outlined in this proposal, we will have somewhat unique data sets. For the Bonanza Creek Experimental Forest (Case 1) calibrated DC-8 SAR (Synthetic Aperture Radar) data and extensive ground truth measurement are already at our disposal. The data set shows documented evidence to temporal change. The Belize Forest Experiment (Case 2) will produce calibrated DC-8 SAR and AVIRIS data, together with extensive measurements on the tropical rain forest itself. The extreme range of these sites, one an Arctic forest, the other a tropical rain forest, has been deliberately chosen to find common problems which can lead to generalized observations and unique problems with data which raise issues for the EOS System.
Remote sensing with synthetic aperture radars (SAR's) is a rapidly developing field. Calibration of these sensors is required for the establishment of relationships between radar backscatter and geographical parameters. A review of recent progress in SAR calibration is presented. The quantities measured by SAR are defined and mathematical formulations of the three basic types of SAR images are developed. The establishment of scientific requirements for calibration and the difficulties involved are discussed. Image quality assessment is reviewed and the problems of radiometric calibration of SAR images using the radar equation and internal and external approaches are considered. Polarimetric radar calibration and the development of the necessary algorithms are described. Interferometric phase calibration and its associated problems are reviewed and future challenges in SAR calibration are discussed.
Report summarizes study of radiometrically and polarimetrically calibrated data from airborne synthetic-aperture-radar images of single-species forest stands in frozen and thawed conditions. Copolarization radar cross sections found as much as 6 dB less in frozen than in thawed condition. Effect attributed to fact that dielectric constants are greater in thawed condition. More-subtle differences determined by computing two features of statistics of Strokes matrices over resolution elements of each forest stand.
Imaging radar polarimeters are usually implemented using a Synthetic Aperture Radar (SAR) approach to give a high resolution image in two dimensions: range and azimuth. For each pixel in the image a polarimetric SAR gives sufficient information to characterize the polarimetric scattering properties of the imaged area (or target) as seen by the radar. Using a polarimetric SAR system as opposed to a single-polarization SAR system provides significantly more information about the target scattering mechanisms and allows better discrimination between different types of surfaces. In these notes a brief overview of SAR polarimetry is offered. The notes are intended as a text to accompany a lecture on SAR polarimetry as part of an AGARD-NATO course. Covered in the notes are the following: the polarization properties of electromagnetic waves; the concepts of radar scattering and measuring radar backscatter with a SAR; polarization synthesis; scattering matrix, Stokes matrix, and covariance matrix representations of polarimetric SAR data; polarization signature plots; design and calibration of polarimetric SAR systems; polarization filtering for target detection; fitting a simple model to polarimetric SAR measurements of naturally occurring features; and supervised classification of polarimetric SAR data.
Report describes international collaboration on experiments in calibration of synthetic-aperture radar (SAR) systems. Purposes of experiments were to: address problems associated with calibration of data, cross-calibrate between sets of imaging-radar data obtained by different systems, test designs of ground calibration equipment, and find candidate sites for calibration of future satellite-borne SAR systems.
Azimuthal symmetry is not assumed. Two sets of equations and algorithms for radiometric and polarimetric calibration of complex quadpolarization data gathered by synthetic-aperture radar (SAR) based partly on known scattering characteristics of artificial targets in scene, and partly on assumption that all other targets scatter reciprocally.
Calibration method accounts for characteristics of real radar equipment invalidating standard 2 X 2 complex-amplitude R (receiving) and T (transmitting) matrices. Overall gain in each combination of transmitting and receiving channels assumed different even when only one transmitter and one receiver used. One characterizes departure of polarimetric Synthetic Aperture Radar (SAR) system from simple 2 X 2 model in terms of single parameter used to transform measurements into format compatible with simple 2 X 2 model. Data processed by applicable one of several prior methods based on simple model.
An equivalence is noted between the backscatter and the polarimetric radar system assumptions for the approaches given by Klein (1992) and van Zyl (1990). It is shown that, to first order in the radar system crosstalk, an exact solution to the Stokes matrix format data calibration problem exists, and that van Zyl's approach can give this first-order solution for appropriately symmetrized polarimetric radar data. If the data are suitably symmetrized, the van Zyl approach can be used to calibrate both scattering matrix and Stokes matrix format data.
Report addresses problem of calibration of differences between phases (relative to transmitted signals) of signals received in two polarization channels of polarimetric imaging radar system. Causes of various types of errors discussed. Calibration necessary to deduce information about target area - type of terrain, presence of vegetation, and land/water boundaries.
Algorithm for calibration of polarimetric synthetic-aperture-radar (SAR) data in Stokes-matrix representation more nearly exact. Present algorithm resembles prior algorithm closely, except requires extraction of one additional parameter from high-resolution raw SAR data in scattering-matrix format before compression and transformation to Stokes-matrix format and slight change in subalgorithm compressing those data.
We seek to combine high-resolution remotely sensed data with models and ground truth measurements, in the context of a Geographical Information System, integrated with specialized image processing software. We will use this integrated system to analyze the data from two Case Studies, one at a bore Al forest site, the other a tropical forest site. We will assess the information content of the different components of the data, determine the optimum data combinations to study biogeophysical changes in the forest, assess the best way to visualize the results, and validate the models for the forest response to different radar wavelengths/polarizations. During the 1990's, unprecedented amounts of high-resolution images from space of the Earth's surface will become available to the applications scientist from the LANDSAT/TM series, European and Japanese ERS-1 satellites, RADARSAT and SIR-C missions. When the Earth Observation Systems (EOS) program is operational, the amount of data available for a particular site can only increase. The interdisciplinary scientist, seeking to use data from various sensors to study his site of interest, may be faced with massive difficulties in manipulating such large data sets, assessing their information content, determining the optimum combinations of data to study a particular parameter, visualizing his results and validating his model of the surface. The techniques to deal with these problems are also needed to support the analysis of data from NASA's current program of Multi-sensor Airborne Campaigns, which will also generate large volumes of data. In the Case Studies outlined in this proposal, we will have somewhat unique data sets. For the Bonanza Creek Experimental Forest (Case I) calibrated DC-8 SAR data and extensive ground truth measurement are already at our disposal. The data set shows documented evidence to temporal change. The Belize Forest Experiment (Case II) will produce calibrated DC-8 SAR and AVIRIS data, together with extensive measurements on the tropical rain forest itself. The extreme range of these sites, one an Arctic forest, the other a tropical rain forest, has been deliberately chosen to find common problems which can lead to generalized observations and unique problems with data which raise issues for the EOS System.
A number of approaches for calibrating SAR imaging data including data from multipolarization and multifrequency systems developed at the Jet Propulsion Laboratory are discussed. It is shown how calibrated SAR data can be used to improve the understanding of the physical properties of the earth's surface layer and how calibration uncertainties and system noise should be handled by the SAR data user. Examples of calibrated SAR data from the NASA/JPL DC-8 SAR are used to demonstrate use of calibrated data to monitor temporal change and to improve the classification of land cover type.
The results of a calibration analysis performed on ERS-1 synthetic aperture radar (SAR) images produced by the Alaska SAR facility (ASF) are presented, together with some preliminary results on change detection on the Alaskan north slope derived from the same images. Image quality, geometric and radiometric fidelity, and repeat pass radiometric stability have all been determined to be satisfactory. A calibration workstation has been designed and implemented for use in operational data quality analysis of ASF data products. Higher-level data analysis programs have been developed for interferometry, texture analysis, and change detection.
Correction algorithm reversible and applicable to both detected and complex images. Method described accounts for major effects of transmitter, antenna, receiver, and certain aspects of SAR signal processing upon calibration and provides for appropriate reversible corrections applied in processing. Applicable to both detected (amplitude-only) and complex (amplitude and phase) SAR image data and provides for optional subtraction of estimated noise floor from image.
The validity of the 2 x 2 receive R and transmit T model for radar polarimeter systems, first proposed by Zebker et al. (1987), is questioned. The model is found to be invalid for many practical realizations of radar polarimeters, which can lead to significant errors in the calibration of polarimetric radar images. A more general model is put forward, which addresses the system defects which cause the 2 x 2 model to break down. By measuring one simple parameter from a polarimetric active radar calibration (PARC), it is possible to transform the scattering matrix measurements made by a radar polarimeter to a format compatible with a 2 x 2 R and T matrix model. Alternatively, the PARC can be used to verify the validity of the 2 x 2 model for any polarimetric radar system. Recommendations for the use of PARCs in polarimetric calibration and to measure the orientation angle of the horizontal (H) and vertical (V) coordinate system are also presented.