Basin Research - Volume 14, Issue 3, 2002

The dynamics between sediment erosion and accumulation at an alluvial basin margin affected by changes in the surface hydrology are explored using scaled analogue models produced in a flume. The presented results differ from previous counterparts in that accumulation or erosion has not been forced at a spreading outlet, but occurred at a slope change produced by previously accumulated sediment. Cyclical upstream incision produced by increased stream discharge generated incised valleys, and these were subsequently filled by sediment carried by less efficient streams generated during the low discharge period. High resolution mapping using 2.5 mm contour maps allowed the study of sediment accumulation and terrain modelling. The results of three selected experiments are analysed. The only variable explored was discharge. The basin margin was simulated by a ramp inserted in a low sloping flume, consisting of two segments of different slopes selected to emulate high and low efficiency flume fans produced elsewhere. Water and fine‐medium sand entered the ramp along a narrow (0.1 m) channel and flow expanded but without occupying the complete 1.2 m flume width. Flows were highly concentrated and noncohesive.

Fan‐like accumulation (slope: 0.11) began during low discharge (LD) periods at the ramp slope break, and proceeded upstream, onlapping quickly at first, but shifting to mostly progradation at the end of the period. High discharges (HD) usually generated two or three incised channels at the beginning of the period, but one of them prevailed and rapidly eroded parts of the LD fan and moved the sediment to a more distal low‐sloping fan (slope: 0.045). Both LD and HD fans passed downstream into a system of small parallel channels resembling a braided alluvial plain ending in sediment lobes. The mapping of the accumulated sediment during the various periods allowed calculation of sediment budgets for the entire flume. The stratal architecture of the deposits was investigated along five parallel trenches cut after experiment termination. The regression analysis of depositional profiles at fan‐like features (expanding flow) and at braided plains (parallel flow) indicated that these fan‐like systems are linear and dependent on applied discharge, while the latter showed an exponential decrease of slope downstream, with a starting value set up by the fan slope.

Two main types of stratigraphic units were generated, the LDST and HDST (system tracts). The LDST has a nonerosive base over ‘bedrock’ and the previous HDST, filling proximal erosional topography and prograding as well, generating an onlap–downlap array. Its geometry is highly variable and dependent on pre‐existing topography. The HDST base is an important erosive surface comparable to sequence boundaries. However, there are places without erosion due to a marginal position with respect to the main stream. Indeed, the results suggest that the three‐dimensional variability of erosion and depositional processes might produce very different architectures along the same basin margin.

The quantitative modelling of fluvial reservoirs, especially in the stages of enhanced oil recovery, requires detailed three‐dimensional data at both the scale of the channel belt and within‐channel. Although studies from core, analogue outcrop and modern environments may partially meet these needs, they often cannot provide detail on the smaller‐scale (i.e. channel‐scale) heterogeneity, frequently suffer from limited three‐dimensional exposure and cannot be used to examine the influence of different variables on the process–deposit relationship. Physical modelling offers a complementary technique that can address many of these quantitative requirements and holds great future potential for integration with reservoir modelling. Physical modelling provides the potential to upscale results and derive reservoir information on three‐dimensional facies geometry, connectivity and permeability.

This paper describes the development and use of physical modelling, which employs generic Froude‐scaling principles, in an experimental basin that permits aggradation in order to model the morphology and subsurface depositional stratigraphy of coarse‐grained braided rivers. An example is presented of a 1:50 scale model based on the braided Ashburton River, Canterbury Plains, New Zealand and the adjacent late Quaternary braided alluvium exposed in the coastal cliffs. Critically, a full, bimodal grain size distribution (20% sand and 80% gravel) was used to replicate the prototype, which allows the realistic reproduction of the surface morphology and importantly permits grain size sorting during deposition. Uncertainties associated with the compression of time, sediment mass balance and the hydrodynamics of the finest particle sizes do not appear to affect the reproducibility of stratigraphy between experimental and natural environments.

Sectioning of the preserved sedimentary sequence in the physical model allows quantification of the geometry, shape, spatial distribution and internal sedimentary structure of the coarse‐ and fine‐grained facies. A six‐fold facies scheme is proposed for the model braided alluvium and a direct link is established between the grain size distribution and facies type: this allows permeability to be estimated for each facies, which can be mapped onto two‐dimensional vertical cross‐sections of the preserved stratigraphy. Results demonstrate the dominance of four facies based on permeability that range over three orders of magnitude in hydraulic conductivity. Quantification of such variability, and linkage to both vertical proportion curves for facies distribution and connectivity presents significant advantages over other methodologies and offers great potential for the modelling of heterogeneous braided river sediments at the within channel‐belt scale. This paper outlines how physical models may be used to develop high‐resolution, geologically‐accurate, object‐based reservoir simulation models.

Our understanding of sedimentation in alluvial basins is best for very short and very long time‐scales (those of bedforms to bars and basinwide deposition, respectively). Between these end members, the intermediate time‐scales of stratigraphic assembly are especially hard to constrain with field data. We address these ‘mesoscale’ fluvial dynamics with data from an experimental alluvial system in a basin with a subsiding floor. Observations of experimental deposition over a range of time‐scales illustrate two important properties of alluvial systems. First, ephemeral flows are disproportionately important in basin filling. Lack of correlation between flow occupation and sedimentation indicates that channelized flows serve mainly as conduits for sediment, while most deposition occurs via short‐lived unchannelized flow events. Second, there is a characteristic time required for individual depositional events to average to basin‐scale stratal patterns. This time can be scaled in terms of the time required for a single channel‐depth of aggradation, and in this form is constant through a four‐fold variation of experimental subsidence rate.

ABSTRACT There is continued interest in how the rate of relative sea‐level rise [A ( > 0)] and the rate of sediment supply [S] function during the growth and evolution of deltaic shorelines. The theory of shoreline autoretreat, recently corroborated in flume experiments, claims that (1) A( > 0) and S can never be in equilibrium, and (2) shoreline or shelf‐edge progradation inevitably turns to retrogradation, when relative sea level is rising even modestly and even if A/S = const (> 0). Autoretreat arises because the area of the clinoform surface of the delta (or shelf edge) per kilometer of shoreline must increase as the relative sea level rises, and the delta (or shelf edge) progrades into deeper water. A finite sediment supply rate is thus liable to become inadequate to sustain progradation. The problem increases further as a rising sea level also greatly increases the delta‐plain volume that needs to be filled, further limiting the progradation of the system. The fundamental trajectory of shoreline migration is thus one characterized by a concave‐landward shape, even under the steady forcing of the basin. The magnitudes of A (> 0) and S, or A/S do not determine whether the landward turnaround of the shoreline is realized or not, but affect merely the length and height of the fundamental trajectory curve. Thus, any attempt to detect and interpret temporal changes in A and S from the observed stratigraphic record of shoreline trajectory needs first to take full account of the inbuilt autoretreat mechanism.

We develop here a simple, semi‐quantitative method of reconstructing the basin conditions (A and S) from the stratigraphic record of prograding deltaic shorelines (or prograding shelf‐margin clinoforms) on the basis of the theory of shoreline autoretreat. The deterministic nature of the autoretreat theory is advantageous in managing this latter issue, because any expected or unexpected change emerges as some discrepancy from a trajectory that was predicted for the initial conditions. The autoretreat theory also provides a convenient graphical method of dealing with the uncertainty of the field data, and with evaluating the accuracy of any reconstruction. Our methodology has been developed to deal with the behaviour of deltaic shorelines, but is basically applicable to any clinoform system, the development of which is affected by relative sea level.

The suggested method is applied to an Early Eocene (Ypresian) regressive shoreline succession in the Central Tertiary Basin on Spitsbergen. The studied regressive wedge developed as a delta‐driven, progradational shelf‐margin system under a regime of overall (i.e. long‐term) rise of relative sea level, but also suffered short‐term sea‐level falls associated with valley incisions on the coastal plain and shelf. On the assumption that S was constant or was steadily decreasing, the analysis of field data obtained from three sites within the basin suggests that the initial water depth in the basin was around 0.45 km, and that the overall relative sea‐level rise (c. 0.80 km) happened largely during an early time period and was followed by a longer period of much lower rate of rise. This pattern of relative sea‐level rise is consistent with the Palaeogene tectonic subsidence trend of the basin which was determined independently through a geohistory analysis. The uncertainty of the field data does not negate our reconstruction.

The combined effects of autoretreat and A/S changes on a deltaic shoreline trajectory are confirmed through the development of an autoretreat‐based methodology. Conventional sequence stratigraphic models that assume a possible equilibrium condition between A and S are both conceptually misleading and insufficient to analyse basin conditions quantitatively. Sequence stratigraphic analyses of shorelines need to incorporate the autoretreat concept.

Sand in laboratory scale analogue models of folding and faulting is usually assumed to behave as a simple Coulomb material in which failure occurs along planes inclined at an angle completely determined by the internal angle of friction. This simple model has inspired use of the inclined simple shear model to explain analogue model behaviour. Unfortunately the inclined simple shear model is incapable of describing the full richness of behaviour in sand‐based analogue models since it cannot cope with rotations and it cannot satisfy general boundary conditions. A rotational simple shear description is developed in this paper which overcomes these obstacles. It is shown that this description explains high‐angle back thrusts in contractional models, that it quantitatively predicts fault throws in listric‐fault extensional models and that it provides a complete kinematic description for domino‐fault block models including the transition from uniform stretch at the model base to rigid rotation in the model upper half.

The Mid‐Palaeocene palaeogeography of Denmark and the surrounding areas have been reconstructed on the basis of published geological data integrated with 3D geodynamic modelling. The use of numerical modelling enables quantitative testing of scenarios based on geological input alone and thus helps constrain likely palaeo‐water depths in areas where the geological data are inconclusive or incomplete.

The interpretation of large‐scale erosional valleys and small‐scale circular depressions at the Mid‐Palaeocene Top Chalk surface in the Norwegian–Danish basin as either submarine or subaerial features is enigmatic and has strong implications for palaeogeographical reconstructions of the eastern North Sea basin.

A 3D thermo‐mechanical model is employed in order to constrain the likely palaeo‐water depths of the eastern North Sea basin during the Palaeocene. The model treats the lithosphere as an elasto‐visco‐plastic continuum and models the lithospheric response to the regional stress field and thermal structure. The model includes the effects of sea‐level change, sedimentation and erosion, from the Mid Cretaceous to the present. Modelling results reproduce to first order geological data such as present sediment isopachs and palaeo‐water depths.

It is concluded that the Mid Palaeocene water depths in the Norwegian–Danish basin were about 250 m. The erosional valleys and circular depressions at the top of the Upper Cretaceous‐Danian Chalk Group are thus interpreted to have formed in relatively deep water rather than due to subaerial exposure. Likely interpretations of the structures are therefore submarine valleys and pockmarks.

A delay in the onset of sedimentation during fault‐related subsidence at a basin margin can occur in both extensional settings, where footwall tilting may cause a diversion of drainage patterns, and in strike‐slip basins, where a source area may be translated along the basin margin. The ‘initial depth’ created by this delay acts as pre‐depositional accommodation and is a partly independent variable. It controls the geometry of the first stratal units deposited at the basin margin and thus modifies the response of the depositional system to subsequent, syndepositional changes in accommodation. In systems with a sharp break in the depositional profile, such as the topset edge in coarse‐grained deltas, the initial depth controls the foreset height and therefore the progradational distance of the topset edge. The topset length, in turn, influences topset accommodation during cyclical base level variations and therefore is reflected in the resulting stacking patterns at both long‐ and short‐term time scales. In the simplified cases modelled in this study, it is the relationship between the initial depth and the net increase in depth over the interval of a relative sea‐level cycle (ΔH) that governs long‐ and short‐term stacking patterns. In situations where the initial depth is significantly larger than ΔH, the topset accommodation of the first delta is insufficient to contain the volume of sediment of younger sequences formed during subsequent relative sea‐level cycles. Therefore, the depositional system tends to prograde over a number of relative sea‐level cycles before the topset area increases so that the long‐term stacking pattern changes to aggradation. Stacking patterns of high‐frequency sequences are influenced by a combination of topset accommodation available and position of the short‐term relative sea‐level cycles on the rising or falling limb of a long‐term sea‐level curve. This determines whether deposits of short‐term cycles are accommodated in delta topsets or foresets, or in both. Variations in stacking pattern caused by different initial depths may be misinterpreted as due to relative sea level or sediment supply changes and it is necessary to consider initial bathymetry in modelling and interpretation of stacking patterns, especially in fault‐bounded basins.

The Quaternary glaciations had a profound impact on the geomorphology and stratigraphy of passive continental margins. The challenge is to resolve the contributions of the main forcing controls relative sea‐level change and sediment flux. The key to answer this question is to understand the interaction between the marine and terrestrial environments, where river dynamics play an essential role. A comprehensible three‐dimensional numerical model is presented in order to investigate quantitatively the behaviour of river–shelf sedimentary systems under glacio‐eustatic conditions. Distinctive features observed in the model results include river avulsion, delta‐lobe switching, incision and knickpoint migration. An important event in the development of the modelled river–shelf system is the establishment of a direct and inextricable link between the drainage basin and the depocentre on the shelf edge, thereby bypassing the exposed shelf. This is termed as ‘drainage connection’. In the model, the timing of drainage connection occurs over a broad interval when the model run is repeated many times with small differences in the initial topography, reflecting the sensitivity of the system to its initial state. It demonstrates the inherent variability in the evolution of a sedimentary system as a consequence of non‐linear behaviour. A statistical approach to modelling is suggested in order to deal with this problem.

This paper describes a new 3‐D forward numerical model (CARBONATE 3D) that simulates the stratigraphic and sedimentological development of carbonate platforms and mixed carbonate–siliciclastic shelves by simulating the following sedimentary processes: (1) Carbonate shallow, open‐marine production, dependent on water depth, restriction and sediment input; (2) Carbonate shallow, restricted‐marine production, dependent on water restriction; (3) Pelagic sediment production and deposition; (4) Coarse and fine siliciclastic input; (5) Erosion, transport and redeposition of sediment, dependent on currents, slope, depth and restriction as well as sediment grain‐size and composition; (6) Dissolution of subaerially exposed carbonate. In this paper the model is used to investigate the controlling mechanisms on the sequence stratigraphy of isolated carbonate platforms and atolls and to predict distinctive architectural signatures from different drowning mechanisms. Investigation of the mechanisms controlling atoll strata shows that although relative sea‐level is the major control, antecedent topography, environmental setting and early diagenesis have profound influence on what stratigraphic geometries and facies develop. Hence care must be taken if sea‐level curves are interpreted from real stratigraphies. Atoll drowning by fast sea‐level rise, by lowered production and by repeated exposure and fast subsequent sea‐level rises are investigated and different stratigraphic signatures for the respective mechanisms predicted. A fast relative sea‐level rise results in a bucket‐shaped morphology developed prior to drowning and a sharp transition from the platform margin facies to a pelagic cover. Drowning caused by lowered platform margin production is predicted to result in the development of a dome‐shaped, shallow‐water shoal over the whole platform top prior to drowning. Fourth order amplitudes of several tens of metres, typical of ‘icehouse’ settings, cause atoll drowning at subsidence rates where atolls subject to fourth order amplitude of only a few metres, typical of ‘greenhouse’ settings, can keep up with the rising sea‐level. In the resultant strata, vertical facies belts are less well developed but horizontally extensive facies bands are more prominent. High fourth order amplitudes (up to 80 m) without sufficient third order scale subsidence will not lead to drowning, however, as the platform can recover in each fourth order lowstand. These results suggest that atolls might be easier to drown in ‘icehouse’ rather than in ‘greenhouse’ conditions but only in situations with suitably high rates of longer‐term relative sea‐level rise or sufficient lag times.

Forward stratigraphic models usually display sediment types on simulated stratigraphic profiles as ‘facies’ defined only by their depth of deposition. More recently, ‘facies’ have been defined and displayed in terms of the dominant processes of deposition (e.g. in situ growth, pelagic production, turbidite deposition). Standard carbonate facies; that is, the Dunham classification, are defined by rock textures and grain composition that imply that a combination of processes acted together to generate a facies. For example, a bioclastic wackestone is a matrix‐supported rock containing up to 90% matrix and > 10% shelly grains. In terms of modelled processes, the muddy matrix could be generated by: (i) reworking of the shallow platform sediments, (ii) from pelagic deposition, or (iii) in situ production. A traditional depth of deposition process display would not be able to distinguish such a wackestone from any other facies deposited at this water depth and a majority process display would not combine reworked, pelagic muds and in situ contribution in one simulated ‘facies’. This paper introduces a new scheme that enables forward models to output simulated facies defined by a range of values for each of the controlling processes and thereby predicts rock textures within simulated stratigraphies. This approach has been applied to the Jurassic carbonate ramps of the Iberian Basin in northeastern Spain. It is shown to provide more accurate information about the processes that are being simulated, allowing more direct comparisons to be made with the facies observed in the field and providing potential for a more rigorous method for assessing the ‘goodness of fit’ of a simulated stratigraphy.

Late Visean (Asbian–Brigantian) platform carbonates in the British Isles show a pronounced cyclicity marked by the alternation of mainly subtidal carbonates and subaerial exposure surfaces. Whereas the cyclicity of shallow‐water limestones of this age has been well‐documented, there has been little attempt to understand the controls on larger‐scale patterns such as those recognized in Pennsylvanian successions in the U.S.A. The principal aim of this study is to test two contrasting theories of cycle stacking via numerical forward modelling. Earlier studies of Pennsylvanian‐early Permian platform carbonates in south‐west U.S.A. suggest that cycle stacking patterns were controlled by the interaction of third‐ and fourth‐order sea‐level oscillations, with relatively uniform fourth‐order oscillations altered mainly by the harmonic effect of lower‐order sea‐level changes. An alternative model is based on an insolation curve for the Carboniferous calculated using Milankovitch parameters. This model predicts a considerable variability in levels of solar insolation that would have affected the amplitude of fourth‐order sea‐level changes and cycle composition. Both of these ideas were examined via numerous model runs using CARBONATE 6.0 and a new program, CARBOSMUT. Model results were evaluated through the use of 3 key criteria derived from well‐documented outcropping stratigraphies in the U.K.: (1) cycle stacking patterns and the stratigraphic position of major transgressions, (2) stratigraphic position of major faunal changes, (3) degree of development of subaerial exposure surfaces. Computer simulations and comparison with outcrop data suggest that a model invoking the interaction of relatively uniform fourth‐ (c.100 Kyr) and third‐order sea‐level oscillations is most appropriate for much of the late Visean, with major lowstands occurring at the mid‐Asbian and Asbian–Brigantian boundary. Late Visean cycles are important exploration targets in the Pri‐Caspian Basin, Kazakhstan and understanding the controls on stratal patterns is important as a potential exploration tool.