A novel large wave flume experiment was conducted on a fixed, barred beach with a sediment pit on the sandbar, allowing for the isolation of small‐scale bed response to large‐scale forcing. Concurrent measurements of instantaneous sheet layer sediment concentration profiles and near‐bed velocity profiles were obtained on a sandbar for the first time. Two sediment distributions were used with median grain diameters, d50, of 0.17 and 0.27 mm. Sheet flow occurred primarily under wave crests, where sheet thickness increased with increasing wave height. A proportionality constant, normalΛ, was used to relate maximum Shields parameter to maximum sheet thickness (normalized by d50), with bed shear stress computed using the quadratic drag law. An enhanced sheet layer thickness was apparent for the smaller sediment experiments ( normalΛ = 18.7), when directly compared to closed‐conduit oscillatory flow tunnel data ( normalΛ = 10.6). However, normalΛ varied significantly (5 <  normalΛ < 31) depending on the procedure used to estimate grain roughness, ks, and wave friction factor, fw. Three models for ks were compared (keeping the model for fw fixed): constant ks = 2.5 d50, and two expressions dependent on flow intensity, derived from steady and oscillatory sheet flow experiments. Values of ks/d50 varied by two orders of magnitude and exhibited an inverse relationship with normalΛ, where normalΛ ∼ 30 for ks/d50 of O(1) while normalΛ ∼ 5 for ks/d50 of O(100). Two expressions for fw were also tested (with the steady flow‐based model for ks), yielding a difference of 69% ( normalΛ ∼ 13 versus normalΛ ∼ 22).

Horizontal and vertical pressure gradients may be important physical mechanisms contributing to onshore sediment transport beneath steep, near‐breaking waves in the surf zone. A barred beach was constructed in a large‐scale laboratory wave flume with a fixed profile containing a mobile sediment layer on the crest of the sandbar. Horizontal and vertical pore pressure gradients were obtained by finite differences of measurements from an array of pressure transducers buried within the upper several centimeters of the bed. Colocated observations of erosion depth were made during asymmetric wave trials with wave heights between 0.10 and 0.98 m, consistently resulting in onshore sheet flow sediment transport. The pore pressure gradient vector within the bed exhibited temporal rotations during each wave cycle, directed predominantly upward under the trough and then rapidly rotating onshore and downward as the wavefront passed. The magnitude of the pore pressure gradient during each phase of rotation was correlated with local wave steepness and relative depth. Momentary bed failures as deep as 20 grain diameters were coincident with sharp increases in the onshore‐directed pore pressure gradients, but occurred at horizontal pressure gradients less than theoretical critical values for initiation of the motion for compact beds. An expression combining the effects of both horizontal and vertical pore pressure gradients with bed shear stress and soil stability is used to determine that failure of the bed is initiated at nonnegligible values of both forces.

A large‐scale laboratory experiment was conducted to evaluate cross‐shore sediment transport and bed response on a sandbar under erosive and accretive field‐scale wave conditions (total of 11 cases). Unprecedented vertical resolution of sediment concentration was achieved through the use of conductivity concentration profilers alongside miniature fiber optic backscatter profilers. Observations were made of intrawave (phase‐averaged) and wave‐averaged cross‐shore sediment flux profiles and transport rates in the lower half of the water column on the crest of a sandbar. The net sediment transport rate was partitioned into suspended sediment (SS) and bed load (BL) components to quantify the relative contributions of SS and BL to the total sediment transport rate. Net SS transport rates were greater than net BL transport rates for the positive (wave crest) half‐cycle in 6 of 11 cases, compared to 100% (11 of 11) for the negative (wave trough) half‐cycle. Net (wave‐averaged) BL transport rates were greater, in magnitude, than net SS transport rates for 7 of the 11 cases. The dominant mode of transport was determined from the ratio of net BL to net SS transport rate magnitudes. The net transport rate was negative (offshore‐directed) when SS dominated and positive (onshore‐directed) when BL dominated. Net BL transport rate correlated well with third moments of free‐stream velocity (r2 = 0.72), suggesting that energetics‐type quasi‐steady formulae may be suitable for predicting BL transport under the range of test conditions.