Assessment of beach erosion risk from seal level rise at island scale
The assessment of the beach erosion under long- and short-term sea level rise at the island level was carried out according to the following procedure: Geo-spatial information from all beaches of the 6 studied islands (e.g. beach dimensions and sediment type, occurrence of infrastructure/assets behind the beach, presence of river outlets etc) was collected/collated using remote-sensed imagery available in the web (e.g. Google Earth imagery) as well as available/archived imagery. Information on current wave forcing was provided through wave hindcasting an/or accessing/analysing available data bases (e.g. the ERA-INTERIM). Information on trends and projections for future mean and extreme sea levels was collated from the scientific literature (e.g. Tsimplis and Shaw, 2010; Hinkel et al., 2014; Vousdoukas et al., 2016) and existing data bases (e.g. the THALES-ISLA beach inventory). Finally, beach erosion and flooding was projected for all beached of the 6 islands, using suitable ensembles of analytical and numerical morphodynamic models. The procedure is summarised in the flow chart below.
Assessment of beach erosion and flooding trends due to long- and short term sea level rise using ensembles of cross-shore morphodynamic models
Seven cross-shore morphodynamic models were used to project beach response to sea level rise (SLR). The models were used in an ensemble mode, based on the concept that as models have differential sensitivity to the controlling factors, ensemble applications may provide ‘tighter’ prediction ranges than individual models. Two model ensembles were created, a ‘long-term’ ensemble consisting of the analytical Bruun, Dean and Edelman models and a ‘short-term’ ensemble comprising the numerical SBEACH, Leont’yev, XBEACH and Boussinesq models. The former is used to assess beach retreat/erosion under MSLR, whereas the latter retreat due to temporary SLR (i.e. episodic storm-induced). With regard to combined SLRs (i.e. storm-induced SLR superimposed on MSLRs), the long-term and short-term ensembles are used consecutively. The models were validated against physical experiments conducted at the GWK wave flume (Hanover, Germany) in 2013. For further details, see Monioudi et al. (2017).
Given the large scale of the application (island) and the lack of more detailed information, the models were set up, using a plausible range of environmental conditions (wave heights of 1 to 4 m and periods of 4 to 8 s, median grain sizes of 0.2 to 5 mm and under 12 SLR scenarios up to 2 m). As initial bathymetry linear profiles were considered with different profile slopes (bed slopes of 1/10 to 1/30 were examined). Totally 5500 numerical experiments were carried out. The above approach is designed to project beach retreat/erosion, but not temporary inundation/flooding due to wave run-up. Therefore, estimations of wave run up excursion/inundation were also undertaken on the basis of run up heights; these were estimated for all tested conditions, using the expressions of Stockdon et al. (2006) which have been validated for the beaches of the Aegean Archipelago (Vousdoukas et al., 2009).
Assessment of beach erosion and flooding trends due to long- and short term sea level rise using ensembles of cross-shore morphodynamic models
Detailed assessment of the erosion risk of the pilot beaches under sea level rise
In order to investigate in detail the morphodynamics of the pilot beaches, two different modelling approaches were used. First, the hydrodynamic regime was assessed through a 2-D wave model (model ALS, for full details, see Karambas et al., 2013) that can simulate coastal hydrodynamics along long coasts and down to deeper water depths of about 40 m with relatively low computational costs. The model is based on wave energy balance considerations in shallow waters (e.g. Battjes and Stive, 1985; Holthuijsen et al., 2003). Diffraction effects are considered, and the solution is based on an implicit backward finite difference scheme (Booij et al., 1999). In nearshore waters a module, based on hyperbolic type, mild-slope equations, is incorporated to deal with compound wave fields under shoaling, refraction, diffraction, reflection (total and partial) and breaking, and the 2-D continuity and momentum equations are used (depth- and short wave-averaged) to simulate wave-induced currents. Experiments were run using observed in the area offshore wind and swell wave conditions. Secondly, a 2-D Boussinesq model was employed to simulate the hydrodynamic and sediment dynamic processes of pilot beaches (Karambas et al., 2013). The model describes non-linear breaking and non-breaking wave propagation, wave-induced currents and sediment transport. Suspended sediment transport is simulated by solving the corresponding advection-diffusion equation in the surf and swash zones. Bed load and sheet flow transport are estimated using an improved formula of Camenen and Larson (2000). Finally a morphodynamic evolution subroutine is included.
The models were set up/forced using topographic-bathymetric, hydrodynamic and sedimentological data of high resolution/accuracy that were collected within the framework of the project. Consequently, errors in the morphodynamic simulations, caused by the use of secondary bathymetric/topographic data of low accuracy (e.g. bathymetric maps of the Hellenic Navy Hydrographic Service), were minimised. In the absence of primary directional wave information, wave forcing was hindcasted from locally available wind data series. Modeling results were verified/validated (where possible) by collected hydrodynamic (field) data sets.
