|
||||||
|
|
||||||
 |
Antarctic Ice Velocity Data
|
|||||
The above overview map is a subscene from the USGS 1-km AVHRR mosaic (Ferrigno et al., 1996). To view a map of station distribution or velocity contours for the Ross Ice Shelf RIGGS project select Station Map, or Velocity Map, or use the link on the overview image.
Excerpts taken from Thomas et al., 1984 (see Related Publications).
Station positions were obtained by the 'point-positioning' method [Defense Mapping Agency, 1972]; data collected by one instrumeat were sufficient to give an absolute position by adopting the ephemeris appropriate to the observed satellite. Most stations were occupied for between 3 and 5 hours and, with precise ephemerides available for two satellites, this generally allowed sufficient time to track at least four satellite passes (4 pass estimated error ~4 m).
Many more than four satellite passes were tracked at each of the base camps, and position errors at these stations are probably less than + or - 3 m. Correction for movement during satellite tracking at these stations was made by assigning the calculated position to an epoch time at the middle of the observation period. At some field stations, fewer than four passes were monitored, and occasionally only one 'good' pass was observed. During data reduction and analysis it was possible to obtain an indication of position errors for these one-pass solutions from the consistency of the data and the geometry of the satellite orbit (W.R. MacDonald, personal communication, 1973).
Most of the stations were reoccupied after one year, but there was a two-year period before reoccupation of stations J4, K3, M3, N4, N8, 05, P6, Q6, and R.I., all near Roosevelt Island. In most cases, the position fixes included estimates of station elevation, which should not change significantly between observations. If the two values for station elevation differed by more than 15 m the position data were not used to calculate ice velocities. The average difference between the remaining pairs of surface-elevation estimates was 4.3 m. This may be the best available estimate of actual position errors.
Only about half of the station positions were fixed by satellite-Doppler tracking. Velocities at other stations were interpolated by using the measured strain rates, which give ice velocity gradients. In general, since the strain rates change slowly between neighboring stations, a good approximation to the strain rates at an arbitrary point can be obtained by linear interpotation between adjacent stations. In principle, the velocity field can be reconstructed from the strain rate field if the absolute velocity and ice shelf rotation rate are known at one point [Thomas, 1976a].
Access the tabular velocity data for the Ross Ice Shelf (RIGGS)
Data Set Source: (also see: Related Publications)
Thomas, R.H., D.R. MacAyeal, D.H. Eilers, and D.R. Gaylord, 1984. Glaciological Studies on the Ross Ice Shelf, Antarctica, 1972-1978. The Ross Ice Shelf: Glaciology and Geophysics Antarctic Research Series, Volume 42, Paper 2, 21-53.
Excerpts taken from Thomas et al., 1984 (see Related Publications).
Most of the stations with unacceptably large errors were near Roosevelt Island, or midway between Roosevelt Island and Crary Ice Rise, where there is strong convergence as two ice streams meet. Significant improvement in the velocity estimates for these stations was achieved by extrapolating from nearby control stations by using both the longitudinal strain rate between stations and the transverse velocity gradients calculated from the observed values of rotation rate and the strain rate tensor, as described in the last section.
Velocity errors at the four base camps are probably less than + or - 5 m/a; at the other control stations, they are less than + or - 15 m/a; at most of the remaining stations, they are less than + or - 30 m/a. The few stations where errors ay be significantly larger are indicated in the appendix Table A1 (Thomas et al., 1984, See Related Publications).
