Flood risk assessment of cultural heritage sites near lakes via advanced hydrodynamic modeling and digital technologies

Cultural heritage flood risk assessments demand high-resolution, physics-based modeling frameworks capable of capturing complex terrain in diverse hydrological environments, and the subtle dynamics that threaten the relics. This study develops and applies such a framework at the lakeside archaeological site of Smuszewo, Poland, where flood hazards could arise from a combination of overland runoff and dynamic lake-stage fluctuations. To this end, we applied a high-resolution modeling approach combining overland terrain obtained via LidAR and drone photogrammetry, and a digitized lake bathymetric model, within a HEC-RAS 2D Rain-on-Grid framework that enables detailed simulation of runoff–lake interactions and site-specific flood scenarios across design storms. We evaluate three scenarios: (i) model calibration using a five-day rainfall-stage event, (ii) the hydraulic impact of including versus omitting explicit lake bathymetry, across five design storms (1–50-year return periods); and (iii) a margin-to-failure analysis simulating lake-level rise from 0.00 to +1.50 m. Results show that omitting bathymetry underestimates peak flows by up to 55 × for frequent storms due to artificial ponding on flat, dry-initiated surfaces (vs. realistic depths and wet cells enabling accurate volume propagation), whereas impact diminishes for the higher return periods. However, even under the worst-case scenario—an extreme storm on top of a +1.5 m lake-stage rise—more than 1 m of freeboard remains. Our findings demonstrate the critical role of multiscale, high-resolution terrain and physics-driven methods—and lay the groundwork for future digital-twin implementations, predictive maintenance strategies, and cloud-based simulations in cultural heritage flood-risk management.