Palaeogeography has too often been considered as simply a backdrop for presentations and posters. But methodological advances over the last 20-30 years, coupled with advances in Geographic Information Systems, mean that plate tectonic modeling and palaeogeography are now becoming an essential and powerful tool for New Ventures exploration. This is nowhere better demonstrated than in the Equatorial and South Atlantic Petroleum Provinces, in which the interplay of plate tectonics and landscape evolution affects not only the geometry and nature of basins around the margins, but source to sink relationships and depositional systems, with direct implications for exploration risking.
The fundamental problem of New Ventures (NV) exploration, or perhaps its very scientific attraction, is that it is focused on areas at the edge of our knowledge. These are geographic areas with few, if any, wells or seismic lines. Such areas provide the geological puzzles that attracted many of us to geology in the first place, but are a serious headache for explorationists.
The South Atlantic and Equatorial Atlantic hydrocarbon provinces, north of the Walvis Ridge, are a case in point. With proven oil reserves of over 285 Bbo (CIA Fact Book, 2010) and recent major discoveries in the sub-salt of the Santos Basin and in the deep water off Ghana, interest in new plays in this region has increased considerably over the last few years. But existing discoveries are not evenly distributed, and with high costs and risks associated with deep water exploration, low cost NV tools to mitigate this risk are essential.
At this early stage in the exploration process, even the most fundamental questions are critically important and need answers before additional, expensive resources are purchased. Quite simply, is there a basin present with enough sediment to generate hydrocarbons and what is the composition and timings of the fill of these basins? Is there a potential source, reservoir and seal rock? Does a play exist?
Industry experience in the South Atlantic means that we can also anticipate more specific facts that will be important, such as the extent of source facies and reservoirs in the sub-salt and the nature of timing of uplift along each margin that can affect maturity modeling. What, for example, is the spatial variation in source and reservoir facies and how does this relate to tectonics, especially the interaction with transforms in the Equatorial region, but also hinterland reactivation and sediment supply. We also need to ask what is the location, timings, volume, provenance and character of clastic accumulations along the margins and how does this relate to changes in hinterland evolution?
Plate Modeling and Palaeogeography
This is where plate modeling and palaeogeography can provide part of the solution.
Plate modeling sets the limits to the geodynamics of an area and thereby the potential development and evolution of basins along a margin (including geometry, crustal type and thickness which lead to heatflow, subsidence and uplift). It also provides the platform on which Gross Depositional Environments (GDE) and landscape models can be compiled and developed; this includes hinterland uplift histories (driven by geodynamics including mantle dynamics), which then affects source to sink relationships, sediment flux and character.
The palaeoenvironmental reconstruction adds flesh to this plate reconstruction, through mapping what we know and what we think may have existed, either by extrapolation from adjacent areas or lithofacies prediction based from process-based modeling. This helps define the depositional setting of frontier areas. By converting this to elevation (palaeotopography and palaeobathymetry) and adding interpretations of palaeodrainage, the maps become more powerful predictive tools, since this level of analysis does not only indicate where sediment might be expected, but goes some way to allowing us to predict what that sediment would comprise (given reconstructions of weathering and erosion, climate, bedrock composition and so forth). This can be taken even further using process-based predictive lithofacies models which use the palaeolandscape as a boundary condition for vegetation, climate, ocean, tide and wave models, that in turn provide the input for lithofacies modeling.
Building the Framework
Building robust plate models and palaeolandscapes is not a trivial matter. Both are labor intensive, requiring large, multi-disciplinary teams of specialists, sizeable databases and constant testing and iteration. But they are only one part of a more extensive integrated workflow designed to solve NV problems, which also includes petroleum geochemistry, stratigraphy, sedimentology and petroleum geology.
To build the foundations of a model in an area of sparse well and seismic data, greater reliance must be placed on potential fields and other remote sensing datasets. Interpretation of these data is the starting point of the workflow and is used to map the structural framework of each area, as well as the extent of volcanics and salt. For the South Atlantic and Equatorial region reconstruction, 16 2D profiles were then used to define the crustal type, geometry and thickness, the limits of stretched crust and position of the Continent-Ocean Transition zone. These were employed as boundary conditions for the plate modeling and to calculate beta-factors as tests of the modeled fits. Differentiation of the transitional crust into attenuated, continental, under-plated and intruded segments provides additional important information for estimating heatflow in developing basins. Depth to basement calculations defined extensions to known depocentres as well as identifying deep-seated structural partitioning in basins with a complex history such as the Santos Basin.
For the South Atlantic about 17,000 structures were compiled using interpretations based on potential fields, Landsat and radar altimetry (SRTM3) data. These were constrained by comparison with seismic lines where available, and by onshore geological maps and publications.
These structures were then used in conjunction with the results of the 2D geophysical modeling to redefine basin geometries along the margin and to understand the nature and kinematics of plate boundaries used in the plate modeling. This includes intra-continental plate boundaries within Africa and South America, which have to be considered in order to account for the timing and geometry of the developing South Atlantic rift during the Late Jurassic to Aptian. This is also true in the model generated in this study, although uncertainties remain about the exact amount of deformation on many of the large, Precambrian Shear Zones. What is clear is that neither the simple northward ‘scissor-opening’ of the South Atlantic of Bullard et al (1965: Philos. Trans. R. Soc. London., Ser. A, 258: 41-51) or the idea of a single, simple translational opening of the Equatorial Atlantic, fits observations. This has a major affect on not only basin development and history, but also the response of the hinterland, which in turn affects source to sink relationships.
Reconstructing Contemporary Base-level
With the structural and tectonic framework established, the next step is to map the depositional systems. Traditionally this is restricted to compiling facies or GDE maps for what is preserved, but in our methodologies this is extended to include the probable contemporary full extent of deposition at the time of the map. Differential symbology is used to distinguish between those interpretations of depositional environment that are known (preserved) or inferred, with a further division applied to lithologies as to whether the designation is based on ‘outcrop’, ‘subcrop’ or ‘inferred’, providing an immediate indication of confidence. Sediment source areas (areas above contemporary base-level) are then mapped and defined according to the last thermo-mechanical event affecting that area (at the simplest level this is assumed to be the reason that such an area is above base-level, but in reality most upland areas are a consequence of multiple causes). Structural and tectonic elements are also mapped on these reconstructions, together with exploration and cultural data, well information, source rock distribution and so forth, including any key datasets held by the NV team. Given the complexity of the system, it is perhaps of little surprise that frequently findings from this phase of the mapping can lead to modifications in the underlying plate model, which increases the accuracy and precision of the map, but adds considerably to the workflow.
Whilst GDE, tectonic and base-level maps provide an explicit representation of source to sink relationships and the extent and nature of depositional environments, high on any NV wish list is the ability to predict confidently the extension of the various play elements. With potential fields data providing a robust means of reconstructing the geometry of basins and crustal types, including possible extensions and structuralization, converting palaeoenvironmental maps to landscape maps provides the boundary conditions for quantitative analysis and lithofacies prediction.
There are several stages to this process. The present day landscape is analyzed to identify the relationship of geology and tectonics in landscape formation, and then the drainage networks are examined in order to identify potential changes in river systems and thus past sediment transport pathways. These analyses point to substantial changes on both West African and South American margins of the South Atlantic, with major changes in the river systems including the Congo, Orange, Niger, Amazon and Sao Francisco. These can be tied to tectonic and base-level related causes that have affected the region through the Mesozoic and Cenozoic, and can therefore be mapped through time.
With rivers defined, elevation is added to the tectonophysiographic and depositional systems. First, the Present Day topography is rotated back through time as a very cursory backdrop to the analysis, and the elevation distribution represented by each tectonophysiographic terrain in the present is applied to its past representation on the palaeogeography maps. Fission track and other palaeoaltimetry data are used to refine elevations, palaeosols to assess areas of low denudation, and mass balance calculations to evaluate whether elevational changes are realistic with respect to downstream sediment accumulations. The methodology is fraught with uncertainties and there are constant iterations back and forth through time to ascertain whether defined changes are consistent.
The resulting landscape maps form the boundary conditions for Earth System Modelling experiments, the results of which can be directly used in predictive lithofacies modeling. In the South and Equatorial Atlantic the results are dominated by two main effects: long wavelength uplifts associated with mantle anomalies in southern and eastern African hinterlands, north-eastern Brazil and northern South America; and reactivation of uplifts, often affecting whole cratonic blocks such as the Gabon Craton, due to changes in the pole of rotation in the South Atlantic (for example, the Santonian Event, which resulted in uplifts across the Equatorial African hinterland and inversion in the Central African Shear Zone). These two types of tectonically driven hinterland change have very different expressions in the landscape. Denudation rates are much lower on the plateau highs associated with the long wavelength uplifts, with sediment supply to downstream basins dominated by incision around the margins of these uplifts, especially if there is associated reactivation of fault scarps. The inversion related uplifts seem to lead to much more intense erosion and sediment supply because of the rate of uplift but especially the resulting relief changes.
The tectonic and landscape evolution of the South Atlantic and Equatorial Atlantic is complex, with numerous periods of reactivation that affects source to sink relationships as well as downstream accommodation space, and the geometry and nature of depositional systems. Only by pulling together all of these various processes can such tools provide robust solutions for NV exploration. But when done correctly, the results are invaluable.