GSA DATA REPOSITORY 2016013

RIFT TO DRIFT TRANSITION IN THE SOUTH ATLANTIC SALT BASINS: A NEW

FLAVOR OF OCEANIC CRUST

Ian O. Norton1,2*, Daniel T. Carruthers1, and Michael R. Hudec1

1
Bureau of Economic Geology

2
Institute for Geophysics (UTIG), Jackson School of Geosciences, University of Texas at Austin

*
E-mail: Norton@ig.utexas.edu

SUPPLEMENTARY INFORMATION

ABSOLUTE AGE OF SALT

Salt in the South Atlantic has not been definitively dated, although its age is generally regarded
as Aptian. More precise age control is lacking, with different methods yielding conflicting
results. To add to the uncertainty, the absolute age of the Aptian stage is poorly constrained. Salt
must be younger than the underlying presalt rift and sag section, but because the basins in which
this section was deposited were isolated from the global ecosystem, stratigraphic ages are
uncertain. Instead, isotopic ages of presalt basalts have been used to constrain ages.
Measurements have been reported from two locations off southern Brazil (W1 and W2, Fig.
DR1). Ages of 117 to 118 Ma were reported from a basalt found in well 1-RJS-625 in the
northern Santos Basin (Moreira et al., 2007; W1 in Fig. DR1). This basalt was interpreted as
being located at the base of the sag section. A younger age of 113 Ma was reported for basalt in
well 1-SCS-2 (Dias et al., 1994; W2 in Fig. DR1). The well is at the west end of the
Florionopolis fracture zone and the dated basalt is overlain by Albian carbonates and anhydrites.
This well is 200 km from the Santos salt basin (Fig. DR1), so why the 113 Ma age has been
extensively used as defining the age of base salt in Brazil (e.g. Moreira et al., 2007) is not
obvious. The South Atlantic was already too wide even at 114 Ma for salt to have been deposited
then (paper Fig. 1).

1 
 
Above the salt, well-developed carbonate platforms had developed on both margins by Albian
time (Chaboureau et al., 2013), although the oldest carbonates directly on top of the salt have not
been dated (Davison, 2007; Davison et al., 2012). Tighter stratigraphic age constraints on
postsalt sediments are given by the occurrence of Early Aptian planktonic foraminifera in an
unidentified Angolan well (Davison, 2007) and by Middle Aptian foraminifera in DSDP hole
364 (Fig. 2 in Caron, 1978; age reported as ‘Gargasian’, which is Middle Aptian (Moullade et
al., 2005)). These data suggest that salt can be no younger than Middle Aptian and could be
older, depending on correlations from the Angolan well (Davison, 2007). Comparison of these
stratigraphic age constraints with isotopic age constraints from the presalt section needs a
geologic time scale, but here we run into some problems. Rather than being defined by
paleontological data, the base of the Aptian stage is defined as the beginning of Chron M0r
(Erba, 1996). It is the only geologic stage boundary defined by a method other than from the
fossil record. Recent ICS time scales (Gradstein et al., 2004; ICS 2015) set the age of base
Aptian at 125 Ma. Time scales that focus only on magnetic anomalies, however, find younger
ages for M0r, such as 121 Ma (Malinverno et al., 2012) or 121.54 Ma Gee and Kent, 2007).
Comparing these ages, the age of base Aptian has a 4-million-year uncertainty. Top Aptian is
113 Ma in ICS 2015 (magnetic anomaly time scales do not apply here because this time is in the
Cretaceous Normal Superchron). Dividing the Aptian into three equal intervals for early, middle,
and late makes Early Aptian (the oldest stratigraphic age for salt) either 125 to 121 Ma (ICS
2015) or 121 to 118.3 Ma (magnetic time scales). Both of these age ranges are older than the 117
to 118 Ma age reported for presalt basalt in Brazil well 1-RJS-625 (Moreira et al., 2007), which
raises the question of what data to use for constraining the age of salt. It is beyond the scope of
this paper to resolve the time scale/age dating issues summarized here, but for our purposes, we
suggest that it is within the range of possible salt ages for salt to have been deposited at the time
of Chron M0r (paper Fig. 4); we assign the Gee and Kent (2007) age of 121 Ma for this time
period, which is consistent with the oldest reported stratigraphic age for the postsalt section.

BASIN MODELING

This section outlines our basin analysis which was used to determine the geometry and
dimensions of the South Atlantic salt basins at the time of salt deposition and the times of
formation of the ‘breakthrough volcanoes’ discussed in the main text. The post-rift evolution of a

2 
 
basin is controlled by two main physical processes: isostatic subsidence caused by sediment
loading, and thermal subsidence caused by cooling of the lithosphere after rifting, both of which
are well-understood (Allen and Allen, 1990; Watts, 2001). The approach used was to construct
present-day crustal-scale cross sections and to use standard basin analysis tools to calculate the
depth to basement on these cross sections at the end of rifting. As discussed in the paper, this is
the time of salt deposition, so the next step was to calculate the salt-loaded shapes of the
sections. Finally, using plate reconstructions to constrain changing section lengths, the structural
geometry of the sections were calculated at the times of formation of the breakthrough
volcanoes. We constructed two crustal-scale cross sections (A-A’ and B, located in Figs. DR1
and DR2) across the Campos-Kwanza and Santos-Namibe margins respectively. Section A was
compiled from the seismic lines in Fig. 2 of the paper, plus a depth converted shelfal line
(Guardado et al., 1989). The Kwanza side of the line, A’, was constructed from 2D-seismic Line
GXT/ION 1800 (Fig. 2). It is not a perfect conjugate but was used because seamounts of the
Sumbe Trend (Fig. DR2) confuse basin structure on the Kwanza margin directly conjugate to the
Campos Basin and line 1800 is outside this area. Line 1800 is also not ideal as the rift-sag
pinchout is not apparent on this line, hence the gap in our LOC (Fig. DR2), but because we are
more interested in gross crustal structure we use it anyway. Line B was constructed from the
seismic section in Fig 5 of Kumar et al., (2012), an east-west line that runs from the coast to the
deep basin (Fig. DR1). We used this line because it is approximately parallel to the Africa-South
America relative motion direction (paper Fig. 1).

The first objective of our analysis was to ascertain the crustal thickness of the basin which is
necessary to calculate the crustal thinning factor, Beta. Crustal thickness for Line A-A’ was
derived from gravity modeling (Fig. DR3). Moho reflectors can be identified in the oceanic
crust portions of the seismic data, and these were used as the main constraints for locating Moho
under continental crust, the depth being adjusted to match observed gravity. A simplified
passive margin sedimentary sequence was used, consisting of Tertiary (mostly clastic) and
Cretaceous (mostly carbonate) sediments and only the thick salt located seaward of the LOC.
These simplifications allow for resolution of the long-wavelength gravity signal, which is the
signal resulting from Moho depth, the main output we need from this analysis.

3 
 
Fig. DR4 shows the flexural back-stripping (Watts, 2001) and restoration of thermal subsidence
using an original crustal thickness of 38 km. for Line A. Calculations were run for several
different values of the lithosphere elastic thickness, Te (Watts, 2001). The output for Te=10 km
produced a cross section (Fig. DR4b) that restored the proximal edge of salt to sea level, which is
appropriate. Fig. DR4b details the geometry and dimensions of the basin with all the post-rift
sediments and Aptian evaporites removed and with thermal subsidence restored. When
combined with Line A’ from the Angolan margin (Fig. DR5), we can derive the approximate
basin geometry at the time of salt deposition. In this study we suggest that salt was deposited at
the transition to sea floor spreading, and that the distal edge of the original salt basin was at the
distal sag pinch-out (LOC). We thus restored lines and joined them at the LOCs to reconstruct
the geometry of the Aptian salt basins (Fig. DR6a, b). The top of the water-loaded basin, that
existed prior to salt deposition, is shown by the green line on Fig. DR4. Pink shading shows the
shape of the basin after salt deposition, assuming that salt filled the basin to sea level. For
calculation purposes we assume that salt deposition was instantaneous (see paper). We also
assume that isostatic loading is instantaneous in our model. Studies of mantle viscosity using
Holocene glacial rebound show that the lithosphere responds to isostatic loads in thousands of
years (McConnell , 1968; Eronen et al., 2001), two orders of magnitude faster than even the
fastest proposed duration of salt deposition (600 kyr, Dias, 2005).

The sections shown in Fig. DR6 were subsequently used to estimate the original thickness and
cross-sectional area of the salt basins. In the Campos Basin, salt area is 254 km2, remarkably
close to the 270 km2 (adjusted for line strike vs. spreading direction) measured on the seismic
line (paper Fig.2 and Carruthers, in prep.). The Santos line also agrees well, with 1433 km2 on
line B (Fig. DR6b) and 1488 km2 measured on the line as published (Kumar et al., 2012). These
close comparisons indicate our results are realistic.

Using the plate reconstructions as guides, we estimate the times at which the salt edges separate.
These times are 118 Ma for the Campos-Kwanza line and 114 Ma (paper Fig. 1) for the Santos
Basin. In Fig. DR7 we show calculations of the shapes of the salt basins at these times, with
basins enlarging by the amount of motion between Africa and South America given by the plate
reconstructions. These calculations assume that oceanic crust subsides at the rate given by
standard subsidence models (Stein and Stein, 1992), with water depth adjusted to maintain the

4 
 
salt areas shown in Fig. DR6. The top salt surface is assumed to be a straight line from zero
motion at the landward salt pinchout to maximum water depth at the ridge axis. This is certainly
an over-simplification, but calculation of the true shape of the salt surface is, for now, an
intractable problem. One observation that is helpful in constraining what the slope can be is
from the front of salt glaciers in the present-day Red Sea, where a dip of 1.27˚ was measured
from Fig. 6 in Augustin et al., (2014). A line showing this dip is included in the Fig. DR7
sections; as can be seen, this is steeper than our assumed salt surface, giving some reassurance
that the slopes shown are possible. The one part of the cross sections that is almost as steep is
the African side of the Santos line, Fig. DR7b. This may be why salt is absent in the Namibe
Basin on that margin: it flowed onto the Santos side of the ocean before the basins separated.
Final separation was also very asymmetric in the Santos-Namibe region, with the final ridge axis
breakthrough close to the Africa LOC.

In our model, the sections shown in Fig. DR7 are for the time of separation of the salt basins and
the start of normal oceanic spreading. We suggest that this happened by the ridge axis breaking
through the overlying salt, forming volcanic constructs which we term ‘breakthrough volcanoes’.
These are preserved today as the landward-facing outer basement ramps seen at the distal salt
limit. An example is shown in paper Fig. 2b. The ramp here is about 2 km high, close to the salt
thickness in our model in Fig. DR7a. From the time of initiation of salt deposition, coincident in
our model with onset of sea floor spreading, until normal oceanic spreading could commence
after the breakthrough volcano formed, oceanic crust was being created underneath a thick layer
of salt. This means that this oceanic crust formed without the usual extrusive component, layer
2A, so we term this crust ‘intrusive oceanic crust’.

5 
 
SUPPLEMENTARY MATERIAL REFERENCES

Allen, P.A. and J.R. Allen, 1990, Basin Analysis, Blackwell Science, pp 449.

Augustin, N., Devey, C.W., van der Zwan, F.M., Feldens, P., Tominaga, M., Bantan, R.A., and
Kwasnitschka, T., 2014, The rifting to spreading transition in the Red Sea, Earth Planet. Sci.
Lett. v. 395, p. 217–230.

Caron, M., 1978, Cretaceous planktonic foraminifers from DSDP Leg 40, southeastern Atlantic
Ocean, in Initial Reports of the Deep Sea Drilling Project, v. 40, (eds Bolli, H. M. et al.),
U.S. Government Printing Office, p. 651–678

Chaboureau, A.C., Guillocheau, F., Robin, C., Rohais, S., Moulin, M., and Aslanian, D., 2013,
Paleogeographic evolution of the central segment of the South Atlantic during Early
Cretaceous times: Paleotopographic and geodynamic implications, Tectonophys. v. 604, p.
191–223.

Davison, I., 2007, Geology and tectonics of the South Atlantic Brazilian salt basins, in
Deformation of the Continental Crust: the Legacy of Mike Coward (eds Reis, A.C., Butler,
R.W. H., and Graham, R.H.), Geol. Soc., London, Spec. Pubs. 272, p. 345–359

Davison, I., Anderson, L., and Nuttall, P., 2012, Salt deposition, loading and gravity drainage in
the Campos and Santos salt basins, in Salt Tectonics, Sediments and Prospectivity (eds
Alsop, G. I. et al.), Geol. Soc., London, Spec. Pubs. 363, p. 159–173.

Dias, J.L., Sad, A.R.E., Fontana, R.L., and Feijó, F.J., 1994, Bacia de Pelotas, B. Geoci.
Petrobras v. 8, p. 235–245.

Dias, J.L., 2005, Tectônica, estratigrafia e sedimentação no Andar Aptiano da margem leste
brasileira, B. Geoci. Petrobras, Rio de Janeiro, v. 13, n. 1, p. 7-25.

Erba, E, 1996, The Aptian stage, Bull. Inst. Royal Sci. Nat. Belg. 66 (suppl.), p. 31– 43.

Eronen, M., Glückert, G., Hatakka, L., van de Plassche, O., van der Plicht, J., and Rantala, P.,
2001, Rates of Holocene isostatic uplift and relative sea-level lowering of the Baltic in SW
Finland based on studies of isolation contacts, Boreas, v. 30, p. 17–30.

6 
 
Gee, J.S., and Kent, D.V., 2007, Source of Oceanic Magnetic Anomalies and the Geomagnetic
Polarity Timescale, Treatise on Geophysics v. 5, p. 455–507.

Gomes, P.O., Kilsdonk, W., Minken, J., Grow, T., and Roberto Barragan, 2009, The Outer High
of the Santos Basin, Southern São Paulo Plateau, Brazil: Pre-Salt Exploration Outbreak,
Paleogeographic Setting, and Evolution of the Syn-Rift Structures, Search and Discovery
Article #10193.

Gradstein, F.M., Ogg, J.G., Smith, A.G., Bleeker, W., and Lourens, L.J., 2004, A new Geologic
Time Scale, with special reference to Precambrian and Neogene, Episodes, v. 27, p. 83-100.

Guardado, L.R., Gambo, L.A.P. and Lucchesi, C.F., 1989, Petroleum Geology of the Campos
Basin, Brazil, a Model for a Producing Atlantic Type Basin, AAPG Mem. A132, p. 3-36.

ICS 2015, http://www.stratigraphy.org/index.php/ics-chart-timescale, accessed March 30, 2015.

Kumar, N., Danforth, A., Nuttal, P., Helwig, J., Bird, D.E., and Venkatraman, S., 2012, From
oceanic crust to exhumed mantle: a 40 year (1970-2010) perspective on the nature of crust
under the Santos Basin, SE Brazil, in Conjugate Divergent Margins (eds Mohriak, W. U. et
al.), Geol. Soc. London, Spec. Publ. 369, p. 147–165.

Malinverno, A., Hildebrandt, J., Tominaga M., and Channell, J.E.T., 2012, M-sequence
geomagnetic polarity time scale (MHTC12) that steadies global spreading rates and
incorporates astrochronology constraints, J. Geophys. Res., 117, B06104,
doi:10.1029/2012JB009260.

McConnell, R.K. Jr., 1968, Viscosity of the mantle from relaxation time spectra of isostatic
adjustment, Journal of Geophysical Research, v. 73, p. 7089–7105,
doi:10.1029/JB073i022p07089.

Moreira, J.L.P., Madeira, C.V., Gil, J.A., Machado, M.A.P., 2007, Bacia de Santos, B. Geoci.
Petrobras v. 15, p. 531–549.

Moullade, M., Tronchetti G., Bellier J.-P., 2005, The Gargasian (Middle Aptian) strata from
Cassis-La Bédoule (Lower Aptian historical stratotype, SE France): planktonic and benthic
foraminiferal assemblages and biostratigraphy.- Carnets de Géologie / Notebooks on
Geology, Brest, Article 2005/02 (CG2005_A02)

7 
 
Sandwell, D.T., and Smith, W.H.F., 2009, Global marine gravity from retracked Geosat and
ERS-1 altimetry: Ridge Segmentation versus spreading rate, J. Geophys. Res., 114, B01411,
doi:10.1029/2008JB006008.

Stein, C., and Stein, S., 1992, A model for the global variation in oceanic depth and heat flow
with lithospheric age, Nature, v. 359, p. 123-128.

Watts, A.B., 2001, Isostacy and Flexure of the Lithosphere, Cambridge Univ. Press, pp 458.

8 
 
 3D

Fig. 2c

W1

W2 Florionopolis fracture zone 

Fig DR1. Data locations, Brazil. Background free air gravity v23.1 (Sandwell and Smith, 2009). Area filled with x’s is salt.  
SOH (diagonal patter) is Santos Outer High (Gomes et al., 2009). Lines A and B are modeled lines. Lines in paper Figs. 2 
and 3 are shown. W1 and W2 are wells mentioned in the supplement.
Fig DR2. Data locations, Africa margin. Background free air gravity v23.1 (Sandwell and Smith, 2009). Area filled with x’s is 
salt. Line A’ is conjugate to line A, Fig. DR2. Line 2400 is line published in Unternehr et al., (2010), mentioned in the main 
text. ST is Sumbe trend seamount chain.
 LOC
|

Line A

Moho

LOC Coast
| |

Line A’

Suppl. Fig. DR3.  Crustal‐scale cross sections across the Brazilian margin (line A) and Angola margin (line A’). Gravity 
modeling was used to extrapolate crustal geometries, especially the Moho, from observations along the oceanic portions of 
the lines.  Gravity profiles were extracted from version 21 of the Sandwell and Smith (2009) gravity data.
 a)
LOC

Edge salt 

b)
LOC

Suppl. Fig. DR4.  Backstripping of Campos Basin Line A. a) is present‐day geometry, derived from the crustal section in Fig. 
DR1a.  b) is the section after flexural backstripping (Watts, 2001) and restoration of 120 m.y. of thermal subsidence (Allen 
and Allen, 1990).
 a)
LOC

b)
LOC

Suppl. Fig. DR5.  Backstripping of Angola Line A’. a) is present‐day geometry, derived from the crustal section in Fig. DR1b.  b) 
is the section after flexural backstripping (Watts, 2001) and restoration of 120 m.y. of thermal subsidence (Allen and Allen, 
1990).
 LOC
A
Campos Kwanza

LOC
B Santos Namibe

Suppl. Fig. DR6. a) Calculation of shape of salt‐filled basin using Line A (Campos) and Line A’ (Kwanza) combined as a single 
profile by joining them at the respective LOCs. Component lines are projected onto the spreading direction following small 
circles in paper Fig. 4.  b) is a similar calculation for the line in the Santos Basin. Green line represents the water‐filled 
original basin shape.  Salt areas are shown, and compare favorably with present‐day vales of 270 km2 in the Campos Basin 
(Dan Carruthers, in prep.) and 1488 km2 measured from Fig. 5 in Kumar et al. (2012).
 Dip of 1.27˚

LOC Ridge LOC

Campos Kwanza
A

Pink = salt
Blue = water Dip of 1.27˚
LOC LOC
B Santos Namibe

Suppl. Fig. DR7. Restoration of cross sections at the times of separation of the distal limits of salt, a) 118 Ma on the Campos‐
Kwanza line and b) 114 Ma in the Santos Basin.  Red vertical lines are at the location of the ridge axes, which break through
the salt to form ‘breakthrough volcanoes’ at the times represented in this figure.  Black vertical dash lines are the LOCs.  
Spreading is assumed to be symmetric on the Campos‐Kwanza line, fully asymmetric on the Santos‐Namibe line with the 
ridge axis remaining at the Namibe LOC.  Topography of the top salt surface is assumed to be a straight line from the 
landward salt limit to the ridge axis, with water depth adjusted to maintain the original salt area from Fig. S4.  For 
comparison, a dip of 1.27˚ is shown, see text.