Regionalization of Parameters of Dynamic Shoreline Evolution Model: A case study from Louisiana

Abstract:

Beach shorelines continuously evolve in response to both natural and anthropogenic factors. Prediction of the coastline evolution is vital for the management of coastal development but is challenging because of the complexity of coastal processes and limitations in coastal erosion measurement. In the last two decades, Louisiana has put significant effort into restoring barrier islands and beaches to defend against hurricanes and rising sea. Data-intensive numerical models are widely used to predict the evolution of beach shorelines and subsequently evaluate the coastal development projects, i.e., beach nourishment. Few recent studies have introduced parsimonious models such as a simple Shoreline Dynamic Simulation Model (SDSM) (Cutler et al., 2019, JCR) to provide a useful low-cost tool for predicting the evolution of the coastline. However, because of the oversimplification in the representation of various processes, the parameters of the SDSM model are conceptual and requires a model calibration technique. Despite being simple, the model, if calibrated robustly, can reproduce the results from more sophisticated data-intensive numerical models.

The objective of this study is to enable the application of the SDSM model to data-poor locations for a quick evaluation of beach nourishment project. In this study, we develop a model parameter regionalization scheme for the SDSM model. First, we conducted a global sensitivity analysis to get insight into the sensitivity of each of the parameters of the SDSM model. After identifying the most influential parameters, we estimated model parameters by calibrating the model against data from the 22 beach nourishment projects from the state of Louisiana. Then we developed and tested a regionalization scheme derived from a set of calibrated model parameters and relevant beach shoreline attributes.