Numerical predictability experiments of cross‐shore sandbar migration

Abstract

[1] Surf zone sandbars, common features along the world's sandy coastlines, continuously change their position in response to time-variable offshore wave conditions. Processbased predictions of cross-shore sandbar migration, relevant to the understanding of autonomous and artificially altered evolution of beaches, are intrinsically imprecise because of uncertainty in the model equations and, potentially, the sensitive dependence on the initial bathymetry. However, the magnitude of the resulting predictability limit and its dominant source are unknown. Here we show that cross-shore sandbar migration on the time scale of years is deterministically forced rather than deterministically chaotic, and that the unpredictability of sandbar migration results primarily from model inadequacy during major wave events. Because the unpredictability of sandbar migration is related to the stochastic nature of the forcing, the predictability limit is not a fixed value but depends on the timing of the wave event. We anticipate that detailed experiments to understand nearshore evolution from the underlying first principles will eventually pay off by extending the range of plausible simulated behavior and, consequently, will increase our ability to understand and predict how coasts may respond to changing climate.

Large-scale laboratory observations of beach morphodynamics and turbulence beneath shoaling and breaking waves

“…The sand-transport module, generally considered to be the weakest link, usually comprises one or more simple equations that semiempirically relate near-bed wave-driven (orbital) flow characteristics to a bulk transport quantity (e.g., Bailard, 1981;Ribberink, 1998). While such a sand-transport module can result in fairly accurate seasonal to multi-year bed-level change predictions in water depths larger than, say, 2 m (e.g., Ruessink et al, 2007;Ruggiero et al, 2009), this is at present essentially impossible in smaller depths (e.g., Ruessink and Kuriyama, 2008). In fact, most morphodynamic models do purposely not attempt to estimate sand transport rates here (e.g., Plant et al, 2004;Ruessink et al, 2007).…”

Large‐scale experimental observations of sheet flow on a sandbar under skewed‐asymmetric waves

“…In fact, models are generally calibrated via tunable parameters to maximize skill in predicting cross‐shore position of the sandbar crest and/or the cross‐shore profile shape [e.g., Hoefel and Elgar , ; Henderson et al ., ; Hsu et al ., ; Ruessink et al ., ]. High‐resolution sediment concentration and velocity profile measurements on a sandbar are needed to better understand the relevance and/or contribution of particular physical processes (that may currently be embedded in the tunable parameters) driving onshore/offshore sandbar migration [ Ruessink et al ., ; Ruessink and Kuriyama , ].…”

“…High-resolution sediment concentration and velocity profile measurements on a sandbar are needed to better understand the relevance and/or contribution of particular physical processes (that may currently be embedded in the tunable parameters) driving onshore/offshore sandbar migration [Ruessink et al, 2007;Ruessink and Kuriyama, 2008].…”

On cross‐shore migration and equilibrium states of nearshore sandbars

“…However, typical timescales of cross‐shore sandbar behavior are in the order of days to several years, and show characteristics that are not related to the timescales of wave climate variability [e.g., Wijnberg and Terwindt , 1995; Ruessink et al , 2003]. Models that describe cross‐shore sandbar behavior as the result of wave and sediment‐transport processes on timescales of seconds to hours are prone to exponential error accumulation due to the nonlinear nature of those processes [see, e.g., Ruessink and Kuriyama , 2008; Pape et al , 2009]. Small errors in the model processes, states and inputs can cause a rapidly increasing divergence between observed and modeled sandbar behavior.…”