H54B-01
The Pulse of Calm Deltas
At the heart of interpreting the history of Earth-surface evolution preserved in the rock record is distinguishing environmental (allogenic) forcing from internally-generated (autogenic) ¢®¡Ænoise¢®¡¾. Allogenic deposits have classically been recognized by their cyclic nature, apparently resulting from periodic changes in base level, sediment supply, or tectonics. Autogenic deposits, although quite variable in their origin and scale, are caused by the nonlinearity of sediment transport and expected to have a random or scale-free (fractal) signature. Here we describe a robust mechanism that generates cyclic deposits by an autogenic process in experimental fan-deltas. Sheet flow over the fan surface induces deposition and an increase in fluvial slope, until a critical slope is exceeded leading to a channelization instability. Channelized flow results in incision and degrading of the fan surface to a lower slope, releasing a pulse of sediment that pushes the shoreline forward. Sheet flow resumes once the surface is regraded, and the cycle repeats in a surprisingly periodic fashion to produce cyclic foreset parasequences. We use a one-dimensional fan evolution model to (1) demonstrate how time-varying flow width can cause pulses in sediment discharge at the shoreline in agreement with the experiments, and (2) scale our results up to field conditions. Alternating sheet- and channelized-flow is known to operate on non-cohesive fans in nature. Our results suggest that, rather than reflecting variation in environmental forcing, many observed cyclic sedimentation may be a signature of the autogenic ¢®¡Æpulse¢®¡¾ of deltas under calm environmental conditions.
H54B-02
Channel Extension in Deep-Water Distributive Systems
The cyclic nature of channel and lobe formation in submarine fans is the result of the unstable and ephemeral nature of newly formed distributary channels. Avulsion cycles are initiated as unconfined sheet flow immediately following avulsion followed by stages of channel incision and extension, deposition of channel mouth deposits, and often channel backfilling. In contrast with those in alluvial and deltaic environments, avulsion cycles in submarine fans are relatively poorly understood due to the difficulty of observing deep ocean processes, either over short timescales needed to measure the hydrodynamics of active turbidity currents, or over longer timescales needed for the morphodynamic evolution of individual distributary channels and avulsion events. Here we report the results of over 80 experiments in a 5m x 3m x1m deep tank using saline (NaCl) density flows carrying low-density plastic sediment (SG 1.5) flowing down an inclined ramp. These experiments were designed to investigate trends observed in earlier self-organized experimental submarine fans with well-developed avulsion cycles, in which distributive lobes were observed to form on relatively high slopes. In particular, we were interested in investigating the relationship between channel extension length (distance from the inlet to the point where the flow becomes de-channelized, transitioning into a mouth-bar/lobe) and slope. The results of the experiments are clear but counter-intuitive. Channels appear to extend in discrete segments and channel extension length is inversely related to slope over a wide range of slopes (5-17 degrees). In addition, channel extension seems largely independent of inlet flow density (salt concentration) over the experimental range (10-24 g/cc). Measurements of densimetric Froude number (Fr') indicate Fr' increases downstream to near critical conditions at the channel lobe transition. Our preliminary interpretation is that distributary channels become unstable due to acceleration to Fr'-critical conditions and the formation of a depositional hydraulic jump, which perturbs sediment transport and ends channel extension. Similar morphodynamic length scale controls are observed in shallow water fan-delta experiments (e.g., SAFL DB-03) and in 2-D depositional cyclic steps. The experiments seem to explain two interesting observations from the earlier self-organized fan experiments and from real submarine fans. Firstly, the observation of ‘perched' fills at the steep entrances to salt withdrawal minibasins (e.g., in the Gulf of Mexico) suggesting higher sedimentation rates (or inefficient sediment transport) on higher slopes (initially higher than at the slope break downstream). Secondly, strong progradation as the fan evolves and slope decreases in ‘perched‘ fans suggests increasing flow efficiency on lower slopes, at least over a certain window of parameter space. Apparently deep water systems have a tendency to self-regulate even when flows differ significantly in initial density. The observed modulation to Fr'-critical flow appears to be an important control on length scales in deep- water distributive channel systems, potentially explaining strong deepwater progradation or ‘delta-like' patterns that have remained paradoxical. Near critical conditions have been inferred from observations of many active submarine fans but the extent to which these results from conservative density currents apply to non-conservative and potentially ‘ignitive' turbidity currents is the subject of ongoing investigation.
H54B-03
Large wave-shaped bedforms in the axial channel of Monterey Submarine Canyon: Monterey Bay, California
Multibeam bathymetric data show that large wave-shaped bedforms exist on the seafloor within the axial channel of Monterey Submarine Canyon offshore northern California (Smith et al., 2006). These features have wavelengths up to 70 m, amplitudes up to 2 m, and distinct asymmetrical crests that are roughly perpendicular to the channel. Comparisons of repetitive multibeam surveys since 2004 shows that the bedforms are active features because their positions change between surveys. Three complementary studies are underway to understand the origin of these features: (1) Vibracoring - In June 2007, the ROV Ventana collected 18 vibracores up to 2 m in length along a 130-m transect in ~285 m water depth that spanned the crests of two and the flanks of three waves. Sediment in these cores is composed of one or more sequences of coarse gravel or multicolored clay-clasts that fine upward into sand. Sometimes individual gravel-clasts or clay-chips occur within sand. The internal stratigraphy of these waves shows they resemble classic gravity-flow deposits. (2) Sediment Movement - A pilot study was conducted to assess whether sediment within the canyon floor moves by traction from currents or mass transport. On February 8, 2007, three acoustic beacons were deployed in ~290 m water depth within the canyon axis using Ventana. The beacons were placed within recesses in 50-cm-high ~45 kg poured-concrete monuments. These boulder-sized monuments were buried leaving only the top of the beacon standing ~6 cm above the sediment surface. Thus, the monuments were largely entombed within the seafloor. We also placed 3 acoustic beacons mounted on trapezoidal frames at the edge of a terrace on the canyon's lower flank. On February 12th, we returned to the area and determined that all three monuments had moved ~150 m down canyon. Two trapezoidal frames were found on their sides entwined with each other 50 and 75 m down canyon from their deployment site. The third frame was never located. A sediment trap mooring downstream at 1,300 m water depth independently showed that there was a gravity flow event on February 9th. The locations of the acoustic beacons have been re-determined on 4 subsequent occasions and they have not changed as of June 28, 2007. Because the monuments moved more than one wavelength down-canyon during the February 2007 event, aliasing problems may occur when trying to track the movement of individual wave crests in the repeat mapping surveys. The movement of the buried monuments suggests that the seafloor was remobilized to more than 50-cm-depth during this sediment transport event. (3) Repeat Mapping - Repeat surveys, including AUV multibeam mapping and chirp sub-bottom profiling, are being conducted 3-4 times per year. These surveys show that the wave-shaped bedforms occur in the canyon axis down to at least 1,100 m water depth. No internal structures are seen in the chirp data. Tracking the movement of these waves is proving difficult because it depends on identifying individual waves in successive surveys. Given that initially buried boulder-sized monuments moved more than two wavelengths down-canyon during an individual event limits our confidence in the identification and tracking of individual waves. Apparently, these wave-shaped bedforms are produced during brief, discrete mass transport and/or gravity flow events.
H54B-04
Characterizing Turbidity Current Flow Conditions From Turbidite Grain Size Distributions, Capistrano Formation, San Clemente, CA
Turbidity currents exert fundamental controls on the evolution of continental margins, yet the physical properties of these flows remain poorly constrained because they are difficult to observe directly in nature or model accurately in experiments. We develop a method to characterize turbidity current flow conditions using grain size and morphologic observations of single-event deposits from the Miocene Capistrano Formation near San Clemente, CA. Observed deposit styles include beds that onlap channel walls indicating time transgressive deposition from a basally charged flow and drape morphologies indicating deposition from flows that concentrate sediment above preexisting topography. While spatial variations in particle size within turbidite channel cross sections exhibiting onlap morphologies provide flow-condition information at a single point through time, particle size variations in drape deposits provide instantaneous vertical flow-structure constraints. We interpret the distribution of fines that does not vary with elevation in the deposits (D<300 microns) to represent the fraction that was fully suspended at some up-dip location and remained distributed throughout the current at the deposition site. Upward fining of the coarse tail of the distributions in an onlap deposit indicates progressive removal of the coarsest particles in suspension from a decelerating current through time. A simple sediment transport model produces vertical sediment concentration profiles as a function of flow conditions that are consistent with deposit grain-size variation. Modeled profiles indicate current velocities of up to 8.4 m/s and deposition times of minutes to hours, and place constraints on the down-channel spatial evolution of the flows. Landscapes are shaped by the net effect of many erosional and depositional events that occur at varying frequencies and magnitudes. Application of this approach to a large number of turbidites would represent a first step towards characterizing the population of possible flow conditions in order to understand how turbidity currents and submarine topography interact.
H54B-05 INVITED
Applying the Concept of Grade to Basin-Scale Stacking Patterns and Stratigraphic Architecture Along the Shelf to Basin Profile: An Outcrop Perspective
The concept of grade is used to describe the net effect of numerous sedimentary processes that, over time, define the morphology of the upper bounding surface of sediment deposition. Two end members characterize the continuum of basin margin morphology: graded and out-of-grade. Recent studies have related basin-margin morphology to specific regional stratigraphic stacking patterns. Graded basin margins contain graded (equilibrium) depositional profiles that advance or prograde into the basin. The morphology of these profiles remains unchanged through time and space. Out-of-grade basin margins contain over-steepened profiles, and as a result, sediments bypass the shelf edge and the proximal slope and accumulate at the base-of-slope position. The morphology of profiles in out-of-grade basin margins changes through time and space. While regional studies have related large-scale stratigraphic stacking patterns to grade, there are no published results documenting the effect of grade on local stratigraphic architecture along shelf to basin profiles. The shelf to basin floor stratigraphic architecture of the Upper Cretaceous Lewis Shale of southern Wyoming, USA and the Upper Carboniferous Ross Sandstone of western Ireland represent graded and out-of-grade depositional systems, respectively. The graded Lewis Shale contains progradational basin-scale stacking patterns, whereas the out-of-grade Ross Sandstone contains aggradational basin-scale stacking patterns. Although the two systems represent end members on the grade to out-of-grade continuum, they have several similarities: (1) both were deposited in epeiric seaways, (2) both were deposited in a time interval of ~1.2 My, (3) both were deposited during high-frequency eustatic changes in sea level, and (4) both have 500m of submarine fan accumulation. Despite these similarities, several differences are noted that appear to be related to grade. They are: (1) architectural and facies associations, (2) presence/absence of incised valleys, (3) presence/absence of submarine canyons, (4) lengths of submarine fans, (5) partitioning of sandstone across the physiographic profile, (6) position of submarine fans through time, and (7) position of submarine fans relative to depocenters. Stratigraphic and structural analysis of the Lewis Shale and Ross Sandstone support predictions made from stratigraphic modeling studies for first-order controls on grade. These controls are initial morphology of the slope, tectonically forced changes during deposition, and sediment supply.
H54B-06
3D stratigraphic modelling of the interaction of marine and land processes
In order to investigate the interaction of marine and land fluvio-deltaic processes we have developed a fast yet representative model with a coupled fluvio-deltaic drainage network module and a marine sedimentation module. The focus of the model is on sediment dynamics and stratification on geological timescales (1000 years and longer), therefore the routines represent only those processes relevant on geological time scales. The algorithms are implemented in the overall architecture of a 3D stratigraphic model, which allows us to evaluate the resulting stratigraphy. The fluvio-deltaic setup is based on subgrid parameterization and is capable of producing convergent and divergent channel networks. The channel network is assumed to be stable, yet avulsions are allowed to develop out of randomly instigated crevasses. Channel stability is modelled one dimensionally by calculating the flow and sediment transport at prospective avulsion nodes. Marine sedimentation and erosion is represented by, longshore currents, river jets and wave influence. Hypopycnal plumes and longshore currents are integrated in one potential flow routine, thus allowing rapid calculation. Wave generation and propagation is modelled independently. The wave-current interaction allows reworking of previously deposited sediments, and consequent resuspension. The sediment dispersal of the resuspended load is assumed to occur exclusively due to longshore currents, as the effective velocity vectors of the waves are mostly orientated crosshore. Previous modelling efforts have shown that, under fully time-invariant forcing changes in avulsion frequency are correlated with the number and length of distributary channels, which are in turn related to alternating phases of progradational and aggradational delta development. This fluctuation in sediment transport results in changes of sediment flux (i.e. storage-and-release events) to the marine basin and ultimately the morphology of the delta front and prodelta. Additionally marine processes such as waves and longshore currents influence the morphology of the delta. Wave reworking may remove nearshore deposits or block potential outflow routes, which will in turn influence the progradation of the delta and consequently the fluvio-deltaic channel network.
H54B-07 INVITED
Scaling Relationships and Evolution of Distributary Networks on Wave Influenced Deltas
Relatively little is known about the temporal and spatial scales of processes that evolve distributary networks, and no genetic model can explain the variability in distributary network pattern on modern deltas. We derived scaling relationships for two fundamentally different processes known to create distributary channels and, with these laws, constructed a simple model for distributary network evolution on wave-influenced deltas. The first process is mouth-bar deposition at the shoreline and subsequent channel bifurcation; the second is avulsion—the wholesale abandonment of a channel in favor of a new path. The former, which is generated by a fluid- mechanical instability at the shoreline and is fueled by local progradation of the delta front, creates relatively small networks with power-law distributions of channel length. In contrast, avulsion arises from a gravitational instability between the superelevated channel and its surrounding floodplain and, in principle, can operate anywhere on the delta surface. The avulsion process generates relatively few but larger distributaries that are bounded in scale by the backwater length. We collected channel statistics from a representative set of modern deltas. Frequency-magnitude plots of channel length agree with our theoretical predictions and show a clear separation in scale that reflects the two channel-forming processes: Mouth-bar distributary lengths scale with the width of the parent channel, and avulsive distributary lengths scale with the backwater length; interestingly, intermediate channel lengths are relatively rare. Wave energy controls network topology by suppressing mouth- bar development, thereby preferentially eliminating smaller-scale, mouth-bar distributaries. Hence, the absence of fine-scale channel structure, as quantified by deviations from expected fractal scaling, provides a metric of the relative importance of waves.
H54B-08
Distinguishing Long-Term Controls on Fluvial Architecture in the Lance Formation, Bighorn Basin, Wyoming
Allogenic processes are considered a prime control on the stratigraphic distribution of channel bodies, however, recent studies have indicated that autogenic stratigraphic organization may occur within fluvial systems on basin- filling time scales (105-106 years). Groupings or clusters of closely-spaced channel bodies can be produced by several different mechanisms, including both allogenic and autogenic processes. Commonly, sand- dominated intervals in stratigraphic successions are interpreted as incised-valley fills produced by base-level changes. In contrast, long-timescale organization of river avulsion can generate similar stratigraphic patterns. For example, sand-dominated intervals in the fluvial Lance Formation (Maastrichtian; Bighorn Basin, WY) have been interpreted as incised-valley fills formed during sea-level lowstand. However, closely-spaced sand bodies in the Ferris Formation (Lance equivalent; Hanna Basin, WY) are interpreted as aggradational in origin, and have been compared to autogenic avulsion stratigraphy produced in experimental basins. We evaluate the Lance Formation in the southern Bighorn Basin in an effort to determine whether these sand-dominated intervals are truly incised- valley fills resulting from sea-level changes, or if they were generated by autogenic processes. The Lance Formation crops out in the western and southern margins of the basin, exposing relatively proximal and distal portions of the system. By comparing alluvial architecture between exposures, we evaluate similarities and differences from upstream to downstream and look for evidence of intrinsic and extrinsic controls on deposition. In both localities, the Lance Formation comprises multi-story sheet sandstones and smaller, single-story sandstones. Observed changes from upstream to downstream in the system include: 1) increasing paleoflow depths (from ~30-60 cm to ~70-120 cm); 2) decreasing preservation of fine-grained material within channel bodies; 3) increasing proportion of amalgamated, multi-story sand bodies; and 4) increasing lateral continuity of multi-story sand bodies. These results indicate that upstream, channel-body spacing is dominantly controlled by aggradational processes and may be the result of autogenic avulsion clustering, whereas downstream, evidence of incision and amalgamation indicate that base-level may have limited and controlled sand-body architecture.