Content uploaded by Dietmar Müller
Author content
All content in this area was uploaded by Dietmar Müller on Oct 05, 2014
Content may be subject to copyright.
MARINE GEOPHYSICS
New global marine gravity model
from CryoSat-2 and Jason-1 reveals
buried tectonic structure
David T. Sandwell,
1
*R. Dietmar Müller,
2
Walter H. F. Smith,
3
Emmanuel Garcia,
1
Richard Francis
4
Gravity models are powerful tools for mapping tectonic structures, especially in the deep
ocean basins where the topography remains unmapped by ships or is buried by thick
sediment. We combined new radar altimeter measurements from satellites CryoSat-2
and Jason-1 with existing data to construct a global marine gravity model that is two times
more accurate than previous models. We found an extinct spreading ridge in the Gulf of
Mexico, a major propagating rift in the South Atlantic Ocean, abyssal hill fabric on
slow-spreading ridges, and thousands of previously uncharted seamounts. These
discoveries allow us to understand regional tectonic processes and highlight the
importance of satellite-derived gravity models as one of the primary tools for the
investigation of remote ocean basins.
Fracture zones (FZs) spanning the ocean ba-
sins reveal the breakup of the continents
and the geometry of sea-floor spreading (1).
The exact intersection points of the FZs
along conjugate continental margins are
used for precise reconstruction of the continents
(2–4). These FZ intersections are commonly
buried by several kilometers of sediments that
flow off the continents to fill the voids created
by the thermal subsidence of the rifted margins
(5). This sediment cover extends hundreds to
thousands of kilometers out onto the oceanic
lithosphere, resulting in a relatively flat and
featureless sea floor. Reflection seismic profiles
can reveal the underlying basement topogra-
phy of the FZs, but the data coverage is usually
insufficient to map out the intersections. In areas
of thin sediment cover, the topographic ridges
and troughs along the FZs produce large gravity
anomalies that are easily traced across the ocean
basins (Fig. 1). However, when the topography
becomes buried by sediment, the original density
contrast of the sea-floor topography is reduced,
resulting in more-subdued, and sometimes sign-
reversed, gravity signatures (6). Moreover, as the
lithosphere ages and cools, the sea floor subsides,
causing a blurring of the gravity anomalies; smaller
wavelengths of the gravity field become less well-
resolved with increasing water depth. Previous
global marine gravity models derived from satel-
lite altimetry had sufficient accuracy and coverage
to map all FZs in unsedimented sea floor (7), but
the 3 to 5 mGal of gravity noise blurred the small
signatures of sediment-covered topography such
as seamounts and FZs. Here we report on a new
global marine gravity model having ~2-mGal ac-
curacy that is providing a dramatically improved
resolution of the 80% of the sea floor that remains
uncharted or is buried beneath thick sediment.
Gravity-field accuracy derived from satellite al-
timetry depends on three factors: altimeter range
precision, spatial track density, and diverse track
orientation. Two altimeter data sets with high
track density have recently become available
(CryoSat-2 and Jason-1) to augment the older
altimeter data (Geosat and ERS-1), resulting in
improvement by a factor 2 to 4 in the global
marine gravity field. Their newer radar technol-
ogy results in a 1.25-times improvement in range
precision that maps directly into gravity-field
improvement (8). The new altimeters also con-
tributemorethan70monthsofdata,ascom-
pared with the 31 months provided by the older
satellites. CryoSat-2 has provided the most dense
track coverage, because although it has a nominal
369-day repeat orbit period, the ground tracks are
allowed to drift within a 5-km band, so after 4 years
in orbit it has provided a nominal track spacing of
about 2.5 km. Jason-1 provided 14 months of dense
track coverage during its geodetic phase, resulting
in a track spacing of 7.5 km.
Most of the improvement in the altimeter-
derived gravity field occurs in the 12- to 40-km
wavelength band, which is of interest for the in-
vestigation of structures as small as 6 km. The
current version of the altimeter-derived gravity
field has an accuracy of about 2 mGal (8). Unlike
terrestrial gravity, where coverage is uneven, these
accuracies are available over all marine areas and
large inland bodies of water, so this gravity pro-
vides an important tool for exploring the deep
ocean basins. At scales smaller than 200 km,
variations in marine gravity primarily reflect
SCIENCE sciencemag.org 3OCTOBER2014•VOL 346 ISSUE 6205 65
RESEARCH |REPORTS
1
Scripps Institution of Oceanography, La Jolla, CA 92093,
USA.
2
EarthByte Group, School of Geosciences, University of
Sydney, New South Wales, Australia.
3
Laboratory for Satellite
Altimetry, National Oceanic and Atmospheric Administration
(NOAA), College Park, MD 20740, USA.
4
European Space
Agency/European Space Research and Technology Centre,
Keplerlaan 1, 2201AZ Noordwijk, Netherlands.
*Corresponding author. E-mail: dsandwell@ucsd.edu
Fig. 1. Ocean gravity maps. (A) New marine gravity anomaly map derived
from satellite altimetry reveals tectonic structures of the ocean basins in
unprecedented detail, especially in areas covered by thicksediments. Land areas
show gravity anomalies from Earth Gravitational Model 2008 (15). (B)VGGmap
derived from satellite altimetry highlights FZs crossing the South Atlantic Ocean
basin (yellow line). Areas outlined in red are small-amplitude anomalies in areas
where thick sediment has diminished the gravity signal of the basement to-
pography.The full-resolution gravity anomaly and VGG models can be viewed in
Google Earth using the following files: ftp://topex.ucsd.edu/pub/global_grav_
1min/global_grav.kmz and ftp://topex.ucsd.edu/pub/global_grav_1min/global_
grav_gradient.kmz.The grids are available in the supplementary material, as well
as at the following FTP site: ftp://topex.ucsd.edu/pub/global_grav_1min.
sea-floor topography generated by plate tecton-
ics such as ridges, FZs, and abyssal hills. Many
FZscannowbetracedmuchmorecloselytothe
continental margins, and one can also better in-
terpret buried, migrating, unstable FZs, which has
the potential to improve the use of FZs as tie
points for reconstructions of the boundaries be-
tween continental fit reconstructions (Fig. 1). In
addition to FZs, there are other tectonic features
associated with continental margins, such as the
boundaries between continental and oceanic crust
[continent-ocean boundaries (COBs)], that can
now be mapped in greater detail.
The first example (Fig. 2) is in the Gulf of
Mexico, where thick sediments obscure the FZs
and extinct ridges. Reconstruction models pro-
vide the overall framework of counterclockwise
rotation of the Yucatan plate with respect to
North America, as well as a generalized position
for the COB (9). The new vertical gravity gradient
images confirm and refine the positions of these
tectonic boundaries. Extinct spreading ridges pro-
duce a negative gravity signature, because the
relatively high-density sediment cover largely
cancels the positive gravity effect of the topo-
graphic ridge, leaving the negative gravity sig-
nature of the compensating Moho topography
(6). In this region, the Moho is more than 15 km
beneath the sea surface, so the effects of upward
continuation reduce and smooth the anomaly.
The second example is on the African ridge
flank, where the new data reveal a major tectonic
feature that was not visible in previous satellite
gravity data sets because of high-frequency noise.
The newly discovered feature is a set of tectonic
lineaments roughly between 8°S and 12°S, strik-
ing northwest-southeast and obliquely dissected
by individual en-echelon faults, stretching from
the Bodo Verde Fracture Zone in the north into
the middle of the Cretaceous Magnetic Quiet Zone
at its southeastern extension (Fig. 1b). This feature
isabout800kmlongand100kmwideanddoes
not follow either the azimuth of nearby sea-floor
isochrons or FZs. A reconstruction of this feature
at magnetic chron 34 [83.5 million years ago
(Ma)] (Fig. 3) reveals that it has a mirror-image
counterpart on the South American plate, but
this conjugate feature is represented only by a
relatively faint gravity lineament (Fig. 3). This
feature is visible in the filtered vertical gravity
gradient image, marking a boundary between
66 3OCTOBER2014•VOL 346 ISSUE 6205 sciencemag.org SCIENCE
Fig. 2. Gulf of Mexico VGG. (A) Uninterpreted. (B) Our interpretation of tectonic structures, after Pindell and Kennan (9). TheVGG reveals subtle signatures of
the extinct spreading ridges and FZs as well as a significant change in amplitude across the boundary between continental and oceanic crust (COBs).This is a
Mercator projection; grayscale saturates at T20 eotvos units.
Fig. 3. South Atlantic filtered VGG. Reconstructed at chron 34 (83.5 Ma, orthographic projection) with
Africa fixed (16). Major tectonic and volcanic sea-floor features and offshore sedimentary basins are
labeled. The mid-ocean ridge is outlined in red, the extinct Abimael spreading ridge is shown as a dashed
red line, and the reconstructed position of the Cardno hot spot (CS) is outlined by a red star. Most of the
sea floor shown in this reconstruction was formed during the Cretaceous Normal Superchron. Also note
that the extinct Abimael spreading ridge between the Santos Basin and Sao Paulo Plateau offshore of
Brazil is now visible as a negativeVGGanomaly (dashed red line), as compared to previous interpretations
(17,18). This region is of great interest for oil and gas exploration, as it is one of the most extensive
deepwater oil and gas frontiers globally, with several recent discoveries (19).
RESEARCH |REPORTS
swaths of differently textured sea-floor fabric to
the east and west of the lineament (Fig. 3). The
geometry of the two features suggests that they
form a pair of an extinct ridge (on the African
side) and a pseudofault (on the South American
side), created by a northward ridge propagation
episode between ~100 and 83 Ma. An absolute
hot spot–based plate reconstruction using the
rotation parameters from O’Neill et al.(10)in-
dicates that the Cardno hot spot (Fig. 3) may
have been situated not far north of the northern
tip of the ridge propagator, where it abuts the
BodoVerdeFractureZone;thisiswherethe
propagator came to a halt. These observations
conform with the inference that ridges have a
tendency to propagate toward hot spots/plumes
and that propagation events and resulting spread-
ing asymmetries are frequently contained within
individual spreading corridors bounded by FZs
(11). The existence of major previously unknown
ridge propagation events will also be relevant for
interpreting marine magnetic anomaly sequences
during the Cretaceous Normal Superchron on
conjugate ridge flanks (12).
One of the most important uses of this new
marine gravity field will be to improve the esti-
matesofsea-floordepthinthe80%oftheoceans
having no depth soundings. The most accurate
method of mapping sea-floor depth uses a mul-
tibeam echosounder mounted on a large research
vessel. However, even after 40 years of mapping
by hundreds of ships, one finds that more than
50% of the ocean floor is more than 10 km away
from a depth measurement. Between the sound-
ings, the sea-floor depth is estimated from marine
gravity measurements from satellite altimetry
(13). This method works best on sea floor where
sediments are thin, resulting in a high correla-
tion between sea-floor topography and gravity
anomalies in the 12-km–to–160-km wavelength
band. The shorter wavelengths are attenuated
because of Newton’s inverse square law, whereas
the longer wavelengths are partially cancelled by
the gravity anomalies caused by the isostatic
topography on the Moho (13). The abyssal hill
fabric created during the sea-floor spreading
process has characteristic wavelengths of 2 to
12 km, so it is now becoming visible in the ver-
tical gravity gradient (VGG) models, especially
on the flanks of the slower-spreading ridges
(14). Additionally, seamounts between 1 and 2 km
tall, which were not apparent in the older gravity
models, are becoming visible in the new data.
As CryoSat-2 continues to map the ocean sur-
face topography, the noise in the global marine
gravity field will decrease. Additional analysis
of the existing data, combined with this steady
decrease in noise, will enable dramatic improve-
ments in our understanding of deep ocean tec-
tonic processes.
REFERENCES AND NOTES
1. J. T. Wilson, Nature 207, 343–347 (1965).
2. S. Cande, J. LaBrecque, W. Haxby, J. Geophys. Res. Solid Earth
93, 13479–13492 (1988).
3. C. Heine, J. Zoethout, R. D. Müller, Solid Earth 4, 215–253
(2013).
4. L. A. Lawver, L. M. Gahagan, I. W. Dalziel, Mem. Natl. Inst. Polar
Res. 53,214–229 (1998).
5. M. S. Steckler, A. B. Watts, Earth Planet. Sci. Lett. 41,1–13
(1978).
6. C. S. Liu, D. T. Sandwell, J. R. Curray, J. Geophys. Res. 87,
7673–7686 (1982).
7. K. Matthews, R. D. Müller, P. Wessel, J. M. Whittaker,
J. Geophys. Res. Solid Earth 116,1–28 (2011).
8. Materials and methods are available as supplementary
materials on Science Online.
9. J. Pindell, L. Kennen, in The Geology and Evolution of the
Region Between North and South America, K. James,
M. A. Lorente, J. Pindell, Eds. (Special Publication, Geological
Society of London, London, 2009), vol. 328, pp. 1–55.
10. C. O'Neill, R. D. Müller, B. Steinberger, Geochem. Geophys.
Geosyst. 6, Q04003 (2005).
11. R. D. Müller, W. R. Roest, J. Y. Royer, Nature 396, 455–459
(1998).
12. R. Granot, J. Dyment, Y. Gallet, Nat. Geosci. 5, 220–223
(2012).
13. W. H. F. Smith, D. T. Sandwell, Science 277, 1956–1962
(1997).
14. J. A. Goff, W. H. F. Smith, K. A. Marks, Oceanography 17,24–37
(2004).
15. N. K. Pavlis, S. A. Holmes, S. C. Kenyon, J. K. Factor,
J. Geophys. Res. 117, B04406 (2012).
16. M. Seton et al., Earth Sci. Rev. 113, 212–270 (2012).
17. W. Mohriak, M. Nóbrega, M. Odegard, B. Gomes, W. Dickson,
Petrol. Geosci. 16, 231–245 (2010).
18. I. Scotchman, G. Gilchrist, N. Kusznir, A. Roberts, R. Fletcher, in
The Breakup of the South Atlantic Ocean: Formation of Failed
Spreading Axes and Blocks of Thinned Continental Crust in the
Santos Basin, Brazil and Its Consequences For Petroleum
System Development (Petroleum Geology Conference Series,
Geological Society of London, London, 2010), pp. 855–866.
19. W. U. Mohriak, P. Szatmari, S. Anjos, Geol. Soc. London Spec.
Publ. 363, 131–158 (2012).
ACKNO WLEDGM ENTS
The CryoSat-2 data were provided by the European Space Agency,
and NASA/Centre National d"Etudes Spatiales provided data
from the Jason-1 altimeter. This research was supported by
NSF (grant OCE-1128801), the Office of Naval Research (grant
N00014-12-1-0111), the National Geospatial Intelligence Agency
(grant HM0177-13-1-0008), the Australian Research Council
(grant FL099224), and ConocoPhillips. Version 23 of global grids
of the gravity anomalies and VGG can be downloaded from the
supplementary materials and also at the following FTP site: ftp://
topex.ucsd.edu/pub/global_grav_1min. The manuscript contents
are the opinions of the authors, and the participation of W.H.F.S.
should not be construed as indicating that the contents of the
paper are a statement of official policy, decision, or position on
behalf of NOAA or the U.S. government.
SUPPLEMENTARY MATERIALS
www.sciencemag.org/content/346/6205/65/suppl/DC1
Supplementary Text
Figs. S1 and S2
References (20–33)
2 July 2014; accepted 2 September 2014
10.1126/science.1258213
NANOPHOTONICS
Chiral nanophotonic waveguide
interface based on spin-orbit
interaction of light
Jan Petersen, Jürgen Volz,*Arno Rauschenbeutel*
Controlling the flow of light with nanophotonic waveguides has the potential of
transforming integrated information processing. Because of the strong transverse
confinement of the guided photons, their internal spin and their orbital angular
momentum get coupled. Using this spin-orbit interaction of light, we break the mirror
symmetry of the scattering of light with a gold nanoparticle on the surface of a
nanophotonic waveguide and realize a chiral waveguide coupler in which the handedness
of the incident light determines the propagation direction in the waveguide. We control
the directionality of the scattering process and can direct up to 94% of the incoupled
light into a given direction. Our approach allows for the control and manipulation of
light in optical waveguides and new designs of optical sensors.
The development of integrated electronic cir-
cuits laid the foundations for the informa-
tion age, which fundamentally changed
modern society. During the past decades,
a transition from electronic to photonic in-
formation transfer took place, and nowadays,
nanophotonic circuits and waveguides promise
to partially replace their electronic counterparts
and to enable radically new functionalities (1–3).
The strong confinement of light provided by such
waveguides leads to large intensity gradients on
the wavelength scale. In this strongly nonparaxial
regime, spin and orbital angular momentum of
light are no longer independent physical quan-
tities but are coupled (4,5). In particular, the spin
dependsonthepositioninthetransverseplane
and on the propagation direction of light in the
waveguide—an effect referred to as spin-orbit in-
teraction of light (SOI). This effect holds great
promises for the investigation of a large range of
physical phenomena such as the spin-Hall effect
(6,7) and extraordinary momentum states (8)
and has been observed for freely propagating light
fields (9,10) in the case of total internal reflection
(11,12), in plasmonic systems (13–15), and for
radio frequency waves in metamaterials (16). Re-
cently, it has been demonstrated in a cavity-
quantum electrodynamics setup in which SOI
fundamentally modifies the coupling between a
single atom and the resonator field (17).
SCIENCE sciencemag.org 3OCTOBER2014•VOL 346 ISSUE 6205 67
RESEARCH |REPORTS
Vienna Center for Quantum Science and Technology, TU
Wien–Atominstitut, Stadionallee 2, 1020 Vienna, Austria.
*Corresponding author. E-mail: jvolz@ati.ac.at (J.V.); arno.
rauschenbeutel@ati.ac.at (A.R.)
www.sciencemag.org/content/346/6205/65/suppl/DC1
Supplementary Materials for
New global marine gravity model from CryoSat-2 and Jason-1 reveals
buried tectonic structure
David T. Sandwell,* R. Dietmar Müller, Walter H. F. Smith, Emmanuel Garcia, Richard
Francis
*Corresponding author. E-mail: dsandwell@ucsd.edu
Published 3 October 2014, Science 346, 65 (2014)
DOI: 10.1126/science.1258213
This PDF file includes:
Supplementary Text
Figs. S1 and S2
References (20–33)
2
Supplementary Text
Gravity Anomaly Recovery
Gravity anomalies are small differences in the pull of gravity associated with lateral
variations in mass. The best approach to measuring marine gravity is to mount a very
precise accelerometer on a ship. Unfortunately this ship coverage of the oceans is very
sparse (20). A second, now equally precise approach is to use an orbiting radar to
measure the topography of the ocean surface, which is nearly an equipotential surface.
The methods for recovering maps of marine gravity anomaly from radar altimeter data
are discussed in many previous publications [e.g., (21-24)]. Some of the key technology
developments related to this new marine gravity model are provided in two recent
publications (25, 26). For our investigation of crustal structure we use Laplace’s equation
to construct the first and second vertical derivatives of the potential called gravity
anomaly and vertical gravity gradient, respectively. Images of these two fields over the
South Atlantic Basin are shown in Fig. 1. The full resolution maps are best viewed using
a computer display program such as Google Earth. The reader can download two small
KMZ files to bring these full resolution maps into their computer. In addition they can
download the gridded files to construct custom maps from
http://topex.ucsd.edu/WWW_html/mar_grav.html .
Improved Radar Technology
The most important contribution of the new altimeters is related to a 1.25 times
improvement in range precision (26). This improvement is mainly related to an increase
in the pulse repetition frequency (PRF) of the newer altimeters with respect to the older
altimeters. The coherent nature of the radar signal results in speckle in the echoes, which
masks the echo waveform and leads to imprecision in retrieval of its parameters. This can
be alleviated by averaging successive echoes, but only up to the point that they become
correlated, the onset of which has been generally assumed at a PRF of somewhat above
2 kHz at the common transmitter frequency of 13.5 GHz (27). The newer altimeters
CryoSat-2 and Jason-1 have PRFs of 1950 Hz and 2060 Hz, respectively while the older
instruments were technologically limited to lower values of 1020 Hz. Theoretically this
approximate doubling of PRF should result in a
2=1.41
improvement in range
precision; the actual improvement is somewhat smaller (1.25) perhaps reflecting the onset
of echo correlation at the 2 kHz PRF. Nevertheless, this improvement in range precision
maps directly into an improvement in gravity field accuracy.
CryoSat-2 was also operated in a new Synthetic Aperture Radar (SAR) mode over
very limited areas of the oceans. This mode has a much higher PRF of 18.2 kHz and the
highly correlated echoes are summed coherently in bursts of 64 pulses to form a long
synthetic aperture. This enhances along-track resolution in the form of a set of narrow
beams distributed in the along-track direction (27–30). Unlike the conventional pulse-
width limited geometry, the resulting echo waveforms have useful information in both
the leading and trailing edges. This, together with an increase in the effective number of
independent samples resulting from the SAR technique, reduces the height noise by a
factor of ~1.4 compared to conventional LRM (31). Comparison of height noise
performance (26) indeed shows this expected improvement for CryoSat-2’s SAR but
similar gains for pulse-width limited echoes are obtained by a two-pass processing
3
scheme in which the slowly varying ocean wave-height is first estimated and smoothed
and then excluded from the estimation process in the second pass. These results show that
CryoSat-2’s LRM performs slightly better than Jason-1 (which is already excellent),
despite its reduced PRF. Much of the design of the two radars is common but it is likely
that the improvements introduced for CryoSat-2’s mission, particularly the higher
transmitter power needed for operation over sloping ice surfaces and the extreme phase
stability required for SAR interferometry, are contributing to this performance.
Coherence-Enhancing Filter
Despite the advances in satellite gravity anomaly image quality described in this
paper, some high-frequency noise remains. In order to further improve the interpretation
of linear tectonic features seen in the new vertical gravity gradient images, we have
applied a filtering technique called coherence-enhancing diffusion to a selected region in
the South Atlantic (Fig. 3) (32). This filter combines anisotropic diffusion (a low-pass
filter) with texture analysis, such that a diffusion tensor is computed from the local image
structure so that the diffusion is parallel to linear features in the data. This type of filter
has been successfully applied for enhancing noisy seismic reflection images to facilitate
improved tracking of seismic horizons (33), and is applied to vertical gravity gradient
data here for the first time. While high-frequency noise has been suppressed, linear,
coherent seafloor structures have been enhanced. In particular the internal en-echelon
structure of the extinct mid-ocean ridge on the African Plate has been enhanced, while
the juxtaposed differences in seafloor structure west and east of the conjugate pseudofault
have been enhanced as well. Deeply buried linear structures of the Santos Basin and Sao
Paulo Plateau offshore Brazil have been equally enhanced (Fig. 3), illustrating that
improved satellite data combined with well-targeted filtering have a great potential to
reveal previously hidden structures on abyssal plains and along passive margins.
Gravity Anomaly Uncertainty
We estimated the uncertainty in the gravity by calculating the rms difference in slope
between individual altimeter profiles and the mean north and east slopes used to compute
gravity (Fig S1.) The uncertainties were calibrated by comparisons with shipboard data
from two completely different proprietary sources. First we computed the rms difference
between the altimeter-derived gravity and more accurate shipboard gravity in a small
region in the Gulf of Mexico. The ship data, provided by EDCON Inc., were collected on
a very fine grid and have an rms crossover error of 0.5 mGal (25). For this first
comparison we found an rms difference of 1.60 mGal. In the second case, the altimeter-
derived gravity data were compared with 30 million of the best shipboard data by the
National Geospatial Intelligence Agency (NGA personal communication) resulting in an
rms difference of 2.6 mGal. The rms difference is somewhat higher (3.6 mGal) in
shallow areas (< 1 km) and somewhat lower (2.3 mGal) in deeper areas (3 - 6 km). On
average, the NGA ship data have an rms error of 1 - 2 mGal. Assuming the mean rms
error is 1.6 mGal then the mean rms error in the altimeter-derived gravity is ~2 mGal in
agreement with the calibration derived from the Gulf of Mexico comparison. As shown
in our previous study (25) most of the error reduction between this new gravity model
and the older models occur in the 12 to 40 km wavelength band.
4
The noise reduction over the short wavelength band provides a dramatic improvement
in the clarity of the vertical VGG signals. We used the Gulf of Mexico region to
illustrate this noise reduction (Fig. S2). The upper plot shows the VGG derived from
only Geosat and ERS-1 altimetry data (24) while the middle plot shows also includes the
new measurements from CryoSat-2 and Jason-1. The reduction in noise between the old
and new models reveals the extinct spreading ridges and transforms as well as the
continent ocean boundary. One can also see some of these features in the old model but
they are largely obscured by noise. The difference between the new and old model (Fig.
S2 c) reveals the noise in the old model. The rms difference between the two models is
6.9 eotvos units and in terms of gravity anomaly (not shown) the rms difference is 2.2
mGal. The rms differences are zero over land, where the VGG and gravity anomaly are
set by the EGM08 model (9). Differences are greatest near the shorelines where the raw
altimeter waveforms are sometimes contaminated by stray echoes off the land. To
understand the contribution of each of the satellite data sets to the accuracy of the gravity
grid, we have constructed a suite of gravity models after removing one of the non-repeat
data sets. We find that because the four non-repeat altimeter data sets have differing
orbital inclinations and noise levels they are all important for achieving the best overall
accuracy.
5
Fig. S1. Gravity error. (a) Estimated error in marine gravity anomaly to 81 degrees
latitude. Color scale ranges from 0-10 mGal. Relatively larger noise occurs in areas of
high mesoscale variability such as the Gulf Stream. Sharp changes in gravity noise occur
at the maximum inclination of Jason-1, Geosat, ERS/Envisat ground tracks. (b)
Longitude-averaged gravity error versus latitude. Noise is higher in polar regions due to
lower track density and altimeter noise caused by sea ice. (Note CryoSat-2 collects data
to 88 degrees latitude but this plot only extends to 81.)
6
Fig. S2. Gulf of Mexico
VGG. (a) Old VGG model
based on Geosat and ERS-1.
(b) New VGG model also
includes data from CryoSat-
2 and Jason-1. (c)
Difference between the two
models plotted using the
same greyscale shows noise
in the old VGG model.
Mercator projection, grey
scale saturates at +/- 20
eotvos units.
References and Notes
1. J. T. Wilson, A new class of faults and their bearing on continental drift. Nature 207,
343–347 (1965). doi:10.1038/207343a0
2. S. Cande, J. LaBrecque, W. Haxby, Plate kinematics of the South Atlantic: Chron C34
to present. J. Geophys. Res. Solid Earth 93, 13479–13492 (1988).
doi:10.1029/JB093iB11p13479
3. C. Heine, J. Zoethout, R. D. Müller, Kinematics of the South Atlantic rift. Solid Earth
4, 215–253 (2013). doi:10.5194/se-4-215-2013
4. L. A. Lawver, L. M. Gahagan, I. W. Dalziel, A tight fit—Early Mesozoic Gondwana, a
plate reconstruction perspective. Mem. Natl. Inst. Polar Res. 53, 214–229 (1998).
5. M. S. Steckler, A. B. Watts, Subsidence of the Atlantic-type continental margin off
New York. Earth Planet. Sci. Lett. 41, 1–13 (1978). doi:10.1016/0012-
821X(78)90036-5
6. C. S. Liu, D. T. Sandwell, J. R. Curray, The negative gravity field over the 85°E
Ridge. J. Geophys. Res. 87, 7673–7686 (1982). doi:10.1029/JB087iB09p07673
7. K. Matthews, R. D. Müller, P. Wessel, J. M. Whittaker, The tectonic fabric of the
ocean basins. J. Geophys. Res. Solid Earth 116, 1–28 (2011).
doi:10.1029/2011JB008413
8. Materials and methods are available as supplementary materials on Science Online.
9. J. Pindell, L. Kennen, in The Geology and Evolution of the Region Between North and
South America, K. James, M. A. Lorente, J. Pindell, Eds. (Special Publication,
Geological Society of London, London, 2009), vol. 328, pp. 1–55.
10. C. O'Neill, R. D. Müller, B. Steinberger, On the uncertainties in hot spot
reconstructions and the significance of moving hot spot reference frames.
Geochem. Geophys. Geosyst. 6, Q04003 (2005). doi:10.1029/2004GC000784
11. R. D. Müller, W. R. Roest, J. Y. Royer, Asymmetric sea-floor spreading caused by
ridge-plume interactions. Nature 396, 455–459 (1998). doi:10.1038/24850
12. R. Granot, J. Dyment, Y. Gallet, Geomagnetic field variability during the Cretaceous
Normal Superchron. Nat. Geosci. 5, 220–223 (2012). doi:10.1038/ngeo1404
13. W. H. F. Smith, D. T. Sandwell, Global seafloor topography from satellite altimetry
and ship depth soundings. Science 277, 1956–1962 (1997).
doi:10.1126/science.277.5334.1956
14. J. A. Goff, W. H. F. Smith, K. A. Marks, The contributions of abyssal hill
morphology and noise to altimetric gravity fabric. Oceanography 17, 24–37
(2004). doi:10.5670/oceanog.2004.64
15. N. K. Pavlis, S. A. Holmes, S. C. Kenyon, J. K. Factor, The development and
evaluation of the Earth Gravitational Model 2008 (EGM2008). J. Geophys. Res.
117, B04406 (2012). 10.1029/2011JB008916 doi:10.1029/2011JB008916
1
16. M. Seton, R. D. Müller, S. Zahirovic, C. Gaina, T. Torsvik, G. Shephard, A. Talsma,
M. Gurnis, M. Turner, S. Maus, M. Chandler, Global continental and ocean basin
reconstructions since 200 Ma. Earth Sci. Rev. 113, 212–270 (2012).
doi:10.1016/j.earscirev.2012.03.002
17. W. Mohriak, M. Nóbrega, M. Odegard, B. Gomes, W. Dickson, Geological and
geophysical interpretation of the Rio Grande Rise, south-eastern Brazilian margin:
Extensional tectonics and rifting of continental and oceanic crusts. Petrol. Geosci.
16, 231–245 (2010). doi:10.1144/1354-079309-910
18. I. Scotchman, G. Gilchrist, N. Kusznir, A. Roberts, R. Fletcher, in The Breakup of the
South Atlantic Ocean: Formation of Failed Spreading Axes and Blocks of
Thinned Continental Crust in the Santos Basin, Brazil and Its Consequences For
Petroleum System Development (Petroleum Geology Conference Series,
Geological Society of London, London, 2010), pp. 855–866.
19. W. U. Mohriak, P. Szatmari, S. Anjos, Salt: Geology and tectonics of selected
Brazilian basins in their global context. Geol. Soc. London Spec. Publ. 363, 131–
158 (2012). doi:10.1144/SP363.7
20. M. T. Chandler, P. Wessel, Improving the quality of marine geophysical track line
data: Along‐track analysis. J. Geophys. Res. Solid Earth 113, B02102 (2008).
doi:10.1029/2007JB005051
21. O. B. Andersen, P. Knudsen, Global marine gravity field from the ERS‐1 and Geosat
geodetic mission altimetry. J. Geophys. Res. Oceans 103, 8129–8137 (1998).
doi:10.1029/97JC02198
22. W. Haxby, G. Karner, J. LaBrecque, J. Weissel, Digital images of combined oceanic
and continental data sets and their use in tectonic studies. EOS 64, 995–1004
(1983). doi:10.1029/EO064i052p00995
23. C. Hwang, B. Parsons, An optimal procedure for deriving marine gravity from multi-
satellite altimetry. Geophys. J. Int. 125, 705–718 (1996). doi:10.1111/j.1365-
246X.1996.tb06018.x
24. D. T. Sandwell, W. H. F. Smith, Global marine gravity from retracked Geosat and
ERS-1 altimetry: Ridge segmentation versus spreading rate. J. Geophys. Res.
Solid Earth 114, B01411 (2009). doi:10.1029/2008JB006008
25. D. Sandwell, E. Garcia, K. Soofi, P. Wessel, M. Chandler, W. H. F. Smith, Toward 1-
mGal accuracy in global marine gravity from CryoSat-2, Envisat, and Jason-1.
Leading Edge 32, 892–899 (2013). doi:10.1190/tle32080892.1
26. E. S. Garcia, D. T. Sandwell, W. H. F. Smith, Retracking CryoSat-2, Envisat and
Jason-1 radar altimetry waveforms for improved gravity field recovery. Geophys.
J. Int. 196, 1402–1422 (2014).
27. D. J. Wingham, C. R. Francis, S. Baker, C. Bouzinac, D. Brockley, R. Cullen, P. de
Chateau-Thierry, S. W. Laxon, U. Mallow, C. Mavrocordatos, L. Phalippou, G.
Ratier, L. Rey, F. Rostan, P. Viau, D. W. Wallis, CryoSat: A mission to determine
the fluctuations in Earth’s land and marine ice fields. Adv. Space Res. 37, 841–
871 (2006). doi:10.1016/j.asr.2005.07.027
2
28. H. D. Griffiths, Synthetic aperture processing for full-deramp radar altimeters.
Electron. Lett. 24, 371–373 (1988). doi:10.1049/el:19880251
29. W. T. K. Johnson, Magellan imaging radar mission to Venus. Proc. IEEE 79, 777–
790 (1991). doi:10.1109/5.90157
30. R. K. Raney, The delay Doppler radar altimeter. IEEE Trans. Geosci. Rem. Sens. 36,
1578–1588 (1998). doi:10.1109/36.718861
31. L. Phalippou, V. Enjolras, Re-tracking of SAR altimeter ocean power waveforms and
related accuracies of sea surface height, significant wave height, and wind speed.
Proc. IEEE IGARSS’07, 3533–3536 (2007).
32. J. Weickert, Coherence-enhancing diffusion of colour images. Image Vis. Comput.
17, 201–212 (1999). doi:10.1016/S0262-8856(98)00102-4
33. G. C. Fehmers, C. F. Höcker, Fast structural interpretation with structure-oriented
filtering. Geophysics 68, 1286–1293 (2003). doi:10.1190/1.1598121
3
