Analogue modelling of the interplay between gravity gliding and spreading across complex rift topography in the Santos Basin

The Santos Basin presents a complex and controversial evolution and distribution of salt tectonics domains. The controversies revolve mainly around the kinematically- linked Albian Gap and São Paulo Plateau. The Albian Gap is a ~450 km long and 60 km wide feature characterized by a post-Albian counter-regional rollover overlying depleted Aptian salt and in which the Albian is absent. The São Paulo Plateau is defined by a pre-salt structural high with significant base-salt topography and overlain by ~2.5 km thick salt. Another prominent feature is the Merluza Graben, a rift depocentre that underlies the southern portion of the Albian Gap and displays significant (3-4 km) of base-salt relief. Two competing hypotheses have been proposed to explain the origin and kinematics of these provinces. One invokes post- Albian extension within the Albian Gap and contraction in the Sao Paulo Plateau. The other invokes post-Albian salt expulsion in the Albian Gap and salt inflation in the São Paulo Plateau without significant lateral deformation. A recent study shows these processes contribute equally to the evolution of these domains, also demonstrating the importance of the previously neglected base-salt relief. We apply 3D physical modelling to test these new concepts and understand the interplay between laterally- variable base-salt relief, gliding and spreading on salt tectonics. Our results show a remarkably-similar salt and post-salt evolution and architecture to the Santos Basin as proposed in recent studies. They improve the understanding on the distribution and interaction of salt-related structural styles and gravity-driven processes, being also applicable to other salt-bearing margins.

4 by slip on a large counter-regional, listric extensional fault (Demercian et al., 1993;  Recent studies, however, demonstrate that salt deformation in the Albian Gap and 47 São Paulo Plateau is three-dimensionally more variable and complex than previously evolution. This is in accordance with the magnitude of translation (~30 km) observed 53 in ramp-syncline basins in the adjacent São Paulo Plateau 2019c). 54 Beneath the salt, the underlying basement and basin fill is characterized by a system 55 of horsts and grabens, including the Merluza Graben (MG) (cf. Garcia et al., 2012;56 Magee et al., 2020) and the Santos Outer High (Fig 1b). The Merluza Graben is 57 partially overlain by the Albian Gap (Fig. 1a), but due to its location and perhaps 58 complexity, it has been relatively understudied in comparison with its adjacent 59 structural provinces. The area is characterized by pronounced (up to 3.5-4 km) base- or numerical models. In this paper, we use a scaled regional (i.e., representing an area 73 100 km long by 60 km wide) physical model of the Santos Basin to test hypotheses 74 related to its salt tectonic evolution and the way in which rift-related relief controlled 75 6 the subsequent salt-tectonic evolution of the basin. The experiment was designed to 76 test the interplay between: i) rift-related base-salt architecture, ii) gliding and iii) 77 spreading associated with the more controversial structures in the Santos Basin, i.e., 78 the Merluza Graben, the Albian Gap, and the São Paulo Plateau (Fig. 2). It is the first 79 physical modelling experiment to study the linked salt-tectonic evolution of three key,      topography, and a mixture of silica sand and clay was used locally to sculpt a steeper 162 dip for the sub-salt faults. We modelled two major landward-dipping, sub-salt faults 163 that display a variable orientation along strike (from orthogonal to oblique to the 164 tectonic transport) (Fig. 4a). The baseplate was tilted 4º and the polymer was 165 emplaced above the pre-salt topography. After 48 hours, when the polymer had 166 settled, a 5 mm-thick, pre-kinematic layer of blue sand was manually poured above 167 the entire model and levelled with a scraper (Fig. 4b). Deformation was triggered by 168 tilting of the baseplate a further 2º basinward (driving gravity gliding) and by adding a 169 wedge of 4 mm thick blue sand (driving gravity spreading) (Fig. 4c).

170
Syn-kinematic prograding wedges of white and coloured (red and yellow) dry silica 171 sand were poured onto the experiment and then levelled every 2 hours. Prior to the 172 deposition of each syn-kinematic wedge, the baseboard was tilted back 2º. The 173 regional datum was progressively raised 1 mm before the deposition of each sand 174 wedge. The roof of the main salt structures elevated above the regional datum during

191
We first show the map-view evolution of the model using overhead time-lapse 192 photography (Figs. 5 and 6) with DIC data (Figs 7 and 8). We then describe multiple        foldbelt is tightened as salt flow is partially buttressed over base-salt steps and by the 265 gradually basinward-thinning salt over the sub-salt plateau (Fig. 6b, 7d and 8d). 266 In the next, the structures within the half-graben become largely dormant due to 267 continuous basinward salt evacuation and overburden translation onto the sub-salt 268 plateau (Fig. 6c). Updip deformation in the northern portion of the half-graben is 269 characterized by development of counter-regional normal faults and basinward-  (Fig. 6d). Extension is thus localized along the largest counter-282 regional fault and its overlying rollover, adjacent to the inflated salt over the sub-salt 283 plateau (Fig. 6d).    Albian rollovers and turtle anticlines (Figs 9-11). It also contains less prominent, post-330 Albian, bowl-shaped minibasins with near-diapir upturned strata in its most distal 331 portions (Fig. 11).     We also observe in our models the presence of Albian-equivalent strata encased 419 within the salt, above the distal sub-salt plateau ( Fig. 6a-b). These features form due 420 to basinward advance of the near-surface salt walls over the previously folded pre-421 kinematic post-salt strata (Fig. 6a-b). Although such structures have not yet been  The presence of seismically reflective, disrupted, and folded strata within some salt 430 walls in the São Paulo Plateau (Fig. 12), like the ones observed in our models (Figs.  prograding sediment loading (i.e., spreading) with gravity gliding (Fig. 14).

441
In our experiments, the landward-dipping normal faults produce half-grabens and  Our experiment shows that differential loading can enhance or reduce the effects of 451 salt flux variations driven by gliding over base-salt topography (Fig. 14c), thus 452 influencing the spatial and temporal distribution of salt structures (Fig. 14b-c). For rollovers that are larger than in a pure gliding scenario (Fig. 14a-b). This is observed 457 over the Merluza Fault in the Santos Basin (Fig. 13) and the equivalent structure in 458 our experiment (Figs. 5-11).  (Fig. 5a-b). In the south domain, the wider and oblique pre-salt graben produces 484 23 an equally wide, inflated salt structure. In both domains, the inflated salt has a broadly 485 similar planform to the underlying graben (Fig. 5). Because of its greater width and the 486 obliquity of base-salt relief, salt inflation and overburden uplift are less and over a wider 487 area than in the north domain, so that the salt is not able to breakthrough to form a 488 diapir (Fig. 5). In the north, the diapir can advance earlier and faster beyond its sub-  (Figs. 6-11). This spatial and temporal variability of salt deformation are also 503 observed by variations in flow vectors in overhead DIC images (Fig. 7). Where base- base-salt step is gentler and smaller (Fig. 7a-b). Conversely, vertical salt flow and 511 inflation are greater in the north where the sub-salt graben and its bounding fault are 512 orthogonal to the transport direction ( Fig. 7a-b). These earlier patterns, however, 513 change at later stages as salt breaks through the cover to form passive diapirs that 514 accommodate greater lateral salt flow than areas covered by sediments (Figs. 7b-d).

515
This is seen at intermediate model stages (Figs. 6 and 7b-c), as lateral salt flow is 516 greater in the north, along the wide diapir formed above the orthogonal segment of the 517 Merluza Fault.  (Fig. 9). This suggests that basinward-dipping 525 normal faults tend to form earlier when the overburden is thinner and thus, primarily 526 by gliding, whereas landward-dipping (i.e., counter-regional) faults form later when 527 differential loading (i.e., spreading) becomes more dominant. Whereas basinward-528 dipping normal faults are readily associated with extensional rollover geometries, the 529 landward-dipping faults have also more variable growth strata (i.e., rollover and/or growth strata that are locally upturned against inflated salt diapirs (Figs 9-11). They 546 occur above either basinward-dipping ( Fig. 9a and 10a) and landward-dipping base-547 salt ( Fig. 10b and 11a) and form in areas of previously inflated salt (Figs. 5-6).