San Rafael Glacier, North Patagonian Icefield
Fig. 7. Topography and glacier velocity of San Rafael Glacier
Fig. 8. Map showing location of SAR image of San Rafael Glacier
Region of San Rafael Glacier, North Patagonian Icefield
Region of Europa and Penguin Glaciers, South Patagonian Icefield
Fig. 9. Map of Penguin and Europa glaciers, South Patagonian Icefield
Fig. 10. Topography of Penguin and Europa glaciers
Fig. 11. Velocity of Penguin and Europa glaciers
Fig. 12. Profile of glacier velocity, Europa GlacierOgives
Fig. 13. X-SAR image of Ogives
Fig. 14. Graph of ogive spacing for three regions
The L-band images obtained during the special one-day repeat-pass interferometry phase accomplished during the final four days of the SRL-2 mission provided unique and unprecedented new information on Patagonian glacier topography and velocities. In addition, the SAR data imaged clear ogive sequences in a number of areas from which velocity information can be extracted. The main results are reported in four papers listed below and summarized in this report.
Rignot, E., R. Forster, and B. Isacks, 1996, Interferometric radar observations of Glaciar San Rafael, Chile, J. Glaciology, 42, 279-291.
Interferometric radar observations of Glaciar San Rafael, Chile, were collected in October 1994 by NASA's Spaceborne Imaging Radar C (SIR-C) at both L- (24cm) and C-band frequency (5.6 cm), with vertical transmit and receive polarization. The C-band data did not yield good geophysical products, because the temporal coherence of the signal was significantly reduced after 24 h. The L-band data were, however, successfully employed to map the surface topography of the icefield with a 10 m uncertainty in height, and measure ice velocity with a precision of 4 mm d-1 1.4 m a-1. The corresponding error in strain rates is 0.05a-1 at a 30 m horizontal spacing. The one-dimensional interferometric velocities were subsequently converted to horizontal displacements by assuming a flow direction and complemented by feature-tracking results near the calving front. The results provide a comprehensive view of the ice-flow dynamics of Glaciar San Rafael. The glacier has a core of rapid flow, 4.5 km in width and 3.50 in average slope, surrounded by slower-moving ice, not by rock. Ice velocity is 2.6 m d-1 or 0.95 km a -1 near the equilibrium-line altitude (1200 m), increasing rapidly before the glacier enters the narrower terminal valley, to reach 17.5 m d-1 or 6.4 km a -1 at the calving front. Strain rates are dominated by lateral shearing at the glacier margins (0.4-0.7 a -1), except for the terminal-valley section, where longitudinal strain rates average close to 1 a -1 . This spectacular longitudinal increase in ice velocity in the last few kilometers may be a fundamental feature of tidewater glaciers.
Figure 7. Topography (top) and glacier velocity (bottom) in direction of radar line of sight for the San Rafael glacier on the western side of the North Patagonian Icefield (see Figs. 2 and 8 for location). Click on the thumbnail view below to see a higher resolution (165 Kbytes) GIF file.
Figure 8. Map of North Patagonian Icefield showing location of images depicted in Fig. 7.
The data were acquired from nearly the same position in space on October 9, 10 and 11, 1994, at L- band frequency (24-cm wavelength), vertically transmitted and received polarization. The area shown in these two images is 50 kilometers by 30 kilometers in size. North is toward the upper right (see Figure 8 for orientation of swath). The top image is a digital elevation model of the scene, where color and saturation represent terrain height (between 0 meters and 2,000 meters) and brightness represents radar backscatter. Low elevations are shown in blue and high elevations are shown in pink. The digital elevation map of the glacier surface has a horizontal resolution of 15 meters and a vertical resolution of 10 meters. The bottom image is a map of ice motion parallel to the radar look direction only, which is from the top of the image. Purple indicates ice motion away from the radar at more than 6 centimeters per day; dark blue is ice motion toward or away at less than 6 cm per day; light blue is motion toward the radar of 6 cm to 20 cm per day; green is motion toward the radar of 20 cm to 45 cm per day; yellow is 45 cm to 85 cm per day; orange is 85 cm to 180 cm per day; red is greater than 180 cm (71 inches) per day. The velocity estimates are accurate to within 5 millimeters per day.
Rignot, E., R. Forster, and B. Isacks, 1996, Mapping of glacial motion and surface topography of Hielo Patagonico Norte, Chile, using satellite SAR L-band interferometry data, Ann. Glaciology, 23, 209-215.
The first topographic and ice-motion maps of the northwestern flank of Hielo Patagonico Norte (HPN, North Patagonian Icefield), in Chile, were produced using satellite synthetic-aperture interferometric radar data acquired by NASA's Spaceborne Imaging Radar (SIR-C) instrument in October 1994. The topographic map has a 10 m vertical precision with a 30 m horizontal spacing, which should be sufficient to serve as a reference for monitoring future mass changes of the icefield. The ice-motion map is accurate to within 4 mm per day (or 1 m per year ). The radar-derived surface topography and ice velocity are used to estimate the ice discharge from the accumulation area of four outlet glaciers, and the calving flux and mass balance of Glaciar San Rafael. The results demonstrate the use of SAR interferometry for monitoring glaciological parameters on a spatial and temporal scale unattainable by any other means.
Forster, R. R., 1996, Chapter 3: Interferometric radar observations of the Hielo Patagonico Sur (South Patagonian Icefield), Chile, in Investigations of Glacial Surface Properties and Ice Velocity of the South Patagonian Icefield from Synthetic Aperture Radar (SAR), PhD Thesis, Cornell University, Ithaca, NY. This study is being prepared for publication with co-authors E. Rignot and B. L. Isacks.
L- band interferometric synthetic aperture radar (INSAR) data for a portion of the South Patagonian Icefield, Chile were collected by the shuttle-based SIR- C radar over a four day span in October 1994. Three co-registered complex SAR images are used to generate phase coherence maps, a digital elevation model (DEM) and an ice velocity map. The phase coherence maps show interferometric data are lost or have poor quality (low coherence) in areas of large, absolute or relative, displacement such as near termini or within shear margins. Low coherence also resulted from a thawing event during data acquisition. The short distance between the space shuttle locations during the acquisitions (baselines) are ideal for displacement measurements, ice displacement estimates to within less then 2 cm, while the DEM is good to about 25 m. The three outlet glaciers in the radar scene act as ice streams flanked by slower moving ice within the icefield. A flow divide between two of the glaciers is mapped by locating a narrow band of near-zero ice velocity. Horizontal ice surface velocity profiles calculated along flowlines show there is a high degree of spatial variability in velocity along the center-lines with velocities reaching over 5.5 m/day. Longitudinal strain rates calculated from these velocities at the locations of the initiation of crevassing agree with theoretical values computed for ice fracture under longitudinal tension. A traverse profile of Penguin and Europa glaciers, the mean slope from the INSAR DEM, and an ice flow model are used to calculate the ice flux for both glaciers.
Figure 9. Map showing locations of location of Penguin and Europa glaciers for which interferometric SAR measurements are presented. The base map is from Aniya et al. (1996). The location of the images used for the interferometry analyses is shown by the oblique rectangle.
Figure 10. Topography of region of Penguin and Europa Glaciers, South Patagonian Icefield (see Figure 9 for location of region shown here). Click on thumbnail view below to see higher resolution (154 kb) GIF file. North is towards the lower left hand side as shown in the high resolution view of Figure 11
Since there are no topographic maps available for this area a filtered version (25 pixel median filter) of the DEM is useful and is shown in Fig. 11 draped over an amplitude SAR image. The DEM covers only the icefield and excludes the nunataks and some areas on the ice where coherence values are low or unwrapping errors occurred. The elevations range from 800 m (dark blue) near the area where phase coherence is lost on Glaciar Europa , up to 2200 m (dark red). The approximate location of the two primary topographic divides on the icefield between the Europa and Penguin glaciers are indicated as black lines and are used in the determination of ice divides.
Figure 11. Velocity along line of sight of radar (from left to right in this figure). Click on thumbnail view to see higher resolution (132 Kb) GIF file.
The ice velocity map shown in Fig. 11 contains only the component of ice motion in the direction of the radar illumination, but the map reveals the great spatial variation in velocity within this geometrically complex portion of the South Patagonian Icefield. The two purple hues represent ice motion toward the radar. The range of colors from blue (low velocity) to red (high velocity) represent motion away from the radar. The narrow range of velocities near zero which includes the transition from ice motion away to ice motion toward the radar is shown in black.
The velocity patterns of the lower potions of all three glaciers resembles those of ice streams within an ice sheet with a concentrated band of high velocity ice bordered by slower moving ice on either side. The same discovery was made by SIR-C InSAR for the San Rafael glacier (Rignot, et al., 1996) and for the Columbia Glacier, Alaska from photogrammetry techniques (Meier et al., 1985).
The ice accelerates toward the terminal edge of the glaciers and reaches a maximum line-of-sight velocity of 0.9, 1.5, and 2.2 m/day for the HPS-19, Penguin and Europa glaciers before phase coherence is lost. There are several circular regions of high velocity separated by troughs of slower velocity on the two upper tributaries of Glaciar Europa. A profile of velocity along the Europa glaciers is shown in Fig. 12.
Figure 12. Profiles of topography, velocity, and strain rate along the profile B-B' shown in fig. 11.
For areas where the flow direction and surface slope is known the radar line-of-sight velocity can be projected onto the ice surface in the direction of flow yielding an ice velocity measurement equivalent to measuring the displacement of a feature on the ice surface (Appendix D). The assumption of ice motion parallel to the ice surface is valid since the vertical component due to ablation is very small compared with the motion over most of the glacier. The horizontal component of the ice surface velocity in the flow direction (vfl-h) is calculated along the path of flow lines that are observed on the amplitude image. The location of the transects are shown in Fig. 11. The surface slope along the transects are calculated from the median filtered DEM. Longitudinal strain rate is computed as Dvfl-h / D dfl where dfl is the distance between velocity measurements along the transect.
The velocity profile shown in Fig. 12 follows a flow line which can be traced almost continuously from the amplitude image. Within the first 9 km of the transect the velocities fluctuate dramatically over short distances, reaching a maximum of 2.6 m/day near 5 km then slowing to a near constant speed of 0.4 m/day before accelerating to 3.3 m/day at the end of the transect (Fig. 12). While the magnitude and fluctuation of these velocities are not typical of mountain glaciers they are very similar to those found on the Columbia Glacier (Meier et al., 1985).
The longitudinal strain rate for this transect gives insight to the flow dynamics. The peaks at 2.3 and 4.3 km coincide with locations where the valley walls narrow the width of the glacier. This causes transverse compression of the ice which is compensated by longitudinal extension (Paterson, 1994). As the valley widens down glacier the flow tends to be more compressive.
The increased noise of the strain rate between 9 and 11 km is a result of low phase coherence in the one-day interferogram caused by the melting episode discussed above. The strain rate begins to increase at 12 km as the glacier accelerates toward the terminus.
The velocity and topographic data can be used to estimate ice fluxes
for the glaciers. The estimates for Penguin glacier, 0.51 cubic km per year,
and Europa glacier, 0.94 cubic km per year, are within the range of values
calculated for the glaciers of the North Patagonian Icefields (1.7 to 0.25
cubic km per year) by Rignot et al. (1996b).
Das, S., R.R. Forster, B. L. Isacks, and C. Malsbury, Brief report on ogives observed by SIR-C/X-SAR on the North and South Patagonian Icefields. (paper in preparation).
The most successful features we have used to measure ice velocities from SAR amplitude data are ogives, or wave bulges. Ogives are formed annually at the base of some icefalls. After formation the bulges move down glacier and a new bulge is formed the following year. The resulting train of ogives extending down glacier can be used to estimate the ice velocity by measuring the peak to peak separation between adjacent waves.
We have observed ogives on glaciers on both the North and South Patagonian Icefields, mostly high in the accumulation zones. Measurements of these ogives provides valuable velocity information in a number of previously unstudied areas. These ogives seem to be imaged best using SAR, as most do not appear on previous TM images or air photos of the same areas.
Figure 14. Graph showing spacing of ogives in three areas as noted.
Fig. 13 shows part of the high accumulation area of Viedma Glacier, one of the largest outlet glaciers draining the east side of the South Patagonian Icefield (SPI.) A train of up to 28 ogives is visible, with an initial spacing of 315 m between the first and second ogives, at the base of an icefall coming off of Viedma Volcano (49.5° S, 73.3°W). This indicates a local annual velocity of approximately 300 meters per year .As can be seen in Fig. 14, the spacing between ogives decreases as the train progresses down glacier indicating the glacier moves towards a zone of higher compressive flow (assuming the rate of flow at the icefall is fairly constant with time.)
Fig. 14 shows separations for two three trains of ogives surrounding Viedma Volcano measured on both the east and west sides of the South Patagonian Icefield. The ogives from high accumulation areas on the west side of the volcano feed Pio XI Glacier, the largest outlet glacier on the west side of the icefield. In the example shown, the east side spacings are about 300 m while the west side spacings are about 500 m, supporting the idea that the west side of the icefield (the maritime side) has a higher mass flux than the east side (continental side.)