Search

Bayesian networks (BNs) are increasingly being used to model complex coastal processes due to their ability to integrate non-linear systems, their transparent probabilistic framework, and low computational cost. A BN may be suited to descriptive or predictive application. Descriptive BNs are highly calibrated models that are useful for better understanding the physics and causal relationships driving a system. Predictive BNs are generalisations of a system that have skill at predicting outside of the training domain. The predictive and descriptive usefulness of a BN depends on its complexity and the amount of data available to train it, but there is often a trade-off; higher descriptive skill comes at the cost of reduced predictive skill. To demonstrate the differences between predictive and descriptive BNs in a coastal engineering context, a BN to predict shoreline recession caused by coastal storm events is developed and tested using an extensive 10-year dataset incorporating 137 individual storms events monitored at Narrabeen-Collaroy Beach, Australia. A parsimonious approach to BN development is used to separately determine the optimum predictive and descriptive BNs for this dataset. Results show that for this dataset two quite different BNs can be developed: one that is optimized to achieve the highest predictive skill, and a second network that is optimized to maximize descriptive skill. The optimum predictive BN is found to comprise 3 nodes (variables) and can predict the shoreline recession caused by unseen storm events with a skill of 65%. The optimum descriptive BN is composed of 5 nodes and can reproduce 88% of the training dataset, but with more limited predictive capabilities. The uses and limitations of these two different approaches to BN formulation are illustrated with example applications to coastal process modelling. It is anticipated that the insights provided in this paper will help to clarify the further development of Bayesian Networks applied to coastal modelling. [ABSTRACT FROM AUTHOR]

This paper assesses the accuracy of 11 existing runup models against field data collected under moderate wave conditions from 11 non-truncated beaches in New South Wales and Queensland, Australia. Beach types spanned the full range of intermediate beach types from low tide terrace to longshore bar and trough. Model predictions for both the 2% runup exceedance ( R 2% ) and maximum runup ( R max ) were highly variable between models, with predictions shown to vary by a factor of 1.5 for the same incident wave conditions. No single model provided the best predictions on all beaches in the dataset. Overall, model root mean square errors are of the order of 25% of the R 2% value. Models for R 2 % derived from field data were shown to be more accurate for predicting runup in the field than those developed from laboratory data, which overestimate the field data significantly. The most accurate existing models for predicting R 2% were those developed by Holman [12] and Vousdoukas et al. [40], with mean RMSE errors of 0.30 m or 25%. A new model-of-models for R 2% was developed from a best fit to the predictions from six existing field and one large scale laboratory R 2% data-derived models. It uses the Hunt [17] scaling parameter tan β H o L o and incorporates a setup parameterisation. This model is shown to be as accurate as the Holman and Vousdoukas et al. models across all tidal stages. It also yielded the smallest maximum error across the dataset. The most accurate predictions for R max were given by Hunt [17] but this tended to under predict the observed maximum runup obtained for 15-min records. Mase's [22] model has larger errors but yielded more conservative estimates. Greater observed values of R max are expected with increased record length, leading to greater differences in predicted values. Given the large variation in predictions across all models, however, it is clear that predictions by uncalibrated runup models on a given beach may be prone to significant error and this should be considered when using such models for coastal management purposes. It should be noted that in extreme events, which are lacking in the dataset, runup may be truncated by beach scarps, cliffs, and dunes, or may overtop, and as a result, the probability density functions will have different tail shapes. The uncertainty already present in current models is likely to increase in such conditions. [ABSTRACT FROM AUTHOR]

Traditional methods for assessing coastal hazards have not typically incorporated a rigorous treatment of uncertainty. Such treatment is necessary to enable risk assessments which are now required by emerging risk based coastal zone management/planning frameworks. While unresolved issues remain, relating to the availability of sufficient data for comprehensive uncertainty assessments, this will hopefully improve in coming decades. Here, we present a modelling framework which integrates geological, engineering and economic approaches for assessing the climate change driven economic risk to coastal developments. The framework incorporates means for combining results from models that focus on the decadal to century time scales at which coasts evolve, and those that focus on the short term and seasonal time scales (storm bite and recovery). This paper demonstrates the functionality of the framework in deriving probabilistic coastal hazard lines and their subsequent use to establish an economically optimal setback line for development at a case study site; the Narrabeen–Collaroy embayment in Sydney, New South Wales. [ABSTRACT FROM AUTHOR]

Multi-purpose Artificial Reefs (MPARs) are structures that may provide aesthetically acceptable coastal protection and improve recreational outcomes. Twenty years after construction of the first MPAR, Narrowneck Reef on the Gold Coast of Australia, most of the available literature is still focused on the planning, design and construction of such structures and peer-reviewed publications on their post-construction monitoring, interaction with sediment transport and impacts on coastal morphology are lacking. The aim of this paper is to evaluate how does Narrowneck Reef influence the sediment transport, and morphological changes around the anthropogenic structure, two decades after construction. To do so, a combination of ten high spatial resolution topo-bathymetric surveys from the top of the dune to the 10 m depth captured over 21 months and a series of 60 simulations using a calibrated numerical model were used. Our results demonstrate that: although not expected during design or reported in similar structures, sand can bypass the MPAR around its offshore end; under oblique waves, the longshore currents are deflected as they pass the reef, resulting in a shadow zone on the downdrift side where sand deposits; the bar crest tends to be higher on the reef's updrift side compared to downdrift, indicating that the MPAR can act as a store for sediments, as initially designed. Furthermore, the MPAR can act to stabilise the bar as it moves onshore with a downdrift offset of the inner bar as a result of low oblique wave incidence. The results presented here demonstrate that the short-term response to the MPAR twenty years after construction is more closely related to the updrift build up of sand and the deflection of longshore currents as they encounter the reef than to the dissipation of wave energy. This is because MPARs are designed to dissipate just enough wave energy so that they can achieve their recreational goal (surfing). • Beach morphodynamic state and tide level influence sediment transport pathways. • Sand bypassing occurs mainly inshore of the structure. • Under large oblique waves, sand bypass can occur offshore of the structure. • Deflection of longshore currents influence morphological response to the reef. • Bar crest is usually higher on the reef's updrift side compared to downdrift. [ABSTRACT FROM AUTHOR]

The rotation of beaches is important and should be more pronounced under high energy conditions with significant changes in wave direction. Shoreline rotation occurs when one end of a beach accretes, while the opposite retreats. This phenomenon is attributed to periodic or long-term changes in wave climate. To simulate the variability of the shoreline orientation in a simple and efficient way, the aim of this paper is to present a new equilibrium-based shoreline rotation model derived from a shoreline migration formulation. This model assumes that beach rotation is induced by the incoming wave power and direction. The proposed model was applied to two embayed beaches with different hydrodynamic and morphodynamic contexts located in the South Pacific Ocean. Almost fifteen years of shoreline positions extracted from hourly time-averaged images captured at Tairua Beach, New Zealand, and thirty-seven years of shoreline position data acquired from beach profile surveys at Narrabeen Beach, Australia, were used to evaluate the model performance. The results from the last beach have been compared and discussed with respect to an existing semi-empirical model, which has shown to successfully reproduce the time-varying shoreline rotation on embayed beaches. Overall, the results demonstrate that the proposed model successfully represents the general shoreline orientation variability both qualitatively and quantitatively. Nevertheless, observations from different sites with diverse beach characteristics and wave conditions are needed to broadly validate the proposed model. • A new equilibrium-based shoreline rotation model is presented. • An analytical Equilibrium Wave Direction Function (EWDF) is obtained. • The model successfully represents the shoreline orientation variability at two embayed beaches of the South Pacific Ocean. [ABSTRACT FROM AUTHOR]