Hydrodynamic Model Sensitivity under Climate Change for Sustainable Aquaculture
Epistemic Uncertainty Analysis of Open Boundary Conditions in Bantry Bay under SSP5-8.5
N. Litzerinou (TU Delft - Civil Engineering & Geosciences)
José A. Á. Antolínez – Mentor (TU Delft - Civil Engineering & Geosciences)
Patricia Mares Nasarre – Graduation committee member (TU Delft - Civil Engineering & Geosciences)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
Coastal aquaculture in data-scarce regions is increasingly exposed to climate change, yet careful calibration of hydrodynamic models in these areas is often limited by a lack of in-situ observations. This thesis examines how the Bantry Bay estuary in Ireland, a stratified coastal system supporting significant kelp and mussel cultivation, responds to 2075 climate projections under the high emission SSP5-8.5 scenario. Given the scarcity of historical tide gauge data, the model calibration was shifted from a conventional deterministic to a probabilistic modelling approach, using a three-dimensional (3D) hydrodynamic model (Delft3D-FM) to test how the bay’s interior environment reacts to plausible variations in open-boundary conditions (temperature, salinity, and sea level).
Physically bounded distributions for temperature and salinity change were built from the pooled projections of four CMIP6 models, and for sea-level rise from the NASA IPCC AR6 percentile bands. A candidate pool of 5,000 parameter combinations was generated by inverse-transform sampling from these
distributions, from which the Maximum Dissimilarity Algorithm selected a final, space-filling ensemble of 50 members. Each member was applied to the 2022 baseline simulation through the delta-change method, superimposing the sampled temperature, salinity, and sea-level offsets onto the observed
boundary forcing to represent plausible 2075 conditions.
The findings show that the bay is boundary-controlled: mean salinity, mean temperature, and mean water level are each governed almost exclusively by their own boundary offset, with no response meaningfully shaped by more than one driver. Salinity uncertainty is small and stable throughout the year,
while temperature uncertainty is seasonal, concentrated in the summer tail, where the ensemble median already exceeds the kelp thermal threshold. This extreme regime tells a more complex story than the mean: thermal stress at cultivation sites is not driven by warming alone, but by the interaction of
local freshening and sea-level rise, with the highest-risk scenarios arising from moderate to low warming compounded by strong freshening rather than warming in isolation. Spatially, the boundary-driven fields are close to uniform across the bay, yet thermal stress itself concentrates in the inner bay. Sea
level rise itself functions primarily as a mean-state shift, with no measurable change to the bay’s tidal range.
This work demonstrates that in data-limited coastal systems, boundary uncertainty, rather than internal model complexity, is what dominates the predictive picture, and that resolving it requires more than a simple temperature projection. Climate-risk assessments for surface aquaculture must account for the complex interplay of multiple physical drivers, as the conditions most likely to cause biological stress arise from their interaction rather than from any single driver in isolation. This probabilistic workflow is designed to be directly transferable to other demonstration sites within the PHAROS project network
and beyond.