coincident profile , ice-free
-~~~---------~U~------~-M~------~_~M-------=~~----~
Latitude (deg)
Fig. 5. ERS-l altimeter profile over ice-covered ocean (top) shows high noise level
due to failure of the standard tracking algorithm to identify the leading edge of the radar
echo. Retracking of the full echo waveform reduces the noise (middle) so the profile
agrees with results obtained during a ice-free period (bottom).
Accuracy and Resolution
The accuracy and resolution of gravity grids constructed using this approach were
recently established through a comparison with accurate shipboard gravity measurements
[Neumann et aI., 1993]. For a small region along the southern Mid-Atlantic Ridge where
there is a dense shipboard survey and large gravity anomalies (140 mgal total variation), the
RMS difference is 7-8 mgal. We have found that individual ship profiles show similar
RMS differences. The particular ship profile shown in Figure 6 traverses the South
Atlantic Ocean along a Geosat altimeter track where 62 repeat profiles are available for
stacking. The ship track deviates from the Geosat trackline at -20.5° latitude to avoid a
small island. The mean difference between the Geosat-derived gravity and the shipboard
gravity is only -0.32 mgal and the RMS difference is 3.57 mgal. We expect that this
stacked Geosat profile reflects the best accuracy that can be achieved with the satellite
altimeter method. A 3.5 mgal error corresponds to a relative height accuracy of only 17
mm over a distance of 5 km (i.e., 114 of the resolution wavelength). Considering that
typical surface wave heights are a meter or more tall, this is a remarkable achievement.
When repeating profiles are stacked, the vertical deflection error decreases as the square
root of the number of profiles used in the stack [Yale et al., 1995]. Thus to make
substantial improvements beyond what will be achieved using both dense ERS-l and
Geosat altimeter profiles will require a dedicated 5-10 year satellite altimeter mission.
17
-~~~---------~U~------~-M~------~_~M-------=~~----~
Latitude (deg)
Fig. 5. ERS-l altimeter profile over ice-covered ocean (top) shows high noise level
due to failure of the standard tracking algorithm to identify the leading edge of the radar
echo. Retracking of the full echo waveform reduces the noise (middle) so the profile
agrees with results obtained during a ice-free period (bottom).
Accuracy and Resolution
The accuracy and resolution of gravity grids constructed using this approach were
recently established through a comparison with accurate shipboard gravity measurements
[Neumann et aI., 1993]. For a small region along the southern Mid-Atlantic Ridge where
there is a dense shipboard survey and large gravity anomalies (140 mgal total variation), the
RMS difference is 7-8 mgal. We have found that individual ship profiles show similar
RMS differences. The particular ship profile shown in Figure 6 traverses the South
Atlantic Ocean along a Geosat altimeter track where 62 repeat profiles are available for
stacking. The ship track deviates from the Geosat trackline at -20.5° latitude to avoid a
small island. The mean difference between the Geosat-derived gravity and the shipboard
gravity is only -0.32 mgal and the RMS difference is 3.57 mgal. We expect that this
stacked Geosat profile reflects the best accuracy that can be achieved with the satellite
altimeter method. A 3.5 mgal error corresponds to a relative height accuracy of only 17
mm over a distance of 5 km (i.e., 114 of the resolution wavelength). Considering that
typical surface wave heights are a meter or more tall, this is a remarkable achievement.
When repeating profiles are stacked, the vertical deflection error decreases as the square
root of the number of profiles used in the stack [Yale et al., 1995]. Thus to make
substantial improvements beyond what will be achieved using both dense ERS-l and
Geosat altimeter profiles will require a dedicated 5-10 year satellite altimeter mission.
17