Risk Analysis of Coastal Areas: An Ai-Based Perspective Using Sar Data

In this work, a Change Detection (CD) analysis was conducted based on the assessment of ground deformation (subsidence) in the Venice Lagoon over recent years. The multi-temporal interferometric Small Baseline Subset (SBAS) technique [1], [2] was employed to analyze the interconnections between the subsidence in the area and the occurrence of extreme flood events, with a specific focus on the floods that occurred in November 2019. Examining the time series of backscattered signals from the Sentinel-1 (S-1) synthetic aperture radar (SAR) sensor, we identified the extent of the flooded regions and evaluated the impact of the floods on the city. The potential of a newly developed Artificial Intelligence (AI) method [3] based on Random Forest [4], [5] was exploited. This methodology leverages the capability of several coherent/incoherent SAR change detection indices (CDIs) and their mutual interaction in a single corpus for rapid mapping of surface changes. This method [3] has shown great success in rapidly mapping land surface changes of areas in Sardinia and Sicily that were affected by large wildfires in the summer of 2021 and flooded areas in Houston and GalvestonBay as a result of Hurricane Harvey in 2017. In conclusion, the comprehensive CD/SBAS analysis provided valuable insights into the relationship between subsidence and recent extreme flood events in the Venice Lagoon, revealing the dynamics of the lagoon and its vulnerability to such events.