Neil Hodgson, Karyna Rodriguez and Lauren Found bring observations and insights from the Pelotas Basin to suggest that sequence stratigraphic models of deepwater require an overhaul.
The use of a eustatic sequence stratigraphic model as a predictive tool for deepwater clastic sedimentation has become problematic. Not only is the model focused on the shelf rather than basin floor (see Figure 1 – the area of interest is the minor section on the far outboard part of this diagram), we now know that multiple processes such as sediment supply, climate, shelf stability and basin tectonics (to name but a few potentially disruptive influences), that have no eustatic drivers (or indirect cryptic ones at best) can generate exceptions to the model. To these we add the contourite current – which in deep water can construct thick shale sequences where none would be predicted, move and rework clastic deposits where they are predicted, and change the drivers on slope channel and basin floor accumulation in ways undescribed so far in current sequence stratigraphic models.
As exploration moved into frontier deepwater areas of the passive margins of the Atlantic, the predictive capability of eustatic sequence stratigraphy could have been of paramount importance. However, exceptions to the model in deepwater are being increasingly reported, because of the availability of modern regional 3D data over the deepwater basin floor, such as the recently acquired regional 3D seismic data from the Pelotas Basin (inset map Figure 2) which now illuminate details of the deepwater system for the first time at this scale. This regional 3D style of dataset is becoming increasingly available, bringing new observations and insights to deepwater systems and we suggest that the model that has prevailed since the 1960s requires an overhaul for the part of the system that lies beyond the shelf edge: the deepwater basin (Figure 1).
In deepwater sediments, variations in sea-level do not create sequence bounding unconformities. Instead, they manifest as sequences that are responding to variations in sediment delivery influenced by facies architecture. Far beyond the shelf edge, the classical hierarchy of sequence-bounding surfaces – famously depicted in two-dimensional dip-oriented models (Figure 1) should include another protagonist coming from out of plane – the contourite or current flowing parallel to the coast. Variation in the character of stacking pattern, levee development, channel–lobe evolution, with sediment supply, grain size variation etc., begin to define sedimentary sequences. Whilst both deepwater sequence stratigraphy and contourites are much studied, the implication of contourite activity on a turbidite driven system in a sequence stratigraphic process is less well documented.
Over the last three acquisition seasons in the Pelotas basin of southeastern Brazil, Searcher with its partner Shearwater have acquired 17,000 km2 of multi-client 3D seismic data, chasing stratigraphic and structurally trapped Albian sands above Aptian source, analogous to the Venus discovery in Namibia’s Orange basin, in addition to Late Cretaceous channel systems charged from the Aptian or indeed the Cenomanian/Turonian source.
By tying into lines from a large regional legacy seismic grid, constrained by all available wells in the Pelotas Basin and extending into Uruguay to the south, we are able to build a regional stratigraphic 3D grid of surfaces with constrained ages (Figure 2) that we subsequently correlated into a sequence stratigraphic framework – the interpretation straw-man in Figure 3. By interpreting the main sequence boundaries (black-dashed lines) on regionally consistent, conformable surfaces, we are assuming that these units were deposited in periods of relatively low input to the basin, that they are mainly fine shale pelagic drape and represent a time when facies belts on the shelf are pushed back towards the coast. These condensed surfaces would include a maximum flooding surface, and whilst lots could be wrong with that assumption, we call this the straw man – a first pass model to test observations against.
Our process was to then examine the sequences between the candidate flooding surface and consider two perspectives: if the sequence was as interpreted, could the facies character be characteristic of a depositional process happening in that part of the system tract, and/or what other processes could be influencing the facies character observed?
Classical sequence stratigraphy is unashamedly shelf-edge-centric, and in deeper water, turbidite centric. Relative sea-level changes produce recognisable surfaces through changing depositional accommodation space and the point of depositional onlap. The model is gloriously simple and in its simplified expression has maximum flooding surfaces develop when relative sea levels are high cycling gracefully into basin ward shifts in facies below the shelf edge when relative sea levels fall. Early attempts to tie sea-level into global phenomena such as climate controlled eustacy, have been challenged over time, and many exceptions have been found where isostacy (dynamic topography) trumps eustacy on the shelf, or sediment supply changing dominates the development of sequence architecture. However, the model can be a useful first pass to set up a straw man for making observations against, as the concept is simple to understand and, on the shelf and upper slope the geometries of systems tracts (LST, TST, HST) in the model can often be confidently observed.
Despite these observations, at the toe of the slope and basin floor the correlatable conformity to unconformities updip and the geometry of sequence tracts can be cryptic – i.e. hard to see and uniquely defined from within the stacked conformable sediments. Of course, 200 km offshore and in 2000-3000 m of water, a sea level drop or rise of less than a 100 m has no practical effect on accommodation space. However, phenomena affecting relative sea level should impact the shelf, thereby indirectly affecting sedimentation on the basin floor fundamentally. A relative rise in sea level and transgression of the shelf may reduce coarse clastic input to a basin as accommodation space is found on the shelf, yet this can also be a time of shelf-edge ravinement and canyon formation, eroding sections of the shelf edge which is taken through turbidites down to the basin floor. A relative sea level drop below the shelf may increase the amount of sediment being dragged off the self and focus this into canyons and channel systems, again transporting sediments down slope to the basin floor. In all this, inverting the shelf tectonically will simulate a relative sea-level fall and increasing the sediment supply through climate changes, weathering, sediment production, fluvial transfer, stream-piracy, shelf storage, canyon connectivity, slope transfer can erroneously simulate events in the eustatic model.
Another factor to consider in deepwater sedimentation sequence stratigraphy is the role of contourites in deposition. Actually, this is true of shelf, slope and basin floor sediments – if there are orthogonal currents then they will be affecting sedimentation in all these arenas. Yet it is often in the lower slope that the silent guerrilla warfare of contourite activity is observed most clearly.
Sequence Stratigraphy has a deep history in outcrop studies (i.e. Book Cliffs in the USA), yet clastic contourites are rarely reported at outcrop. This is not necessarily because they are not there but because you need scale of exposure to see them, and you also need to be looking at deepwater sediments in basins that had contourite currents active (often passive margin basins). It is thus unsurprising that contourites have been extensively studied on deepwater modern seismic in passive margins, rather than outctrop. One such margin that displays Cretaceous and Tertiary contourites is the South Atlantic Brazil-Uruguay Pelotas Basin.
Contourites can repurpose the sediments entrained in turbidite flows to build contourite drifts down flow, and create colossal sediment-wave fields (Stow et al., 2009), as well as remobilising dumps of coarse clastics into mobile sand dune fields (Mutti et al., 2014). On the slope, as seen in the recent work by Bryan Cronin on the Jubilee and TEN fields of Ghana, channels confined by levees can be affected by contourites that develop asymmetric-sized levees, facilitating channel avulsion to occur in a particular direction, i.e. into the flow of the contourite. None of these architectures yet have a system tract to call their own in a turbidite mindset dominated by sequence stratigraphy models. Or to look at this another way – to date we have very little understanding of the relation between eustacy and even relative seal level changes and contourite activity, if any, as they are often complex current systems even within a single paleo-seabed equivalent strata, so we do not know what contourites might add to the characterisation of system tracts, though they have to be at least considered as modifiers.
The first systems tract we observe in the basin lies directly above oceanic crust in Figures 3 and 4. It is Aptian in age and represents the first marine transgression into the spreading basin between Africa and South America. This was a tectonically driven first transgression, reflecting subsidence of newly formed oceanic crust during continental drift. Existing sequence-stratigraphic models do not explicitly include this first transgression as the starting point.
One limitation of the conventional model is that it is ‘tectonic setting-independent’. Therefore, it was never designed to predict the development of a world-class source rock directly over oceanic crust. Furthermore, the concept of identifying source rocks directly from seismic data (Loseth et. al., 2009, Davison et al., 2018) is relatively recent, and now that the significance of this first source-rock depositing event is established, there is value in incorporating it into an updated sequence-stratigraphic model for passive margins which could be used more widely. This would be a combination of eustacy and mega-regional tectonics related to continental drift initiation.
In the Pelotas Basin, the first Aptian source-rock event was deposited in the depression formed at the transition between SDRs (Seaward Dipping Reflectors) and oceanic crust, and directly over oceanic crust. As SDRs were deposited subaerially, we envisage an initially relatively shallow-water basin. As transgression continued, the source-rock package was deposited over an increasingly extensive area (Figure 5).
Moving into the Albian, continued basin deepening could be interpreted within the classical model as continued transgression. Yet, we observe series of north-south-oriented clastic channels and fans developed above the Aptian source rock. Based on their seismic character and their position within the conventional model, these channel and fan systems have been interpreted as part of a lowstand systems tract, labelled Lowstand 1 and 2 in Figure 6.
A series of large clastic fans, oriented north-south and deposited at base of slope (Figure 6A), prograde, with lateral accretion, onlapping onto the Aptian source rock to the southeast, and sitting in counter regional dip configuration. This is very similar to the configuration for Venus in Namibia’s Orange Basin, especially the counter-regional depositional onlap. Lateral accretion of fans may be occurring as slope-channels feeding them migrate laterally too, or it could just reflect successive fans moving to fill the topology on the basin floor. Figure 6B is a surface at the top of the Albian sequence, showing the lateral movements of the basin floor fans at that time. Their trajectory is almost north to south, whilst slope updip is NW-SE implying that as turbidites hit the basin floor their trajectory is modified into a north-south orientation. This could be because of the geometry of a trough at the base of slope (Figure 6) deflected flow to the south, oblique to slope-dip, and nearly orthogonally in contourite orientation constrained by the rising basement moving outboard, that creates the counter regional dip. A gravity-driven turbidite that ends up flowing parallel to contours is deliciously close to having converted itself into a contourite, but of course it isn’t that at all, and lacking further dip to follow will lose momentum and settle out on that basin floor nadir, as a fan. This horizon (Figure 6B) is from the quieter later part of the upper low stand, and contourite sand waves to the west may have developed due to a lack of turbidite input, and a dominance of contourite influence in the late low stand, transgressive system tract and high stand. However, the contourite currents are unlikely to have appeared only when the turbidite input starts to wane – it is just that this is when we start to see their influence. At other times the contourites will be modifying channel development and stripping turbidite plumes of their fines, generating coarse clastic turbidite flows that have higher net to gross.
In the Albian sequence, Figure 6, we observe a lower section which has highly reflective soft horizons pinching out against counter regional dip to the Southeast which is associated with the lobe dominated fan (Figure 6A), below a more well-bedded, conformable and lower amplitude section, which is associated with the more channel-dominated fans and the contourite bedforms to the west. If the basal sequence was a low stand, then coarse clastics, eroded off the shelf and focused down dip during the initial relative sea level fall, could have ended up on the basin floor. Then subsequent coarse clastics in the low stand phase could have been caught in the proximal prism on the slope, with finer grained materials reaching the basin floor.
At the base of this Albian sequence we see no clear evidence of contourites. Whilst it’s possible that there were contourite currents in that basin, its isolation and restricted nature suggests not, and we have only the lateral migration of the fans as evidence which frankly could be due to several other factors. In fact it would be a surprise perhaps to have strong contourites in the basin at this time as the Atlantic gateway between north and south Atlantic does not open until latest Albian-Cenomanian times (demonstrated by Duarte et al 2025) and it is then that we might expect the South Atlantic to begin to experience strong, persistent contourite currents.
At the top of the Albian, a well-bedded, conformable maximum-flooding/highstand condensed sequence boundary passes upwards into a lower Cenomanian-Turonian lowstand (Figure 7A). This comprises a mixture of well-developed contourite bedforms to the west and channels and fans to the east.
Above these, to the west, we observe a package characterised by contourite drifts, formed by relatively low-velocity contourite currents transporting fine-grained sediment, together with small turbidite channels migrating westwards. These pass laterally eastwards into parallel-bedded, low-amplitude events interpreted as the Cenomanian-Turonian source-rock package (Rodriguez et al., 2026, GeoExpro). The contourite drift and source rock form part of a transgressive systems tract at the top of the Turonian (Figure 7B).
The Coniacian is an impressive expanded sequence in the basin, containing a thick succession of amalgamated channel complexes and basin-floor fans (Figure 8), apparently reworked by strong, high-velocity contourite currents. Most of the sequence is interpreted as lowstand. Several clear channels can be identified on legacy 2D strike lines on the slope, supporting the transport of coarse-grained sediment into the basin floor.
The Coniacian also appears to correspond to a broader global event, with several reports of thick deepwater successions of clastic channels and fans. The Mopane and recently announced Merlin discoveries in Namibia’s Orange Basin, including an intra-MTC fan, are reported to be Coniacian in age.
At Coniacian level (Figure 8) we see the clear signs of strong contourite wave-fields in the basin. Relict channels coming from the NW can still be interpreted. Indeed, we see sediment waves and drifts adjacent to channel levees. Contourite currents cannot be assumed to be constant with time and may be hugely affected by climate (so may relate chronologically to eustacy) and would change drastically during the opening or modification of sea gateways. Periods of dominant contourites influence on sedimentation might be required for the development of contourite sediment waves, but at other times, when turbidite flows dominate, they would be quietly influencing channel avulsion, levee asymmetry and lobe switching.
As we move towards integrating contourite observations with deepwater sequence stratigraphy, we recall that classic models once assumed lowstands generated all basin-floor fans. Indeed, variability in the sedimentological record that didn’t fit the model, were considered to reflect the unique tectonic and depositional settings of each sedimentary basin, and needed to be rationalised on a case-by-case basis. The occurrence of contourite modified bedforms, and just the ubiquitousness of contourites in the post-Albian section suggest that all such sequences should be considered with an eye to contourite activity.
Our observations in the Pelotas Delta are consistent with the introduction of coarse clastics during early stages of the lowstand, and a sediment supply that wanes as sediments are caught in lowstand wedges on the slope. Yet, other clastics also appear through the late lowstand, transitional and highstand system tracts as well and all of these sands will be modified by contourite currents by fine stripping pre-deposition, modification to sediment routing and deposition, creation of sediment waves, contourite drifts, or reworking of deposited coarse clastics.
The implication is not that contourites replace sequence stratigraphy, nor that every deepwater architectural element should be attributed to contourite processes. Rather, the deepwater model needs another degree of freedom. In addition to relative sea level, sediment supply, tectonics and accommodation, contour-parallel currents should be considered as an active control on sediment routing and preservation.
As the original sequence stratigraphic model was developed to be tectonic-setting independent, the first transgressive source-rock event above oceanic crust provides another missing element. Together, this event and the subsequent contourite influence suggest that the classic shelf-edge sequence-stratigraphic model does not fully capture the architecture of the deepwater basin.
The expansion of the model into deep water (Figure 9), with systems tracts modified to account for contourite activity and the first transgressive source rock above oceanic crust, would therefore be a useful step forward in reapproaching the characterisation of deepwater sediments within a sequence stratigraphic model. The Pelotas 3D dataset provides an opportunity to test that proposition at a scale that was simply unavailable when the original models were developed.
Acknowledgements
Searcher wishes to thank its partner Shearwater for its contribution to the Pelotas 3D projects, in particular Sindre Bull-Jansen for the insightful Paleoscan images in Figures 7 and 8 above herein.
