Abstract Turtle structures are generally regarded to be the result of passive or extensional collapse of underlying salt bodies. However, subsurface examples in the Columbus basin (Trinidad), onshore south Texas (U.S.A.), and offshore Golden Lane (Mexico), exhibit four-way inversion structures that look at least superficially like turtle structures that formed with little or no salt at the detachment level. Furthermore, many anticlines in the shelf region of the northern Gulf of Mexico are possibly more kinematically akin to extensional turtles than traditional turtle structures. I propose the term ‘extensional turtle’ to denote a class of turtle structures at the opposite end of the spectrum from traditional turtles, where a mock turtle lies between. Unlike mock turtles, extensional turtles are generated purely by extension. Even so, they commonly have some mobile-rock withdrawal, as they are always associated with a very weak detachment. The distinguishing feature of such structures is that they are bound by a kinematically linked set of regional and counter-regional, large-offset, listric normal faults. Local complications may occur where younger regional faults repeatedly cut the counter-regional fault, resulting in the counter-regional fault becoming reset downdip. The counter-regional fault may then remain active, resulting in multiple turtle anticlines. Counterregional rollover followed by regional rollover on opposing arcuate listric faults produces four-way plunging turtle structures. Offshore Louisiana has numerous turtle structures, extensional faults, and salt diapirs. I suggest that many of these turtles bear significant similarities to extensional turtles, and it is likely that a continuum of structural styles exist between true turtles, mock turtles and extensional turtles.

Abstract The conceptual breakthroughs in understanding salt tectonics can be recognized by reviewing the history of salt tectonics, which divides naturally into three parts: the pioneering era, the fluid era, and the brittle era. The pioneering era (1856-1933) featured the search for a general hypothesis of salt diapirism, initially dominated by bizarre, erroneous notions of igneous activity, residual islands, in situ crystallization, osmotic pressures, and expansive crystallization. Gradually data from oil exploration constrained speculation. The effects of buoyancy versus orogeny were debated, contact relations were characterized, salt glaciers were discovered, and the concepts of downbuilding and differential loading were proposed as diapiric mechanisms. The fluid era (1933–1989) was dominated by the view that salt tectonics resulted from Rayleigh-Taylor instabilities in which a dense fluid overburden having negligible yield strength sinks into a less dense fluid salt layer, displacing it upward. Density contrasts, viscosity contrasts, and dominant wavelengths were emphasized, whereas strength and faulting of the overburden were ignored. During this era, palinspastic reconstructions were attempted; salt upwelling below thin overburdens was recognized; internal structures of mined diapirs were discovered; peripheral sinks, turtle structures, and diapir families were comprehended; flow laws for dry salt were formulated; and contractional belts on divergent margins and allochthonous salt sheets were recognized. The 1970s revealed the basic driving force of salt allochthons, intrasalt minibasins, finite strains in diapirs, the possibility of thermal convection in salt, direct measurement of salt glacial flow stimulated by rainfall, and the internal structure of convecting evaporites and salt glaciers. The 1980s revealed salt rollers, subtle traps, flow laws for damp salt, salt canopies, and mushroom diapirs. Modeling explored effects of regional stresses on domal faults, spoke circulation, and combined Rayleigh-Taylor instability and thermal convection. By this time, the awesome implications of increased reservoirs below allochthonous salt sheets had stimulated a renaissance in salt tectonic research. Blossoming about 1989, the brittle era is actually rooted in the 1947 discovery that a diapir stops rising if its roof becomes too thick. Such a notion was heretical in the fluid era. Stimulated by sandbox experiments and computerized reconstructions of Gulf Coast diapirs and surrounding faults, the onset of the brittle era yielded regional detachments and evacuation surfaces (salt welds and fault welds) along vanished salt allochthons, raft tectonics, shallow spreading, and segmentation of salt sheets. The early 1990s revealed rules of section balancing for salt tectonics, salt flats and salt ramps, reactive piercement as a diapiric initiator resulting from tectonic differential loading, cryptic thin-skinned extension, influence of sedimentation rate on the geometry of passive diapirs and extrusions, the importance of critical overburden thickness to the viability of active diapirs, fault-segmented sheets, counter-regional fault systems, subsiding diapirs, extensional turtle structure anticlines, and mock turtle structures.

Abstract The conceptual breakthroughs in understanding salt tectonics can be recognized by reviewing the history of salt tectonics, which divides naturally into three parts: the pioneering era, the fluid era, and the brittle era. The pioneering era (1856-1933) featured the search for a general hypothesis of salt diapirism, initially dominated by bizarre, erroneous notions of igneous activity, residual islands, in situ crystallization, osmotic pressures, and expansive crystallization. Gradually data from oil exploration constrained speculation. The effects of buoyancy versus orogeny were debated, contact relations were characterized, salt glaciers were discovered, and the concepts of downbuilding and differential loading were proposed as diapiric mechanisms. The fluid era (1933–1989) was dominated by the view that salt tectonics resulted from Rayleigh-Taylor instabilities in which a dense fluid overburden having negligible yield strength sinks into a less dense fluid salt layer, displacing it upward. Density contrasts, viscosity contrasts, and dominant wavelengths were emphasized, whereas strength and faulting of the overburden were ignored. During this era, palinspastic reconstructions were attempted; salt upwelling below thin overburdens was recognized; internal structures of mined diapirs were discovered; peripheral sinks, turtle structures, and diapir families were comprehended; flow laws for dry salt were formulated; and contractional belts on divergent margins and allochthonous salt sheets were recognized. The 1970s revealed the basic driving force of salt allochthons, intrasalt minibasins, finite strains in diapirs, the possibility of thermal convection in salt, direct measurement of salt glacial flow stimulated by rainfall, and the internal structure of convecting evaporites and salt glaciers. The 1980s revealed salt rollers, subtle traps, flow laws for damp salt, salt canopies, and mushroom diapirs. Modeling explored effects of regional stresses on domal faults, spoke circulation, and combined Rayleigh-Taylor instability and thermal convection. By this time, the awesome implications of increased reservoirs below allochthonous salt sheets had stimulated a renaissance in salt tectonic research. Blossoming about 1989, the brittle era is actually rooted in the 1947 discovery that a diapir stops rising if its roof becomes too thick. Such a notion was heretical in the fluid era. Stimulated by sandbox experiments and computerized reconstructions of Gulf Coast diapirs and surrounding faults, the onset of the brittle era yielded regional detachments and evacuation surfaces (salt welds and fault welds) along vanished salt allochthons, raft tectonics, shallow spreading, and segmentation of salt sheets. The early 1990s revealed rules of section balancing for salt tectonics, salt flats and salt ramps, reactive piercement as a diapiric initiator resulting from tectonic differential loading, cryptic thin-skinned extension, influence of sedimentation rate on the geometry of passive diapirs and extrusions, the importance of critical overburden thickness to the viability of active diapirs, fault-segmented sheets, counter-regional fault systems, subsiding diapirs, extensional turtle structure anticlines, and mock turtle structures.

Abstract Seismic reflections interpreted to be top Oligocene, top Wilcox (approximately base middle Eocene), top Cretaceous, top Jurassic, and basement were mapped across portions of the Green Canyon, Keathley Canyon, Walker Ridge, Lund, Sigsbee Escarpment, Amery Terrace, and Lund South OCS areas of the central northern Gulf of Mexico (Fig. 1). 3D Pre-stack depth migrated data were used for mapping the areas covered by allochthonous salt. 2D Pre-stack time migrated data were used for mapping the area on the abyssal plain beyond the Sigsbee Escarpment. These data cover approximately 50,000 km 2 (19,500 miles 2). Well control was obtained from data available through the Minerals Management Service. Figure 1. Location map. Black line encloses the area of data coverage. Dashed line marks the transition from 3D prestack depth migrated (PSDM) data to the west and north to 2D pre-stack time migrated (PSTM) data to the east. Numbered segments refer to figures with those numbers. Abbreviations for deep-water OCS areas: AC–Alaminos Canyon; AM–Amery Terrace; AT–Atwater Valley; EB–East Breaks; GC–Green Canyon; GB–Garden Banks; KC–Keathley Canyon; L–Lund; LS–Lund South; MC–Mississippi Canyon; SE–Sigsbee Escarpment; WR–Walker Ridge. Structure maps on the top Oligocene, top Wilcox, top Cretaceous, and basement formed the regional surfaces between which isopach/isochron maps were created to analyze depositional patterns. As might be expected, basement structure displayed the greatest relief and complexity. Outboard from the allochthonous salt of the Sigsbee Escarpment, half-graben structures indicative of rift basin topography were clearly imaged (Fig. 2). Elsewhere on the abyssal plain isolated, sharp-peaked, elevated basement features were observed between more numerous gently sloped highs. These basement structures typically had reflection terminations against their margins or flanks and continuous reflections draping them. Figure 2. 2D Pre-stack time migrated line showing rift basin structure in the basement, Wilcox strata down lapping onto the Cretaceous and thinning to the east, and Oligocene strata down lapping onto the Wilcox and thinning to the north. The vertical scale is in seconds of two-way time (TWT). The horizontal scale is in feet (100,000 feet ~ 18.94 miles or 30.55 kilometers). Abbreviations for horizons: Olig=Oligocene (orange); Wx=Wilcox (blue); K=Cretaceous (green); J=Jurassic (pink); and Bsmt=basement (yellow). The top Cretaceous and top Wilcox surfaces show broad regional similarities and show less structural complexity than the basement. Outboard of the Sigsbee Escarpment, both surfaces are broadly lobate and have relatively gentle inclinations which rise to the east. The main observable differences between the two are: (A) the Cretaceous surface has several isolated high points reflecting underlying basement structures and (B) the Wilcox surface has a more lobate/interdigitate contour character. The top Oligocene surface is less lobate in appearance than either the Cretaceous or Wilcox surface and rises to the southeast (Fig. 3). Figure 3. Time structure map on the top Oligocene. The contour interval is 50 milliseconds. Isochron maps between the four structural surfaces reflect the underlying structure and depositional trends of the interval. Thus the basement to Cretaceous isochron shows thick Jurassic infill, Cretaceous drape in the grabens (Fig. 2), and thin to no cover over highs in the rifted basement topography. The Cretaceous to Wilcox isochron has a broad lobate form that thins gently from west to east. A very subtle down-lapping pattern is visible within the Wilcox interval on Figure 2 . Deviations from this pattern occur primarily where basement structures produce isolated thins. The Wilcox to Oligocene interval shows a regional gradient of north to south thickening and only a slight influence from deeper structure. Down-lapping and thinning to the north strongly suggest a southerly source for the Oligocene interval. Beneath the allochthonous salt of the Sigsbee Escarpment, all surfaces deepen northward and show much greater local variability. Basement is only occasionally visible as it generally lies below the fifteen kilometer limit of the available PSDM data. The deepest area mapped is in Green Canyon where the top Oligocene approaches twelve kilometers depth, the top Wilcox approaches thirteen kilometers, and the top Cretaceous almost fourteen and one half kilometers. These surfaces shallow to less than eight kilometers deep on the abyssal plain. Three coincident lows roughly oriented north-south suggest preferred sediment pathways and possibly areas of thicker original autothonous salt. A change on the structure and isopach maps from smooth broadly spaced contours on the abyssal plain to highly variable tightly spaced contours suggests the location for the original limits of salt deposition in this area. This location often lies close to but not exactly in line with the present day Sigsbee Escarpment (Fig. 1). Of key interest to hydrocarbon explorationists are any factors that would effect Wilcox deposition. We have observed three factors that influence the deposition and thickness of Wilcox age strata in this area: Pre-existing basement highs have caused the Wilcox to be thin or absent around those structures. Although basement topography is mostly smoothed over by the end of the Cretaceous, a few large structures still influenced deposition in the Wilcox on the abyssal plain beyond the Sigsbee Escarpment. Salt nappes and salt pillows have caused thinning of Wilcox strata over those structures. Our interpretation indicates multiple kilometer thick salt nappes extruded beyond the limits of the original salt basin during the Cretaceous (Figs. 4 and 5). Inflated salt pillows associated with the nappes lay along the boundary of the salt basin. Though now deflated, the presence of these salt pillows and other salt pillows updip are recorded by the depositional thinning of Wilcox strata above them. These allochthonous bodies provided the core structure over which Wilcox and Miocene reservoirs are folded or draped at Chinook, Atlantis, Das Bump, and other important deepwater discoveries. The location of allochthonous salt at the onset of Wilcox deposition is apparently coincident with the pronounced increase in northerly dips of the Mesozoic and Paleogene strata. This relationship is consistent with originally thick autochthonous salt above the deepest mapped basement. Sites of continued salt withdrawal from the autochthonous level into growing salt structures directly affected Wilcox sediment thickness. Such sites would have been primary candidates for the location of Wilcox sediment fairways. Identification and elimination of salt feeders would help in refining/defining these pathways. Figure 4. 3D Pre-stack depth migrated line showing a Cretaceous age salt nappe and its associated deformation front. Both features lie just basinward of the modern Sigsbee Escarpment. Thinned Wilcox and Oligocene strata show where a now evacuated salt pillow once existed. The vertical and horizontal scales are in kilometers. Abbreviations for horizons: Olig=Oligocene (orange); Wx=Wilcox (blue); K=Cretaceous (green); J=Jurassic (pink); and Bsmt=basement (yellow). Figure 5. 3D Pre-stack depth migrated line showing a second Cretaceous age salt nappe. This one lies about thirty kilometers shoreward of the Sigsbee Escarpment. Thinned Wilcox strata and an Oligocene turtle structure show where a now evacuated salt pillow once existed. The vertical and horizontal scales are in kilometers. Abbreviations for horizons: Olig=Oligocene (orange); Wx=Wilcox (blue); K=Cretaceous (green); J=Jurassic (pink); and Bsmt=basement (yellow). Deposition of the Wilcox strata can be broadly divided into two paleogeographic domains: (A) a relatively complex north-westerly region characterized by pre-existing, elevated sea-floor, salt-cored structures and sites of contemporaneous salt evacuation, and (B) a relatively simple south-easterly region characterized by a near flat and smooth sea-floor rarely punctuated by unburied basement structures. The transition between these two regions should mark changes in Wilcox depositional styles. In the more complex topographic region, Wilcox depositional events were forced to interact with relatively rapid changing sea-floor dips. Whereas in the more simple region to the southeast, a much more unconfined sea-floor presented limited impediment to widespread expansion of depositional events exiting the more complex region to the north-west. Drilling of Wilcox strata to-date has been mainly in the simpler south-easterly region and in the transition zone to the more complex Wilcox geometries towards the north-west. Figure 4 shows an example of one salt nappe and its contractional deformation front that lies in close proximity but basinward of the Sigsbee Escarpment. Thrust relationships suggest that the nappe continued to move/inflate until the end of the Cretaceous. An inflated salt pillow associated with the nappe is present through the Oligocene but then deflates during the Miocene. This interpretation is supported by the thin but depressed Wilcox and Oligocene section behind the nappe today. We predict that the edge of the salt basin lies behind the nappe, below where the Wilcox and Oligocene intervals begin dipping to the north. Figure 5 shows another example of a salt nappe that lies in about thirty kilometers inside of the Sigsbee Escarpment. This nappe does not have a deformational front associated with it. But an inflated salt pillow is associated with this nappe as in Figure 4 . Similar to Figure 4 , the interpretation is supported by a thin but depressed Wilcox section behind the nappe. In contrast, evacuation of the pillow begins in the Oligocene, as evidenced by the Oligocene age turtle structure. Evacuation continues into the Miocene until the pillow is completely deflated. The nappe remnant is all that remains of this salt body. Unique to these two examples, but possibly typical of most salt pillows around the edge of the salt basin, loading has forced salt backwards (updip) into the salt basin. In Figure 4 , the reversal of salt movement is about ten kilometers. In Figure 5 , the reversal of salt movement may be twenty to twenty-five kilometers.

Abstract A combination of 3D pre stack and post-stack time-migrated seismic data was used to examine salt structures, stratigraphy, and hydrocarbon potential in the BES 2, 100, 200 and BMES 1, 2, and 9 blocks of the Espirito Santo Basin, Brazil. Salt structures display a proximal to distal basinward transition from salt rollers, to vertical diapirs, to diapirs with overhangs and alloch-thonous tongues, and finally to salt canopies. The original autochthonous salt thickness increases following a similar proximal to distal basinward gradient. Deformation, driven by a combination of gravity gliding and gravity spreading, has been a relatively continuous process in the Espirito Santo Basin. Contraction began early in the Albian and continued unabated up to the present-day. However, individual structures ceased movement at different times depending on geometry, salt supply, and overburden thickness. A major thermal pulse affected the basin in the early to middle Eocene, associated with emplacement of the volcanic Abrolhos Plateau. Both intrusive magmas and extrusive flows are interpreted to exist. Intrusive dikes and sills display characteristic saucer shapes in cross section, elliptical to circular shapes on time slices, and cone shapes in 3D. Magmas appear to have used existing salt structures and associated fault planes as preferred pathways to reach shallower levels. Extrusive flows were identified only where seismic character, clear stratal relationships, and direct ties to intrusive geometries allowed. All the elements for excellent hydrocarbon potential exist in the Espirito Santo Basin. The main Syn-rift II source bed found in the Campos Basin exists across the basin. Several other less documented source intervals also exist. Numerous contractional folds, turtle structures, and diapir-flank traps are present. Reservoir intervals exist in Albian carbonates, Upper Cretaceous transgressive sands, and Cenozoic regressive sands. At least one deepwater hydrocarbon system is operating, as evidenced by numerous shallow bright spots, gas chimneys, and a recent major deep-water discovery. The presence of intrusive magmas may adversely affect deeper source intervals in some places, but could locally bring immature source rocks into the oil window.

Abstract Industry, regional deep-resolution seismic reflection profiles along the shallow to ultra-deepwater regions of the South Atlantic sedimentary basins allow the interpretation of different salt tectonic compartments for the whole Aptian basin, which extends from the rift border towards the oceanic crustal limit. Geological and geophysical interpretation, particularly along regional seismic transects in the eastern Brazilian and western African margins, reveals six major salt tectonic provinces associated with gravitational sliding and spreading. These tectonic provinces are characterized by particular salt families that have counterpart structural elements in the Gulf of Mexico, which shows better development of Jurassic salt basins onland and much thicker clastic input offshore in the Late Tertiary. The tectonic domains in the offshore region (platform and in deep waters) are characterized by halokinetic structures that developed huge salt diapirs, an extensional province associated with turtle structures, roll-overs and evacuation grabens, and particularly, by major development of allochthonous salt tongues. Extensional elements predominate in the proximal and intermediate provinces, whereas compressional features characterize basinward provinces near the continental—oceanic crustal boundary. Volcanic features, igneous intrusions, and wedges of seaward-dipping reflectors characterize the transition from rifted continental to oceanic crust. The deep-resolution seismic profiles also allow the identification of autochthonous and allochthonous salt structures near the crustal boundary, forming bathymetric escarpments and salt nappes that extend towards the undeformed post-salt sedimentary sequences overlying oceanic crust. Physical modelling experiments and seismic restoration using balancing techniques can constrain the seismic interpretation. We also propose conceptual plays that have not been tested in the South Atlantic but have been proved elsewhere, both in the Gulf of Mexico and in the North Sea.

Abstract Reservoirs at Pony-Knotty Head Field consist of stacked, middle Miocene (Serravallian) turbidites deposited as high-frequency low-stand successions within an increasingly ponded basin. Depositional elements include: (1) high to moderate permeability channel axes, channel margins, channelized lobes, and amalgamated lobes; and (2) those having low-permeability, such as marginal to distal lobes, levee-overbank debrites, slumped mudsheterolithics, and pelagic/hemipelagic muds. Fluid pressure data demonstrate that the Pony-Knotty Head Field is segmented into pressure compartments at multiple scales. Although the field is a low-dip, faulted, four-way turtle structure, interpreted faults are neither long enough nor have sufficient throw to segment reservoirs into observed pressure cells. Analyses of individual reservoir units indicate that variations in fluid potential are often greater vertically within wells than laterally between wells. This pattern indicates that at least some segmentation at this scale is due to low-dip stratigraphic barriers between depositional elements rather than to steeply dipping barriers, such as faults. At the field scale, both fluid pressures and depositional elements change vertically. Excess pressure was used to help define compartments at Pony-Knotty Head Field. (“Excess pressure” is the difference between pressure measured in a well and pressure calculated using a datum with an expected fluid gradient.) The deepest reservoirs have the lowest excess pressures. They are dominated by laterally continuous, unconfined depositional elements that bled excess pressure laterally. Progressively shallower reservoirs have progressively higher excess pressures in progressively more confined depositional elements. Between reservoirs of different depths and ages, stratigraphic complexity increased with time as increasing structural confinement of the depocenter above mobile salt drove stratigraphic evolution from a lobe-dominated system to a channelized lobe and levee-channel complex system. We propose that compartmentalization at this scale results directly from stratigraphic responses to the structural evolution of depocenters.

Abstract This seismic line, as interpreted, illustrates salt "rollers" or "anticlines" as they appear on the west flank of the East Texas salt basin. The rollers occur in a belt approximately 32 km (20 mi) wide parallel to structural and depositional contours. Downdip (east) of the zone of salt rollers is a zone of piercement salt domes, which represent the maximum original thickness of salt in the basin. The piercement salt domes are described on a seismic line edited by Engleman and Kemmer (this volume); the line continues eastward from the east end of this line. Updip (northwest) the salt thins to a feather edge within 64 km (40 mi) of the northwestern end of this line. Salt deformation in this zone is restricted to a few low relief pillows. An extensional fault system, the Mexia-Talco Fault Zone, is located approximately 9.5 km (6 mi) northwest of the northwestern end of this line. This fault system is believed to represent a regional pull-apart, from which salt moved downslope (east) following deposition of the Jurassic carbonates. This movement, in addition to overburden loading is thought to have been largely responsible for the anticlinal forms seen here. Timing of salt movement is complex, and appears to be episodic in some areas and continuous in others. Salt can be observed to have withdrawn almost completely on both flanks of the roller at shot point 425, as well as on the eastern flank of the roller at shot point 625. The up-turn of the Jurassic carbonates at the extreme eastern end of this line may represent the west flank of a "turtle" structure. Hydrocarbon production is associated with many of the rollers, where it is obtained principally from Jurassic carbonates. The well at shot point 425 produces from a structure in the Smackover Limestone overlying a well-developed roller. Structures over rollers may persist high in the stratigraphic section, as exemplified by the well at shot point 620 which produces from Lower Cretaceous carbonates.

Abstract Examination of 2D seismic data in South Texas has identified what is now interpreted to be a large, rafted block of Eocene, Paleocene, and Cretaceous strata, analogous to rafts identified in the Kwanza Basin of Angola. Preliminarily named the “Wilcox raft” because of its association with the Wilcox depotrough, it has been identified in the subsurface extending from Starr County on the Texas–Mexican border, northward over 200 kilometers into Live Oak County, Texas. The actual extent of rafted material may extend farther to the north and/or south. The raft’s detachment surface is interpreted to be at the base of the Jurassic Louann salt. The Wilcox raft contains one primary block more than 150 kilometers long and 15 to greater than 30 kilometers wide. The primary raft block may be segmented, and the entire rafted unit may include a number of smaller branching arms, ramps, and offset fault blocks. Various portions of the raft have downdip displacements from 5 to greater than 30 kilometers. The raft is bound on the west by expanded upper Wilcox (early Eocene) strata and on the east by expanded Queen City (middle Eocene) strata. Other incompletely detached blocks lie to the west of the raft across the Wilcox depotrough. Raft geometries suggest that at least one additional rafted block lies farther basinward of the Wilcox raft, possibly beneath expanded Vicksburg (early Oligocene) strata. A proposal for rafting in this area of South Texas is not entirely new. Earlier modeling and restorations across the Wilcox depotrough have incorporated rafts. However, these models are predicated on large-scale salt withdrawal and incorporate more than three kilometers (>10,000 feet) of autochthonous salt occupying the area of the Wilcox depotrough. We believe that a much thinner autochthonous salt layer existed beneath South Texas. In other areas of the northern Gulf of Mexico, where thick autochthonous salt existed, salt stocks are abundant. Although a few salt structures do exist in South Texas, there are very few compared with other interior salt basins. Forward modeling suggests that large sedimentary structures in the Wilcox depotrough, which can be misinterpreted as turtle structures, are related strictly to deposition during raft extension and not salt withdrawal. The geometries can be produced purely by extension on multiple detachments (Louann and lower Paleocene Midway shales) linked by ramps that dip both basinward and landward.

Abstract The petroleum geology of the Mississippi Canyon, Atwater Valley, western DeSoto and western Lloyd Ridge protraction areas, offshore northern Gulf of Mexico, is controlled by the interaction of salt tectonics and high sedimentation rate during the Neogene, and has resulted resulting in a complex distribution of reservoirs and traps. Seventy-eight fields/discoveries are evaluated and comprise structures with four-way closures (18), three-way closures (46), and stratigraphic traps (14). Three of these discoveries are in Upper Jurassic eolian reservoirs, the remainder are in Neogene deep-water reservoirs. The tectonic-stratigraphic evolution of the area is analyzed at eleven discrete intervals between 24 Ma and Present. The analyses show how the allochthonous salt systems evolved over time, and their effect on sedimentation patterns and sub-basin evolution. The study area includes some of the largest fields in the northern deep Gulf of Mexico. Thunder Horse produces from an anticlinal (turtle) structure that developed with a basement-controlled allochthonous system. The greater Mars-Ursa sub-basin has nine fields with > 1.5 BBBOE EUR, including Mars, Ursa and Princess, that developed with a counterregional allochthonous salt system. The remaining fields have considerably smaller reserves, which are controlled by the area within closure and number of reservoir intervals. Many of the smaller fields are produced from one well subsea tiebacks. Most of fields in the study area are contained within sheet-like or wedge-shaped stratigraphic intervals and have four-way or three-way trapping configurations. These findings reflect the profound effect that mobile salt has had on the petroleum geology of the region.