AMSR 2002presentSwath width 1445 km

Electronics Repair Manuals

Schematic Diagrams and Service Manuals

Get Instant Access

Table 73.3 Details of SAR missions suitable for remote sensing of the cryosphere

SAR Time period Frequency (GHz) Incidence angle Polarization

Table 73.3 Details of SAR missions suitable for remote sensing of the cryosphere

SAR Time period Frequency (GHz) Incidence angle Polarization

ERS-1

1991-

1995

5.3

23

VV

ERS-2

1995-

2002

5.3

23

VV

ENVISAT

2002-

present

5.3

23

VV, HH, VH

SRTM

2000

5.3 and 10

JERS-1

1992-

1998

1.2

35

HH

RADARSAT-1

1995-

2003

5.3

20-

50

HH

RADARSAT-2

2006

5.3

20-

50

HH

Figure 73.2 Schematic diagram illustrating the geometry of repeat pass synthetic aperture radar interferometry. Ai and Aj are the positions of the satellite at epochs i and j, B is the separation in space, or baseline, between the two measurements and r and rj are the ranges to a target on the surface.

Figure 73.2 Schematic diagram illustrating the geometry of repeat pass synthetic aperture radar interferometry. Ai and Aj are the positions of the satellite at epochs i and j, B is the separation in space, or baseline, between the two measurements and r and rj are the ranges to a target on the surface.

and open water, for example, must rely on backscatter and textural differences, which is not always adequate.

73.3.3.1 Interferometric synthetic aperture radar (InSAR)

The fact that SARs record the phase of the signal has allowed for the possibility of combining images taken at different times and/or locations to produce interference patterns caused by differences in phase in the two images. This is known as interfero-metric SAR or InSAR and its use has resulted in some remarkable results. Repeat pass interferometry is where pairs of images are combined that have been taken at different times (and from slightly different positions along an orbit) and began to be developed for glaciological applications after the launch of ERS-1 in 1991. Single pass interferometry (as exemplified by the shuttle radar topography mission, SRTM) is where two images are recorded at the same time but from different positions. Differences in path length of a fraction of a wavelength can be measured from the phase offset between the two images allowing, for example, millimetric displacements to be observed. The interference pattern is a function of (i) the topography, (ii) any displacement of the surface that has taken place between the two image acquisitions and (iii) the separation in space (known as the baseline) of the SAR when the two images were acquired (Fig. 73.2). A brief outline of the concept is given below.

For the ith and jth observations in repeat pass observations, the interferometric phase difference at any point in the interferogram is given by

Dfij = (4p/1)Bijsin(0 - aij) + (4p/1)Apij = ftopography + fmotion (1)

where the baseline, Bij, is the distance separating the two points, with baseline Bij, 0 is the observation angle of the SAR, and a is the tilt of the baseline with respect to the horizontal (Fig. 73.2). The first term, ftopography, contains phase information related to the topography of the surface with respect to the baseline. If the targets are displaced by Apij in the range direction between the two observations, then the observed phase will include a second contribution due to this displacement. The sensitivity of the measurements to topography is proportional to the baseline B. For example, if the baseline is zero then ftopography is zero and the interference pattern is dependent on displacement only. Variable fmotion, however, is independent of B. When the ith and jth observations are acquired at the same time (single-pass interferometry), only ftopographyis non-zero.

Normally, InSAR can provide only relative height information, and ground control points (GCPs) in the form of either GPS data or a pre-existing course resolution DEM are required for three purposes: (i) to provide absolute height control, (ii) to improve the baseline estimate, which is crucial to obtain accurate results, and (iii) to separate ftopography from fmotion. It is also possible to achieve (iii) by using more than two images to produce two or more interferograms, while making the assumption that fmotion is a constant term in all of them. More often, a DEM derived from some other source is used. Perhaps of greater value than the topographic term in Equation (1), however, is the component fmotion. Major advances in our understanding of, and ability to monitor ice motion for glaciers, ice caps and sheets, has been achieved over the past decade through the use of InSAR techniques (Joughin et al., 1996, 1998, 2002; Rignot, 1996, 2002b; Dowdeswell et al., 1999a; Luckman et al., 2002). Joughin (this volume, Chapter 74) illustrates the utility and power of interferometry for land-ice applications and this is further discussed in section 73.4.2.

One mission has been flown, however, where deriving topography was the sole objective. The shuttle radar topography mission (SRTM) was fundamentally different from the others listed in Table 73.3 as it was a short, 11-day flight specifically designed to map topography of land surfaces between 60°N and 56°S at a resolution of 30 m and with a vertical accuracy of better than 16 m. The SRTM deployed two different antennae flown on the same platform to achieve simultaneous observations with a fixed, known baseline (Rabus et al., 2003; Smith & Sandwell, 2003). For many regions the accuracy achieved by SRTM was considerably better than 16m (Rignot et al., 2003), providing the potential for the use of these data to map lower latitude glaciers and ice caps with sufficient accuracy to detect elevation changes over time (see later).

Was this article helpful?

0 0

Post a comment