John B. Shaw

Backwater and river plume controls on scour upstream of river mouths: Implications for fluvio‐deltaic morphodynamics

Sediment flux from rivers to oceans is the fundamental driver of fluvio-deltaic morphodynamics and continental margin sedimentation, yet sediment transport across the river-to-marine boundary is poorly understood. Coastal rivers typically are affected by backwater, a zone of spatially decelerating flow that is transitional between normal flow upstream and the offshore river plume. Flow deceleration in the backwater zone, as well as spreading of the offshore plume, should render rivers highly depositional near their mouths, leading to sedimentation and eventual elimination of the backwater zone at steady state. This reasoning is counter to observations of riverbed scour, erosional bed forms, and long-lived backwater zones near the mouths of some coastal rivers (e.g., Mississippi River, United States). To explain these observations, we present a quasi-2-D model of a coupled fluvial backwater and offshore river plume system and apply it to the Mississippi River. Results show that during high-discharge events the normal-flow depth can become larger than the water depth at the river mouth resulting in drawdown of the water surface, spatial acceleration of flow, and erosion of the riverbed. As proposed by Lane (1957), the transition to drawdown and erosion is ultimately forced by spreading of the offshore river plume. This points to the need to model coupled river and river plume systems with a dynamic backwater zone under a suite of discharges to accurately capture fluvio-deltaic morphodynamics and connectivity between fluvial sediment sources and marine depositional sinks.

The importance of erosion in distributary channel network growth, Wax Lake Delta, Louisiana, USA

A river delta’s shape and kinematics are dictated by the three-dimensional evolution of its distributary channels on the delta front, yet measurements of this evolution are scarce. We supply four bathymetric surveys documenting this evolution for Gadwall Pass—part of the Wax Lake Delta, one of the few rapidly prograding regions of the greater Mississippi Delta in coastal Louisiana, United States. This distributary channel extends 2–6 km beyond the sub-aerially emergent delta (dependent on water surface elevation) and bifurcates into four similarly sized distributary channels (average channel width = ∼150 m) in this sub-aqueous reach. Distributary channel growth proceeds primarily through erosion of the unchannelized foreset deposit, and growth patterns differ between high and low river flow. During high river flow, high upstream sand supply acts to aggrade the bed both inside and outside of the channel network. Erosion during high flow is focused at the sand shoals that define the sidewalls of the bifurcate channels, causing channel network rearrangement into a single primary channel with the remaining secondary channels branching off of it. During low river flow, bed erosion is focused at channel tips and the beds of all of the subaqueous distributary channels, leading to a bayward extension of each channel tip by ≥0.87 channel-widths. Channel-bottom erosion during low river flow is enhanced by tidally modulated currents that support sand suspension and transport in the subaqueous channels during ebb tide while receiving only a small sand supply from upstream.

Can relative paleointensities be determined from the normalized magnetization of the wind-blown loess of China?

Although the wind-blown loess and interbedded paleosol sequence in the Loess Plateau of China has been extensively studied and regarded as one of the most complete terrestrial records of both the geomagnetic field and climatic changes during the Quaternary, little attention has been paid to it as a source of relative paleointensity. In this study, we examine the Malan loess (L1), the last glacial sediments, in the Lingtai section to determine whether loess is able to reliably record the relative paleointensity. Toward that end both the conventional normalizing method as well as the pseudo-Thellier method [Tauxe et al., 1995] were used in conjunction with the examination of the rock magnetic properties and natural remanent magnetization (NRM). Rock magnetic properties of L1 show the uniformity in terms of magnetic mineralogy and grain size, suggesting that they may be suitable for relative paleointensity studies. Normalized remanences derived from L1 show highs between both 72-66 ka and 62-51 ka and lows near 63 ka, 42 ka and between 20 and 10 ka. Comparable results using different normalization parameters suggest that large-amplitude variation of normalized magnetization may reflect the intensity variation of the geomagnetic field. However, comparisons of the normalized magnetization with contemporaneous marine records show some dissimilarities. In particular, the intensity low at 20-10 ka, corresponding to the last glacial maximum, is not compatible with the Sint-800 composite record. On the other hand, spectral coherence analyses of middle- and high-frequency components of the normalized magnetizations suggest some climatic influence. Thus this apparent discrepancy may be explained by climatic changes in the Loess Plateau and the resulting effect on the NRM

Linking river-flood dynamics to hyperpycnal-plume deposits: Experiments, theory, and geological implications

Turbid river plumes entering ocean or lake water of lesser density (i.e., hyperpycnal plumes) can plunge to form turbidity currents providing an important link between terrestrial sediment sources and marine depositional sinks. A leading hypothesis suggests that hyperpycnal-plume deposits accurately record the rising and falling discharge of a flooding river (in terms of sediment-size grading, bedform sequence, and deposit thickness), which, if correct, has significant implications for unraveling river dynamics, reservoir potential, and Earth history from marine-event beds. Herein, we present one of the first experimental flume studies aimed at testing this hypothesis. Results indicate that depth-averaged hyperpycnal-plume velocities can be uncorrelated or even anti-correlated with river discharge at certain seabed locations because of translation of the plunge point resulting from temporal variations in discharge and sediment concentration through the duration of a river flood. An advection length scale of settling sediment is found to be an important control on hyperpycnal-plume deposits, where coarse sediment (sand) is most likely to record multiple flow accelerations and decelerations related to plunge-point translation even for a river flood with a single-peaked hydrograph. In contrast, fine sediment (mud) is relatively insensitive to local plunge-point dynamics and is most likely to preserve directly rising and falling river discharge. Finally, it was found that the necessary fluvial sediment concentration to form a plunging plume can be much larger than the concentration typically used assuming density equivalence because of deposition upstream of the plunge point.

Flow patterns and morphology of a prograding river delta

The transition of flow between laterally confined channels and the unchannelized delta front controls the morphodynamic evolution of river deltas but has rarely been measured at the field scale. We quantify flow patterns and bathymetry that define the evolution of the subaqueous delta front on the Wax Lake Delta, a rapidly prograding delta in coastal Louisiana. A significant portion of flow (∼59%) departs the channel network over lateral channel margins as opposed to the downstream channel tips. Bathymetric surveys and remotely sensed estimates of flow direction allow spatial changes in flow velocity to be quantified and patterns of erosion and deposition to be estimated. Shallowing along channel margins produces spatial acceleration and erosion. Lateral spreading, deceleration, and deposition occur within three to eight channel widths outside of the channel margins. In interdistributary bays, the shape of each flow path is constrained by “nourishment boundaries” that separate the outflows from neighboring channels. Deposit elevation decreases with a basinward slope of 2.4 × 10−4 with distance from a channel margin along any flow path, regardless of the channel or location that flow departed the network. Bathymetric depressions called “interdistributary troughs” form along nourishment boundaries where flow paths are the longest and deposit elevation is correspondingly low. We conclude that the deposit morphology exerts a strong control on bathymetric evolution and that interaction between neighboring channels and even neighboring deltas can influence delta front morphology.

Airborne radar imaging of subaqueous channel evolution in Wax Lake Delta, Louisiana, USA

Shallow coastal regions are among the fastest evolving landscapes but are notoriously difficult to measure with high spatiotemporal resolution. Using Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) data, we demonstrate that high signal‐to‐noise L band synthetic aperture radar (SAR) can reveal subaqueous channel networks at the distal ends of river deltas. Using 27 UAVSAR images collected between 2009 and 2015 from the Wax Lake Delta in coastal Louisiana, USA, we show that under normal tidal conditions, planform geometry of the distributary channel network is frequently resolved in the UAVSAR images, including ~700 m of seaward network extension over 5 years for one channel. UAVSAR also reveals regions of subaerial and subaqueous vegetation, streaklines of biogenic surfactants, and what appear to be small distributary channels aliased by the survey grid, all illustrating the value of fine resolution, low noise, L band SAR for mapping the nearshore subaqueous delta channel network.

Quantifying the stratigraphic completeness of delta shoreline trajectories

Understanding the incomplete nature of the stratigraphic record is fundamental for interpreting stratigraphic sequences. Methods for quantifying stratigraphic completeness for one-dimensional stratigraphic columns, defined as the proportion of time intervals of some length that contain stratigraphy, are commonplace; however, quantitative assessments of completeness in higher dimensions are lacking. Here we present a metric for defining stratigraphic completeness of two-dimensional shoreline trajectories using topset-foreset rollover positions in dip-parallel sections and describe the preservation potential of a shoreline trajectory derived from the geometry of the delta surface profile and the kinematics of the geomorphic shoreline trajectory. Two end-member forward models are required to fully constrain the preservation potential of the shoreline dependent on whether or not a topset is eroded during base level fall. A laboratory fan-delta was constructed under nonsteady boundary conditions, and one-dimensional stratigraphic column and two-dimensional shoreline completeness curves were calculated. Results are consistent with the hypothesis derived from conservation of sediment mass that completeness over all timescales should increase given increasing dimensions of analysis. Stratigraphic trajectories and completeness curves determined from forward models using experimental geomorphic trajectories compare well to values from transects when subsampled to the equivalent stratigraphic resolution as observed in the actual preserved sequence. The concept of stratigraphic completeness applied to two-dimensional trajectory analysis and the end-member forward models presented here provide novel tools for a conceptual understanding of the nature of stratigraphic preservation at basin scales.

Backwater number scaling of alluvial bed forms

The backwater number, Bw, compares the backwater length scale to the length scale of alluvial bed forms. We derive theory to show that Bw plays an important role in determining the behavior and scaling of morphodynamic systems. When Bw≫1, spatial patterns in deposition and erosion derive from flow accelerations associated with changes in flow depth, and bed evolution is akin to a kinematic wave. When Bw≪1, the spatial pattern of shear stress is determined by variations in energy slope, and alluvial beds experience topographic dispersion. This theory is confirmed using a numerical model and data compiled from the literature. We present a data set of Bw for bed forms ranging from dunes to river deltas, including field and experimental measurements. For field-scale measurements, we find that dunes have Bw>49, braid bars exist in the range Bw = [7.1, 17], meanders have a range Bw =[7.1, 18], and river mouth deposition ranges over Bw=[7.4, 29]. Further, alluvial morphologies that are easily recreated in the laboratory (dunes and avulsions) have overlapping field and laboratory Bw ranges. In contrast, alluvial forms that have traditionally been difficult to recreate (meanders and river mouth processes) have field Bw that are difficult to match in laboratory settings. Large experimental Froude numbers are shown to reduce experimental Bw and incite diffusional behavior. Finally, we demonstrate the utility of Bw scaling for estimating fundamental scales in sedimentary systems.

Measuring Subaqueous Progradation of the Wax Lake Delta with a Model of Flow Direction Divergence

This article presents and validates a technique for estimating the location of subaqueous channel tips based on the divergence of the flow field on the foreset of river deltas. The technique is referred to as the “Flow Divergence to Channel Tip” (FD2C) model. It builds on the previous work of Shaw et al(2016, JGR-ES) in which the authors first presented the method of mapping “streak lines” to estimate flow direction offshore of the Wax Lake Delta. Here the use of streak lines is used to provide input data (divergence field) for the FD2C model.

Congruent Bifurcation Angles in River Delta and Tributary Channel Networks

We show that distributary channels on river deltas exhibit a mean bifurcation angle that can be understood using theory developed in tributary channel networks. In certain cases, tributary network bifurcation geometries have been demonstrated to be controlled by diffusive groundwater flow feeding incipient bifurcations, producing a characteristic angle of 72∘. We measured 25 unique distributary bifurcations in an experimental delta and 197 bifurcations in 10 natural deltas, yielding a mean angle of 70.4∘±2.6∘ (95% confidence interval) for field-scale deltas and a mean angle of 68.3∘±8.7∘ for the experimental delta, consistent with this theoretical prediction. The bifurcation angle holds for small scales relative to channel width length scales. Furthermore, the experimental data show that the mean angle is 72∘ immediately after bifurcation initiation and remains relatively constant over significant time scales. Although distributary networks do not mirror tributary networks perfectly, the similar control and expression of bifurcation angles suggests that additional morphodynamic insight may be gained from further comparative study.