Rock-fall simulation and validation with in situ data: the case of Moira settlement in Western Greece.

Rock-falls are catastrophic phenomena, which have been studied widely through remote sensing data and special software analysis. Rock-fall investigation is quite crucial and especially the rock-fall prediction, as these phenomena cause serious damages and fatalities that could be avoided with a correct susceptibility mapping. In this paper, the RocFall software was chosen aiming to achieve a rock-fall simulation. Kinetic energy illustration, velocity calculation, end points detection and bounces heights depiction can be calculated by the specific program. A well-studied landslide in Moira settlement, near to the city of Patras (Western Greece) was selected for the simulation as many geo-data sets already exist. Slope maps and elevation profiles were extracted, in GIS environment, from different Digital Surface Models (DSMs). Those derived products were implemented in RocFall software for further simulations. More specifically, free available DSMs such as DSM from the Greek Cadastral DSM, ALOS AW3D30 DEM, ASTER GDEM, SRTM30 DEM, SRTM90 DEM, TanDEM_X as well as UAV DSM created by field campaigns were used for slope profile extraction. The results were assessed based on the spatial resolution of DSMs and were validated with in situ observations and measurements. The current study has two objectives: Firstly, to evaluate RocFall software outcomes with the field measurements and secondly, to estimate the influence of the DSM spatial resolution and accuracy to rock-fall simulation. According to previous studies, the spatial resolution is affecting the vertical accuracy of the DSMs. Among the aforementioned DSMs, UAV DSM was proved more appropriate for landslides simulation.