Quantifying Uncertainties in Numerical and Financial Modelling for Gravel and Composite Sand-Gravel Coasts

Phillips, Benjamin ORCID: 0000-0001-7210-2880 (2021) Quantifying Uncertainties in Numerical and Financial Modelling for Gravel and Composite Sand-Gravel Coasts. PhD thesis, University of Liverpool.

Text 200839825_May2021.pdf - Unspecified Access to this file is embargoed until 1 August 2025. Download (9MB)

Abstract

The coastal systems of focus in this thesis are gravel and composite sand-gravel coasts. The study sites used in this thesis are in the United Kingdom, where gravel coasts form 1000 km of coastline and act as crucial, naturally evolving coastal defences for many coastal communities. These systems face increasing risks of flooding and erosion driven by sea-level rise. Consequentially, transitions in shoreline management plans from `hold the line' to `no active intervention' are now planned in response to these amplified risks. This thesis uses numerical models to explore uncertainties in the the long-term and event-driven morphological evolution of gravel coasts, and hinterland inundation resulting from overwash of gravel coasts. COastal Vector Evolution Model (COVE), a `one-line' model which treats the coastline as a single contour line, is first developed for applications to gravel coasts. The model development is motivated by a need to project future morphological evolution of these crucial, naturally evolving defences to inform nourishment strategies and shoreline management plans. COVE was validated against alongshore topographic surveys, but under-predicts evolution when applied with a transport coefficient capable of maintaining model stability. The model was then used to compare coastal evolution under wave climates characterised with different offshore wave directions, heights and periods with that using a baseline time series. Predicted evolution shows similar agreement whether driven by a time-series of data or a wave climate that is sampled statistically. Uncertainty in the net present value of future flood losses from wave-water level conditions that have equal probability but with variable severity and financial consequences is then explored. The case study used is Fairbourne, Gwynedd, an example of where the shoreline management is stipulated to transition to no active intervention in response to 21st Century sea-level rise. Understanding mechanisms of uncertainty such as sea-level rise, discount rate and overwash are critical to ensure that coastal adaptation decisions are not taken too early or too late. Expected annual damage (financial risk) and the net present value of flood damage throughout the 21st Century are calculated using XBeach-X and LISFLOOD-FP to resolve overwash and hinterland inundation, respectively and deterministic and probabilistic methods for representing overwash and sea-level rise. Differences of orders of magnitudes in the net present value of future flood damage were identified between 25th, 50th and 75th percentiles of overwash severity. A probabilistic method for representing uncertainty in overwash is proposed and found to yield flood damages which are three times higher when compared with a deterministic approach using average overwash conditions. This highlights the importance of representing overwash as a mechanism of uncertainty in flood damage valuation and demonstrates the risk of under-predicting flood damage under average conditions represented deterministically. Intertidal foreshore evolution has been largely neglected in previous applications of XBeach-X to gravel beaches. The morphodynamic feedback between the sandy, intertidal foreshore and supratidal gravel barrier components of a composite sand-gravel beach was investigated by varying the frequency at which the sandy, intertidal foreshore evolution is updated during the gravel simulation. Constraining intertidal foreshore evolution was shown to amplify wave runup and erosion of the gravel barrier. The simulation in which the foreshore was not morphologically updated led to a 30-fold increase in hazardous freeboard values, and an order of magnitude increase in barrier erosion when compared to the scenarios where the foreshore was updated whilst the water level exceeded the toe of the barrier. Representing intertidal foreshore evolution when applying XBeach-X to composite sand-gravel beaches is shown to be important to avoid over-prediction or under-prediction of wave runup and resulting gravel barrier erosion. The most important findings from the thesis are two-fold. First, that neglecting intertidal foreshore evolution within XBeach-X can amplify erosion of the crest of a composite sand-gravel beach. This may have implications for the development of early warning systems for locations where the beach protects critical infrastructure, and implications for cost-benefit analysis if the beach is artifically maintained. Secondly, that overwash is a critical source of uncertainty in the financial valuation of flood damages. Overwash should therefore be considered probabilistically to ensure optimal investment in coastal adaptation and mitigation measures in response to sea-level rise.

Item Type:	Thesis (PhD)
Divisions:	Faculty of Science and Engineering > School of Environmental Sciences
Depositing User:	Symplectic Admin
Date Deposited:	09 Sep 2021 13:29
Last Modified:	18 Jan 2023 22:37
DOI:	10.17638/03124167
Supervisors:	Plater, Andrew ORCID: 0000-0001-7043-227X Brown, Jennifer
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3124167