Landslide studying and monitoring by combining digital models from aerial visible and infrared photogrammetry

Abstract In this paper, a methodological framework to study and monitor a landslide through visible and infrared photogrammetry is presented. The procedure is applied to a coastal landslide that occurred within a nature reserve in 1993, which was surveyed using aerial sensors. Two types of landslide dense point clouds were generated through photogrammetry algorithms: the first arose from photos taken in daylight, while the second was realized with infrared images taken in dark environmental conditions, providing an innovation hint to the study. The combined analysis of the two models allowed achieving a morphological characterization of the landslide, through the recognition and mapping of the main features. Both models gave also hints on the presence of major structural planes controlling the landsliding motif of the studied cliff, which were measured at both the cliff and rock mass scales, through digital rock mass surveys at selected outcrops. Results allowed recognizing two main discontinuity sets, likely related to the major tectonic systems crossing the area. Finally, since the daylight surveys were repeated in 2021 and 2024, the resulting models were compared also to a preliminary dense point cloud built by using some 2014 landslide perspective photos. The so set 10-year monitoring suggested that the zone of depletion is affected by secondary movements, such as minor rockfalls and shallow denudation, while the foot is undergoing erosion by wave action. The combined analysis of visible and thermal dense point clouds proved a successful procedure for a complete landslide morphological and structural characterization, resulting in a reliable operation alternative to conventional surveys in restricted areas. Moreover, the digital monitoring allowed evaluating the evolution of key parts of the landslides, providing useful information in the frame of possible stabilization works or territorial planning.