A.L. Gray

Synthetic aperture radar calibration using reference reflectors

A simple expression for the terrain backscatter coefficient is derived in terms of the integrated power of an adjacent known radar reflector in a synthetic aperture radar (SAR) image. It is shown that this technique for SAR image calibration is independent of the radar system focus or partial coherence and thereby possesses an important advantage over the usual technique, which relies on an estimate of the peak of the reflector impulse response. Results from airborne SAR overflights of corner reflectors and active radar calibrators are used to demonstrate the validity and consistency of the method and to show that the method is robust under defocus caused by an incorrect FM rate or inadequate motion compensation of data collected during turbulence. It is also shown that the fading errors associated with the integral method are comparable to or slightly worse than those associated with the peak estimation method. However, this small disadvantage is outweighed by the fact that the integral method is independent of actual resolution. >

Repeat-pass interferometry with airborne synthetic aperture radar

It is shown that interference can be observed by coherently combining pairs of either X- or C-band airborne synthetic aperture radar (SAR) images from separate passes over the same test site. Coherence between separate images is obtained only if the aircraft is flown, and the data are processed in such a way that each resolution cell in the two images is viewed with very nearly the same geometry. Successful repeat-pass interferometric results were obtained from those passes flown by the CCRS Convair 580 aircraft with flight-line offsets of less than a few tens of meters. A summary of the experiment, the phase correction of nonrectilinear aircraft motion, and the subsequent data processing is provided. >

InSAR results from the RADARSAT Antarctic Mapping Mission data: estimation of glacier motion using a simple registration procedure

An interferometric method is used to derive ice motion from RADARSAT data collected during the Antarctic Mapping Mission. Although one cannot solve for both topography and ice motion using one interferometric pair, it is possible to use a coarsely sampled digital terrain model to estimate ice motion using an image registration method. Less accurate than the usual fringe counting method for estimation of radial displacement, the image registration method allows useful motion estimation in both range and azimuth. The method is described and some results shown for a large area (/spl sim/17000 km/sup 2/) including ice flow into the Filchner Ice Shelf.

Validation of along-track interferometric SAR measurements of ocean surface waves

C-band HH-polarized along-track interferometric synthetic aperture radar (ATInSAR) images of ocean waves from the Canada Centre for Remote Sensing (CCRS) CV-580 SAR are used to study ATInSAR wave imaging fidelity. ATInSAR phase spectra are derived using cross spectra between the interferogram phase generated from individual look complex SAR images. The resulting observed phase spectra are quantitatively compared with forward-mapped in situ directional wave spectra collocated with the ATInSAR observations. The forward mapping models include two quasilinear forward transforms that include varying degrees of velocity bunching nonlinearity, both of which include an azimuth null in the modulation transfer function, and an implementation of the fully nonlinear forward transform. The null arises from the cancellation of upward and downward orbital velocity components within a resolution cell. The second of the two quasilinear transforms and the nonlinear forward transform are shown to have agreement with the observed velocity spectra, as measured by correlation coefficients between the observed spectra or forward-mapped spectra and the ATInSAR spectra. ATInSAR can allow a direct measurement of the orbital velocity spectrum. There is, however, an added layer of complexity caused by the cancellation of upward and downward orbital velocity components within a resolution cell. Furthermore, the SAR complication of velocity bunching nonlinearity is not avoided by using an ATInSAR.

Airborne interferometric SAR results from mountainous and glacial terrain

In February 1992, the CCRS Interferometric SAR (CCRS C-InSAR system) was flown over a region in the Canadian Rockies. The test site covered part of the Kananaskis Valley and the Spray Lakes Reservoir and included terrain relief changes of over 1500 metres. The test site was imaged from the 4 cardinal points of the compass, however, the relief changes were so extreme that it was not possible to fill in every gap in the derived digital terrain model caused by layover and radar shadow. The DEM results have been checked by comparing the consistency of the results from opposing passes, by comparison with results generated from aerial stereo photography, and by comparison with GPS ground survey sites. It will be shown that the consistency of results from opposing passes is good and the implied system accuracy compares favourably with that obtained by comparison with the reference DEM and ground survey points. Deriving terrain elevation from Arctic glacier covered terrain using aerial stereo photography is difficult because of weather and flight requirements, coupled with the difficulty of identifying tie points in photography of terrain which is snow covered much of the year. To test InSARs ability to map glacial terrain, the CCRS system was flown over a mountainous test site (with over 1600 metres relief change) on Bylot island in the Canadian Arctic. The test has confirmed the suitability of radar interferometric techniques for this type of terrain. Initial results are shown.<<ETX>>

Time-delayed reflections in L-band synthetic aperture radar imagery of icebergs

In some L-band synthetic aperture radar imagery of icebergs, there are relatively strong reflections which do not exist on the coincident X-band imagery and which do not correspond to attributable surface roughness or features. As these false images appear on the down-range side of the iceberg, they were originally explained as multiple reflections within the iceberg. It is shown that the majority of the observations correspond to the expected positions of a transit to and from the ice-water interface at the bottom surface of the iceberg. Factors governing observation of the bottom surface reflection and its significance in terms of underwater draught are discussed. >

Progress in the development of the CCRS along-track interferometer

The CCRS C-band SAR owned and operated by the Canada Centre for Remote Sensing (CCRS) was modified in 1991 to operate in an across-track C-band interferometric mode for derivation of terrain elevation. This mode has now been extended such that the two C-band receivers can be used with two separate, almost identical, antennas aligned in the along-track direction thereby creating a SAR capable of detecting and measuring target radial motion. Two new C(H) antennas have been mounted on the right-hand side of the Convair 580 to form the along-track C-band interferometer. The antennas share a common, rigid, mounting structure and the phase centres are separated by 0.5 m. The primary application for this mode will be in ocean monitoring R&D; SAR wave and wake imaging, measurement of coastal and ocean currents, estimation of pack-ice drift, detection of sub-surface sand waves through current modulation, etc. while the short baseline dictated by the existing radome leads to a relatively low sensitivity to radial motion, approximately 24/spl deg/ per m/s, the time between image formation at the same azimuth geometry is small (less than 2 ms) with respect to typical C-band ocean coherence times (around 50-100 ms). This, combined with a good signal-to-noise ratio, will allow relatively low phase noise on the interferometric products and therefore adequate sensitivity to radial motion for most situations. A description of the new system and of ground and airborne testing are given.<<ETX>>

The Canadian airborne R&D SAR facility: the CCRS C/X SAR

The CCRS C- and X-band SAR systems have evolved continuously since their commissioning in 1988. This paper summarizes the present state of the radar system and outlines its capabilities.