[Coastal flood hazard for Lecce, Italy, from breaches in the dunes] - Supplement

Journal of Water & Climate Change Supplementary Material for: Coastal flood hazard for Lecce, Italy, from breaches in the dunes Gianandrea Mannarinia, Mario Leonardo Salinasa, Giorgia Verria, Vladimir Santos Da Costaa, Riccardo Barzaghib and Daniela Carrionb a) CMCC Foundation - Euro-Mediterranean Center on Climate Change, Italy b) Dipartimento di Ingegneria Civile e Ambientale, Politecnico di Milano, Italy (August 2025) 1 GEODETIC ADJUSTMENT!................................................................................................................!2 2 COMPARISON OF LAND COVER DATA!.....................................................................................!4 3 FLOODING MAP ZOOMS!................................................................................................................!6 4 SECTORAL HAZARD MAPS!.............................................................................................................!7 4.1 Torre Rinalda ................................................................................................................................................. 7 4.2 Spiaggiabella .................................................................................................................................................. 9 4.3 Idume ........................................................................................................................................................... 10 4.4 Torre Chianca ................................................................................................................................................ 11 4.5 Zona Montegrappa ....................................................................................................................................... 12 4.6 Acquatina North ........................................................................................................................................... 13 4.7 Acquatina South ........................................................................................................................................... 14 4.8 Frigole .......................................................................................................................................................... 15 5 ADDITIONAL REFERENCES!..........................................................................................................!16 Supplementary Material 2 1 Geodetic adjustment First, the fundamental relationship between ellipsoidal height h, orthometric height H, and geoid undulation N is recalled [Heiskanen_1967, Wang_2021]: ℎ" = "𝐻" + "𝑁 (1) The EGM-96 geoid is provided on a 30' grid, while IT-05 is on a 2' grid. Thus, IT-05 can capture much higher frequency components of the geoid undulation. EGM-96 ellipsoid heights were brought to the reference ellipsoid and datum used to produce IT-05 by evaluating, on the same grid, the quantity: ∆𝑁 = 𝑁!"# − 𝑁$%&' (2) Then, a three-parameter (∆𝑥, ∆𝑦, ∆𝑧) functional [Fotopoulos_2013] ∆𝑁 (= " ∆𝑥" cos 𝜑cos 𝜆 +"∆𝑦" cos 𝜑sin 𝜆 +"∆𝑧" sin 𝜑 (3) was identified via a least-square fit on the ∆𝑁 data. The resulting ∆𝑁 4(= " ∆𝑁 4((∆𝑥, ∆𝑦, ∆𝑧) surface is shown in Supplementary Figure S1(a) and was used to define the undulation 𝑁!"#/$%&' of EGM-96 in the IT-05 system: 𝑁!"#/$%&'*= 𝑁!"# "−""∆𝑁 4( (4) This way it was possible to define ellipsoidal heights on the IT-05 system of coordinates. In analogy to Eq.1, following equalities hold: ℎ$%&' (,-.)=𝐷𝑇𝑀$%&' + 𝑁$%&'" (5a) ℎ$%&' (012)=𝐸𝑆𝐿!"# + 𝑁!"#/$%&' (5b) For a flooded location, the two left hand sides of Eq.5 are identical, leading to the relation: 𝐷𝑇𝑀$%&' =𝐸𝑆𝐿!"# +"∆ (6) With 𝛥 = 𝑁!"#|$%&' −"𝑁"$%&'. Thus, to compare to the orthometric height from the DTM, the Kirezci et al. (2010)’s values ESLegm must be increased by 𝛥. As seen from Supplementary Figure S1(a), along a transect on the coastal strip of Lecce the average value of 𝛥 is +2.7 cm. This constant offset was added to all ESL values provided by Kirezci et al. (2010). Nevertheless, it should be kept in mind that the 3 standard deviation on the IT-05 undulation is also in the order of 3 cm (Albertella et al., 2008) and even larger errors are estimated for EGM-96 (Lemoine et al., 1998). The so adjusted ESL values for Lecce with their lower and upper bounds are provided in the Supplementary material data. Supplementary Figure S1 (a) Transformation functional ΔNf from EGM96 to IT-05. The Lecce coastal strip is depicted in red; (b) Correction Δ of ESL values for passing from EGM96 to IT-05, along the red strip in a). Supplementary Material 4 2 Comparison of Land cover data We used data and class labels from the Corine land cover system (CLC) in its 2018's version, the latest available. We retained the original nomenclature of the land cover classes. Alternative land cover databases might lead to different outcomes. To assess it qualitatively, we compare CLC imagery to Sentinel-2 Global Land Cover project (S2GLC)'s one. A third option, here unexplored, would be to utilize Landsat data. It would enable the direct execution of a seamless change detection and classification algorithm on Google Earth Engine (Arevalo et al., 2020). While CLC has a lower spatial resolution (about 100 m, Büttner et al.,2017) compared to both S2GLC (tens of meters, Demirkan et al., 2020), it offers a standardized land cover dataset for Europe. Near Lecce’s coast, S2GLC identifies a greater extent of urbanized territory compared to CLC (cf. Supplementary Figure S2). While some of this discrepancy accounts for buildings not detected in CLC, a significant portion of the additional urban area classified by S2GLC is not corroborated as urban upon verification with orthophotos. We thus used the CLC dataset to calculate flooded areas by land use type. These areas were derived using qGIS built-in functions (DIFFERENCE, INTERSECTION) to compute overlaps between polygons. The resulting data are visualized in the column bars of Supplementary Figure S5(d)-S12(d) and stored numerically in the Flood_metrics_2060.xlsx database, part of the Supplementary material data. A potential limitation of this approach lies in the horizontal resolution of the CLC dataset (100 m), which is significantly coarser than that of the DTM (1 m) used for Lecce. This disparity may result in mismatches between the total flooded area estimated from the DTM and the aggregated values by land use type. We acknowledge this limitation and provide both sets of information in the database for full transparency. 5 Supplementary Figure S2. Comparison of Corine (CLC, red) and Sentinel-2 (S2GLC, green) land cover databases for the urban class, both for year 2018. (a) overview map with Places of Interest (POI, numbers next to black dots) for Lecce PUG, cf. Supplementary material data; (b) subregions with false positives for Sentinel-2; c) false positives for both Sentinel-2 and Corine. Supplementary Material 6 3 Flooding map zooms Supplementary Figure S3 Close-up of model flooding (depth according to the shades of blue) and future coastline (orange line) for a portion of the Lecce coast next the locations of main manuscript’s Figure 5(a) and Figure 5(b). The solid black line in the sea is the 0-meter elevation contour. Panel (c) provides the topography along the two transects shown in black in (a) and (b) where the ticks are spaced 50 meters. Supplementary Figure S4 As S3, but for the locations shown in main manuscript’s Figure 5(c) and Figure 5(d). A C B RCP 4.5 71 cm 2030 Elevation [m] 0m coast 0m flood viewcone 0 50m 100m 150m 200m 250m v. Portofino v. Porto Torres A C B RCP 4.5 67 cm 2020 Elevation [m] 0m coast 0m flood viewcone 0 50m 100m 150m 200m 250m 300m 350m 400m v. Torvajanica v. Ortona 7 4 Sectoral hazard maps 4.1 Torre Rinalda Supplementary Figure S5. Flooded areas in the "Torre Rinalda" sector during extreme sea level (ESL) events in selected years of the 21st century: 2020 (a), 2060 (b), and 2040 (c). The orange contour line outlines the flood extent, while flood depth is indicated in shades of blue. The background imagery includes an NDWI map (a) and the orthophoto A B C D NDWI 2023-12-20 Canals Supplementary Material 8 (b,c). Panel (d) shows a time series of flooded area by land cover type, using the same color scheme as in Figure 2 of the main manuscript. 9 4.2 Spiaggiabella Supplementary Figure S6. As S5, but for the "Spiaggiabella" sector. A B C D NDWI 2023-12-20 Fiumicelli lagoon Fiumicelli lagoon Rauccio Swamp Rauccio Swamp Supplementary Material 16 5 Additional References Albertella A., Barzaghi R., Carrion C., Maggi A. (2008) The joint use of gravity data and GPS/levelling undulations in geoid estimation procedures. Bollettino di Geodesia e Scienze Affini, n°1, 47-57 Büttner, G., Kosztra, B., Soukup, T., Sousa, A. and Langanke, T., 2017. CLC2018 technical guidelines. European Environment Agency: Copenhagen, Denmark, 25. Demirkan, D.Ç., Koz, A. and Düzgün, H.Ş., 2020. Hierarchical classification of Sentinel 2-a images for land use and land cover mapping and its use for the CORINE system. Journal of applied remote sensing, 14(2), pp.026524-026524. Fotopoulos, G. (2012) Combination of heights. In Geoid determination: Theory and methods (pp. 517-544). Berlin, Heidelberg: Springer Berlin Heidelberg ISBN: 978-3-540-74699-7 Heiskanen, W.A. and Moritz, H. (1967) Physical geodesy. W.H. Freeman and Company, San Francisco, ISBN: 9780716702337 Lemoine F.G., Kenyon S.C., Factor J.K., Trimmer R.G., Pavlis N.K., Chinn D.S., et al. (1998) The development of the Joint NASA GSFC and the National Imagery and Mapping Agency (NIMA) Geopotential Model EGM96. National Aeronautics and Space Administration, Goddard Space Flight Center, Greenbelt, Maryland 20771, NASA/TP-\1988-20681 Wang, Y.M., Sánchez, L., Ågren, J. et al. (2021). Colorado geoid computation experiment: overview and summary. J Geod 95, 127