**4. Results**

Figure 8 shows the 100-year joint exceedance isocontours for all the points analyzed around Corsica. Different extreme regimes are here outlined: the dependence between the extreme triplets on the western part appear to be less marked than for the eastern part (as indicated by the convexity and range covered by the isocontours). This means that the 100-year triplets on the western coast deviate less from the 100-year return levels estimated using the marginal (i.e., without accounting for the dependence). This is clearly not the case on the eastern part as illustrated for instance by the Bastia analysis (black box) which shows large convexity of the isocontours. This 100-year joint exceedance isocontours' spatialization also highlights higher values of Hs (>6 m) and U (>18 m/s) offshore on the western part of the island compared to the eastern part of Corsica. This is in agreemen<sup>t</sup> with the different wind regimes (and induced waves) presented in Figure 1b. Indeed, on the west side of the island, wind and wave features (Hs > 6 m, U > 18 m/s, Tp > 12 s, 500 km < fetch < 1300 km) sugges<sup>t</sup> that extreme scenarios may be driven by swells, whereas offshore conditions on the east part of Corsica mostly by wind waves

**Figure 8.** The 100-year joint exceedance isocontours for all of the boxes (in color) analyzed around Corsica, the colorbar corresponds to SWL (m).

The SWAN and SWASH-2DH simulations (steps 6–7) provide the contributions to compute total water levels (TWL) (step 8) at the shoreline for each offshore condition combination tested (see Section 2.2 for details on the methods). Figure 9 illustrates the distribution of TWL along the shoreline (a) for current conditions, and (b) for future conditions (2100). For current conditions, the static TWL along the shoreline in the study area is between 0.80 and 1.8 m. For "2100" future conditions, the values are between 1.2 and 2.2 m. Clearly, outside the area on the west of the Cap Corse in the north of the island, this value is hardly reached even under future conditions. Moreover, we note that values of TWL are higher on the west coast rather than the east coast, especially in the north west and south west, for both current and future conditions.

Finally, we tested the sensitivity of the results to the choices of the offshore conditions' combinations: this is generally below 0.05 m and can be considered not significant. Only a few combinations show differences greater than 0.05 m. In particular, Figure 10a shows that combinations 19 and 27 lead to the highest total water levels at the shoreline in the Gulf of Porto considering the specific directions Dp = 270◦ and Du = 240◦. Moreover, the directions Dp and Du of waves and winds influence the scenarios leading to maximum water level at the shoreline. Indeed, when applying other directions (even with small changes) Dp = 240◦ and Du = 240◦ to force the SWAN and SWASH-2DH models, the scenario leading to the

highest total water levels is different: combination 17 (Figure 10b). The differences are particularly marked in the southern Gulf of Porto where total levels at the shoreline are higher when applying Dp = 270◦ and Du = 240◦ (Figure 10c) instead of Dp = 240◦ and Du = 240◦ (Figure 10d).

**Figure 9.** Total water level at the shoreline obtained: (**a**) for current conditions; (**b**) for future conditions (2100).

**Figure 10.** Scenarios leading to maximal water level at the shoreline in Gulf of Porto and Gulf of Girolata: (**a**) for directions Dp = 270◦ and Du = 240◦; (**b**) for directions Dp = 240◦ and Du = 240◦ in current conditions. Total water level at the shoreline obtained in Gulf of Porto and Gulf of Girolata: (**c**) for directions Dp = 270◦ and Du = 240◦; (**d**) for directions Dp = 240◦ and Du = 240◦ in current conditions.
