Photogrammetric Analysis of Post-Flood Geomorphological Changes along the Lilas River (Evia Island, Central Greece) Using UAS Data

Geomorphological transformations represent one of the most significant outcomes of high-magnitude flood events, as intense hydraulic forces have the capacity to rapidly reshape river channels, redistribute sediments, and modify the connectivity and functionality of adjacent floodplains. Understanding these processes is crucial for both hazard assessment and sustainable river management. In this context, the present study employs a multi-temporal approach using Unmanned Aerial Systems (UAS) combined with Structure-from-Motion (SfM) photogrammetry to detect, visualize and quantify geomorphological changes induced by flooding along selected sections of the Lilas River, located on Evia Island in Central Greece. These particular river reaches were strongly affected by the extreme flash flood that occurred in August 2020, an event that caused significant geomorphic disruption.

High-resolution aerial surveys were carried out both before the flood event, and shortly thereafter, in June 2018 and in September 2020 respectively. These surveys enabled the generation of highly detailed Digital Surface Models (DSMs) and orthomosaics, with a ground sampling resolution of approximately 2.5 cm. By performing differential analyses of the DSMs, the study was able to capture detailed patterns of erosion and deposition along the river corridor. The results indicate a pronounced spatial variability, with areas of intense erosion exhibiting local vertical lowering exceeding 7 meters, while zones of sediment accumulation showed depositional aggradation of up to approximately 5 meters after corrections for vegetation cover. Such extreme geomorphic changes highlight the uneven distribution of flood-induced forces along the river channel.

One of the most striking findings of the study is the substantial channel widening that occurred in response to the flood. At specific locations, cross-sectional widths expanded by factors ranging from three to nine, primarily as a result of lateral bank erosion. These findings underscore the complex interactions between natural geomorphic processes, extreme hydrological forcing, and anthropogenic landscape modifications, demonstrating that flood impacts cannot be understood without considering the coupled effects of these factors.

Overall, the study illustrates the capability of repeatable UAS–SfM workflows to provide high-resolution, quantitative evidence of flood-driven geomorphic change. Such data are invaluable for supporting post-event assessments, informing river restoration planning, and guiding the design of infrastructure adaptation strategies. Moreover, the results contribute to broader efforts in flood risk management, particularly in Mediterranean catchments that are highly susceptible to extreme weather events. By integrating detailed topographic measurements with hydrological and ecological considerations, the methodology presented here represents a powerful tool for anticipating and mitigating the consequences of future floods.