Understanding of the formative conditions of fluvial bedforms is significant for both river management and geological studies. Diagrams showing bedform stability conditions have been widely used for the analyses of sedimentary structures. However, the use of discriminants to determine the boundaries of different bedforms regimes has not yet been explored. In this study, we use discriminant functions to describe formative conditions for a range of fluvial bedforms in a 3‐D dimensionless parametric space. We do this by means of discriminant analysis using the Mahalanobis distance. We analyzed 3,793 available laboratory and field data and used these to produce new bedform phase diagrams. These diagrams employ three dimensionless parameters representing properties of flow hydraulics and sediment particles as their axes. The discriminant functions for bedform regimes proposed herein are quadratic functions of three dimensionless parameters and are expressed as curved surfaces in 3‐D space. These empirical functions can be used to estimate paleoflow velocities from sedimentary structures. As an example of the reconstruction of hydraulic conditions, we calculated the paleoflow velocity of the 2011 Tohoku‐Oki tsunami backwash flow from the sedimentary structures of the tsunami deposit. In so doing, we successfully reconstructed reasonable values of the paleoflow velocities.

Sedimentary bedforms such as ripples and dunes are generated both by river flows and sediment-laden gravity currents. Gravity current deposits are usually parameterized using existing bedform phase diagrams which are based on data from laboratory experiments and field observations of openchannel flows. Yet, it is not evident that open-channel flow bedform phase diagrams are applicable to gravity current deposits. Gravity current hydrodynamics are dependent on vertical density variation, that is density stratification, and therefore are fundamentally different from open-channel flows. New experiments to produce gravity current deposits are conducted and compared to existing open-channel flow data. It is shown that a parameter phase-space based on the lower layer of stratified gravity currents (i.e. that part below the velocity maximum) significantly improves the prediction of bedform type compared to bedform phase diagrams derived from layeraveraged parameters. These results confirm that bedforms produced by gravity currents can only be predicted accurately using the characteristics of the lower layer of stratified flow.

No abstract

We study the behavior of a confined stream made of two immiscible fluids when it reaches a T junction. Two flow patterns are witnessed: the stream is either directed in only one sidearm, yielding a preferential flow pathway for the dispersed phase, or splits between both. We show that the selection of these patterns is not triggered by the shape of the junction nor by capillary effects, but results from confinement. It can be anticipated in terms of the hydrodynamic properties of the flow. A simple model yielding universal behavior in terms of the relevant adimensional parameters of the problem is presented and discussed.

Sedimentary structures in ancient deposits are clues to reconstruct past geohazards. While parallel lamination formed by plane beds is one of the most common sedimentary structures in event deposits such as turbidites, the formative conditions for plane beds remain unclear. In the literature, two types of plane beds (upper and lower plane beds) exist and are supposed to develop under different shear stresses, particle sizes, and flow regimes. Here, we present new phase diagrams based on the compilation of existing data regarding formative hydraulic conditions for plane beds to clarify the formation processes associated with the two types of plane beds. The diagrams indicated that the data form two separate populations and the gap between them corresponds to the threshold condition of the particle entrainment into suspension. Lower plane beds form when sediment particles move only as bed load. This phase space can be discerned from fine sand to gravel and differs from the conventional view in which the formation of the lower plane bed is limited to grain sizes above 0.7 mm. In addition, our phase diagrams suggest that upper plane beds appear under conditions of the active suspended load. Our analyses demonstrate that the suspended load contributes to the formation of plane beds, whereas other mechanisms can also produce fine-grained plane beds in flows with low bed shear stress. Thus, the results of this study suggest that the existing interpretations on fine-grained parallel lamination such as Bouma’s Td division need to be reconsidered. The bedform phase diagrams newly established in this study will be useful for estimating the flow conditions from the geologic records of event beds.