Blind testing of shoreline evolution models

Abstract

Beaches around the world continuously adjust to daily and seasonal changes in wave and tide conditions, which are themselves changing over longer timescales. Different approaches to predict multi-year shoreline evolution have been implemented; however, robust and reliable predictions of shoreline evolution are still problematic even in short-term scenarios (shorter than decadal). Here we show results of a modelling competition, where 19 numerical models (a mix of established shoreline models and machine learning techniques) were tested using data collected for tairua beach, new Zealand with 18 years of daily averaged alongshore shoreline position and beach rotation (orientation) data obtained from a camera system. in general, traditional shoreline models and machine learning techniques were able to reproduce shoreline changes during the calibration period (1999-2014) for normal conditions but some of the model struggled to predict extreme and fast oscillations. During the forecast period (unseen data, 2014-2017), both approaches showed a decrease in models' capability to predict the shoreline position. this was more evident for some of the machine learning algorithms. A model ensemble performed better than individual models and enables assessment of uncertainties in model architecture. Research-coordinated approaches (e.g., modelling competitions) can fuel advances in predictive capabilities and provide a forum for the discussion about the advantages/disadvantages of available models. Quantitative prediction of beach erosion and recovery is essential to planning resilient coastal communities with robust strategies to adapt to erosion hazards. Over the last decades, research efforts to understand and predict shoreline evolution have intensified as coastal erosion is likely to be exacerbated by climatic changes 1-5. The social and economic burden of changes in shoreline position are vast, which has inspired development of a growing variety of models based on different approaches and techniques; yet current models can fail (e.g. predicting erosion in accreting conditions). The challenge for shoreline models is, therefore, to provide reliable, robust and realistic predictions of change, with a reasonable computational cost, applicability to a broad variety of systems, and some quantifiable assessment of the uncertainties.

16 years of topographic surveys of rip-channelled high-energy meso-macrotidal sandy beach

“…This monitoring program is also entering a new era as, since 2016, 4 km of coastal dunes are now surveyed at least quarterly 54,55 , together with notches dug in the dune to reinstate dune mobility and address plant community restoration along with biological monitoring 37 . With the growing interest in long-term beach datasets 23,30 and growing need of validation data for a wealth of coastal models 56 , it is timely to facilitate the unrestricted use of this unique high-energy meso-macrotidal beach dataset, including the daily surveys during the ECORS'08 field campaign. The archived dataset includes all the raw survey points and suitably interpolated digital elevation models (DEMs), and time series of astronomical tide and offshore wave forcing.…”

Impact of model free parameters and sea‐level rise uncertainties on 20‐years shoreline hindcast: the case of Truc Vert beach (SW France)

“…This generation of RCMs has been applied successfully at many sites where longshore processes have little influence on shoreline variability (e.g. Castelle et al, 2014; Dodet et al, 2019; Ludka et al, 2015; Lemos et al, 2018; Montaño et al (2020); Robinet et al, 2020; Splinter et al, 2014). However, these RCMs rely heavily on a data‐driven approach to search for the best free parameters.…”

Efficient Modeling of Complex Sandy Coastal Evolution at Monthly to Century Time Scales

“…Ensemble forecasts can then be made with the relatively swift ShorelineS model to assess a number of realizations of the future coastline position accounting for variations in initial and boundary conditions as well as uncertainty in model parameters. The potential for such approaches was recently shown by Montaño et al (2020) in a unique comparison of a number of mostly equilibrium-type cross-shore models and data-driven models against a 15-year calibration dataset followed by a blind 3-year 'shorecast.' The relatively straightforward model input of the ShorelineS model provides a unique opportunity to automatically process large data sets of remotely sensed coastal information, as the precise definition of the coastal reference lines does not affect the forecasts.…”