Multi-Scale Forward Modelling of Microbial Lacustrine Carbonates

Microbial lacustrine carbonates have been the focus of investigation in recent years due to the discovery of giant oil fields in the South Atlantic. Two numerical forward stratigraphic models, one modified and one entirely new, are used to explore controls on large-scale and smaller-scale processes and characteristics of carbonate lacustrine strata.
The large-scale model (Carbo-CAT) focuses on exploring kilometre scale carbonate stratal heterogeneity developing in extensional settings. The modelled processes include spatial facies distributions calculated with a cellular automata (CA), carbonate production controlled by water depth- dependent curves and spatial distribution of dissolved carbonate and sediment transport as a slope-dependent mechanism. Finally, the subsidence produced by various 3D fault configurations can be included. The model predicts carbonate platform geometries controlled by active extensional faults and spatially variable platform production. The model also explores stratal heterogeneity generated by factory competition modelled with a cellular automata.
The small-scale model (Mounds-3D) investigates the controls on microbial mound development in the metre to decametre scale. Modelled microbial growth includes precipitation and trapping and binding processes. These are affected by energy, slope and spatial distribution of the microbial community. The model incorporates a depth-averaged hydrodynamic model to assess the impact of transported sediment deposition, erosion and trapping and binding in mound development. The shape, size and distribution of mounds are controlled by rates of in-situ production, rates of sediment transport, hydrodynamic energy and the rate of accommodation creation.
The number of processes included in each model, and the high number of parameters that can be defined by the user to control them, require a systematic and thorough analysis of the relationship between input and output to extract correct conclusions. The inherent cell-size and time-step sensitivity of the cellular automata in Carbo-CAT and the coupled growth and hydrodynamic models in Mounds3D require being careful when reporting results. The models are more suitable for addressing the relative impact different processes have on the resulting strata, rather than offering absolute answers on the parameter values that produce certain features.
To reduce the uncertainty that the initial parameters controlling the CA can produce over the facies distribution in Carbo-CAT, the results are measured in terms of the vertical and horizontal heterogeneities. Although the CA typically exhibits non-linear responses to changes in the initial parameters, this work shows that trends can be extracted to predict the outcome of the facies population process in terms of the vertical heterogeneity produced. The question remains on why a particular set of cellular automata rules should be picked over another. As best practice, the uncertainty that different rules introduced in the model should be quantified and reported.
Exploring the effect different fault behaviours have on the carbonate strata with Carbo-CAT show the relative impact fault evolution can have on the stratal architecture. Examples of equifinality suggest that simple conceptual models of syn-rift carbonate deposition have limited predictive value. The inherent complexity of carbonate systems requires diverse depositional models, since direct cause and effect behaviours need to consider a wide range of aspects of the depositional setting. For example, the size of relay ramps can affect the development of carbonate platforms over linking faults, which in turn affects transport paths and the resulting deposition in the hanging-wall.
Microbial mounds modelled in Carbo-CAT show that both autogenic and allogenic processes and controls can affect their characteristics (shape, size and distribution). The tests show that, although the initial bathymetry exerts a first order control on their initial location, other processes are also key on their final distribution. Although further testing is required, the dynamic behaviour of the modelled microbial systems suggests that mound spacing is also controlled by local variations on the hydrodynamic and depositional conditions.
Both models are used to evaluate, challenge and improve large and small-scale outcrop-derived depositional models of the Mupe Member microbialites (Purbeck Limestone Group). The impact of active tectonics in the distribution of thicknesses and facies in the basin is explored with Carbo-CAT. This study refines tectono-stratigraphic models linking fault activity to specific depositional patterns and internal stratal relationships. A fair match between measured and modelled data in terms of thickness in scenarios with and without active subsidence suggests that this data alone are not enough to disregard any of the two options. In some cases, the models show the same deepening-shallowing cycle development that was interpreted from facies stacking in the field. The models can be used to improve correlations of these cycles, where the resulting architecture can be associated to different tectonic activity.
Controls on the outcrop-scale distribution and shape of metre-high mounds are studied with Mounds3D. With this study, the outcrop-derived depositional mound model is expanded with quantitative understanding of internal system dynamics that control mound geometries. The study supports the outcrop-derived suggestion that lake level can have a major impact on the expansion and contraction of mounds. It also suggests that minor inter-fingering with intermound sediment can be a function of the system’s dynamic. The model also allowed investigation of the origin of microbial mounds not directly linked to bathymetric highs, which can be reproduced in the model. Their inception requires coherent feedback between the in-situ production and sediment transport processes.