errors for TIP are at the few cm level or less: JGM-l & 2 (Nerem et al., 1994b) and JGM3 (Tapley et al., 1994a). The latest of these model includes GPS tracking of TIP
(Melbourne et aI., 1994; Yunck et al., 1994; Schutz et aI., 1994); a technique which
promises to make significant improvements to the gravity field model when implemented
on a lower orbit. This has allowed orbits to be computed with an unprecedented total
radial accuracy of 2-3 cm (Tapley et aI., 1994b; Marshall et aI., 1995) with almost no
geographic correlation (Christensen et aI., 1994). Gravity-induced orbit errors are much
larger (> 5 cm) for other lower satellites such as Geosat, ERS-l and ERS-2; thus further
improvement of the models is required for these missions. Improving the gravity field for
a particular orbit requires using precise tracking data at that altitude and inclination.
Therefore, gravity errors for Geosat can be expected to be reduced in the future by
incorporating GPS tracking data from the Geosat Follow-On mission. Likewise, gravity
model errors for ERS-l & 2 orbit determination should be reduced by incorporating
PRARE tracking data (Wilmes and Reigber, 1989) of ERS-2 (once the PRARE tracking
network is fully deployed).
While precision orbit determination for altimeter missions has been a topic of research
for two decades, its application to SAR interferometry has emerged as an important
application only in the last few years. Currently, orbit errors are empirically removed
from SAR interferograms using an assumption of no motion at the margins of the
interferogram (Massonnet and Rabaute, 1993), but future applications will clearly require
precision orbit determination at the few cm level, and this may in turn place additional
requirements on improving gravity models.
GEOPHYSICAL APPLICATIONS
There are a wide variety of geophysical studies that require detailed knowledge of the
gravitational field. The external gravitational field is a reflection of the planet's internal
density variations; thus gravity field models provide one of the primary constraints for
determining the internal structure of a planetary body. The long-wavelengths of the
gravity field are determined almost exclusively from satellite tracking data, but
determining the short-wavelengths is accomplished mainly through surface gravity and
satellite altimeter data.
Surface gravity data provide most of the short-wavelength coverage over the land. For
global models, the 360 x 360 OSU91A model (Rapp et al., 1991) is considered to be the
current state-of-the-art. Regional models provide much higher spatial resolution, such as
the GEOID93 model for North America (Milbert and Schultz, 1993). Both of these
models depend heavily on accurate surface gravity measurements to provide wellresolved fields, as do the JGM models (Nerem et al., 1994b; Tapley et al., 1994a). A
good example of the geophysical application of these models over the continents is the
work of McKenzie (1995) who used a filtered version of the JGM-2 geoid model to
analyze intraplate dynamics in Africa. A collaboration between NASA/GSFC and the
Defense Mapping Agency (DMA) will result in the release of a wealth of previously
classified surface gravity data in 1996, and the 360 x 360 gravity model resulting from
this effort should provide considerable improvement in our knowledge of the gravity field
over the continents (Nerem et al., this issue) and polar ocean regions.
Satellite altimetry is the principal technique used to define the short wavelengths of the
marine gravity field in current models (Rapp et aI., 1991; Rapp and Basic, 1992). ERS-l
altimetry and Geosat Geodetic Mission (GM) altimetry south of 30 0 S latitude have led to
significant advances in our knowledge of the short wavelengths of the marine gravity
field (McAdoo and Marks, 1992; Sandwell, 1992; Livermore et al., 1994). The model of
the gravity field in the Arctic Ocean has been significantly improved using ERS-l
altimeter measurements over sea ice through the innovative work of Laxon and McAdoo
(1994). The recent declassification of the entire Geosat Geodetic Mission (GM) altimeter
data set should provide some improvement to the marine gravity field north of 30 0 S
3
(Melbourne et aI., 1994; Yunck et al., 1994; Schutz et aI., 1994); a technique which
promises to make significant improvements to the gravity field model when implemented
on a lower orbit. This has allowed orbits to be computed with an unprecedented total
radial accuracy of 2-3 cm (Tapley et aI., 1994b; Marshall et aI., 1995) with almost no
geographic correlation (Christensen et aI., 1994). Gravity-induced orbit errors are much
larger (> 5 cm) for other lower satellites such as Geosat, ERS-l and ERS-2; thus further
improvement of the models is required for these missions. Improving the gravity field for
a particular orbit requires using precise tracking data at that altitude and inclination.
Therefore, gravity errors for Geosat can be expected to be reduced in the future by
incorporating GPS tracking data from the Geosat Follow-On mission. Likewise, gravity
model errors for ERS-l & 2 orbit determination should be reduced by incorporating
PRARE tracking data (Wilmes and Reigber, 1989) of ERS-2 (once the PRARE tracking
network is fully deployed).
While precision orbit determination for altimeter missions has been a topic of research
for two decades, its application to SAR interferometry has emerged as an important
application only in the last few years. Currently, orbit errors are empirically removed
from SAR interferograms using an assumption of no motion at the margins of the
interferogram (Massonnet and Rabaute, 1993), but future applications will clearly require
precision orbit determination at the few cm level, and this may in turn place additional
requirements on improving gravity models.
GEOPHYSICAL APPLICATIONS
There are a wide variety of geophysical studies that require detailed knowledge of the
gravitational field. The external gravitational field is a reflection of the planet's internal
density variations; thus gravity field models provide one of the primary constraints for
determining the internal structure of a planetary body. The long-wavelengths of the
gravity field are determined almost exclusively from satellite tracking data, but
determining the short-wavelengths is accomplished mainly through surface gravity and
satellite altimeter data.
Surface gravity data provide most of the short-wavelength coverage over the land. For
global models, the 360 x 360 OSU91A model (Rapp et al., 1991) is considered to be the
current state-of-the-art. Regional models provide much higher spatial resolution, such as
the GEOID93 model for North America (Milbert and Schultz, 1993). Both of these
models depend heavily on accurate surface gravity measurements to provide wellresolved fields, as do the JGM models (Nerem et al., 1994b; Tapley et al., 1994a). A
good example of the geophysical application of these models over the continents is the
work of McKenzie (1995) who used a filtered version of the JGM-2 geoid model to
analyze intraplate dynamics in Africa. A collaboration between NASA/GSFC and the
Defense Mapping Agency (DMA) will result in the release of a wealth of previously
classified surface gravity data in 1996, and the 360 x 360 gravity model resulting from
this effort should provide considerable improvement in our knowledge of the gravity field
over the continents (Nerem et al., this issue) and polar ocean regions.
Satellite altimetry is the principal technique used to define the short wavelengths of the
marine gravity field in current models (Rapp et aI., 1991; Rapp and Basic, 1992). ERS-l
altimetry and Geosat Geodetic Mission (GM) altimetry south of 30 0 S latitude have led to
significant advances in our knowledge of the short wavelengths of the marine gravity
field (McAdoo and Marks, 1992; Sandwell, 1992; Livermore et al., 1994). The model of
the gravity field in the Arctic Ocean has been significantly improved using ERS-l
altimeter measurements over sea ice through the innovative work of Laxon and McAdoo
(1994). The recent declassification of the entire Geosat Geodetic Mission (GM) altimeter
data set should provide some improvement to the marine gravity field north of 30 0 S
3