A Time-Series Analysis of Rockfall Evolution in a Coastal Region Using Remote Sensing Data

The evolution of technology, particularly the integration of Unmanned Aerial Systems (UAS), earth observation datasets, and historical data such as aerial photographs, stand as fundamental tools for comprehending and reconstructing surface evolution and potential environmental changes. In addition, the active geodynamic phenomena in conjunction with climate crisis and the increasing frequency of extreme weather phenomena can cause abrupt events such as rockfalls and landslides, altering completely the morphology on both small and large scales.

This study deals generally with the temporal evolution of landscapes and specifically focuses on the detection and quantification of a significant rockfall event that occurred at Kalamaki Beach on Zakynthos Island, Greece – a very popular summer destination. Utilizing UAS surveys conducted in July 2020 and July 2023, this research revealed a rockfall that has significantly altered the coastal morphology. During this period, two severe natural phenomena occurred, one of which could potentially be the cause of this rockfall event. Initially, the Mediterranean hurricane (‘medicane’) ‘Ianos’ made landfall in September 2020, affecting a large part of the country including the Ionian Islands. The result was severe damage to property and infrastructures, along with human casualties, induced by intense precipitation, flash flooding, strong winds, and wave action. Second, in September of 2022, an M_L=5.4 earthquake struck between Cephalonia and Zakynthos Islands in the Ionian Sea, triggering considerable impact in both islands. The study employs satellite images postdating these natural disasters, to detect the source of the rockfall in Kalamaki Beach. Additionally, historical analog aerial images from 1996 and 2010 were used as assets for understanding the surface’s evolution. For the quantitative analysis, we applied 3D semi-automated change detection techniques such as the M3C2 algorithm, to estimate the volume of the rockfall.

The results provide insights into the complex interplay between natural disasters and geological processes, shedding light on the dynamic nature of landscapes and the potential implications for visitor-preferred areas.

This research not only contributes to our understanding of landscape evolution but also underscores the importance of integrating modern and historical datasets to decipher the dynamic processes shaping the Earth's surface. The methodology proposed, serves as a valuable approach for assessing and managing geological hazards in coastal regions affected by both climatic events and geodynamic activities.