Enhancing Photomonitoring techniques in landslide studies in the frame of Geosciences IR project

Landslides represent a significant geological hazard globally, with over 635,000 landslides identified in Italy alone. Despite their prevalence, only a fraction of these phenomena is actively monitored. Advancements in monitoring technologies offer promising tools for improving landslide management, but their application requires further validation and dissemination within the technical community.

This study, conducted under the PNRR “GeosciencesIR” project, investigates the use of photomonitoring techniques across fifteen landslide sites in Italy, where continuous or periodic monitoring activities are conducted. Monitoring setups feature ten ground-based cameras, while periodic drone-based acquisitions or photographic surveys provide supplementary observations for the remaining sites. The landslides encompass diverse mechanisms and kinematics, offering a robust basis for comparative analysis and evaluation of technique applicability.

The deployed monitoring systems utilize various hardware configurations, including optical cameras, robotic heads with Reflex cameras, and mobile devices. Images are predominantly captured in RGB format, and analyses are performed using the proprietary software “IRIS”, developed by NHAZCA S.r.l., employing change detection and digital image correlation algorithms. The techniques allow to identify variations (e.g., appearance or disappearance of objects in the FOV) or the track of object motion caused by landslide displacements between successive images over time. Additionally, time series of displacements have been extracted, providing insights into temporal evolution and supporting comparative validation against other monitoring data.

These sites have been continuously monitored since early 2024 and to date, over 140,000 images have been acquired, amounting to a dataset of more than 370 GB. Preliminary results include the identification of rockfalls, their size and timing, and the detection of retrogressive failure processes. For landslides with complex or flow mechanisms, the estimated 2D velocities provide consistent insights into motion trends, acknowledging the optimal performance and the inherent limitations of 2D analyses compared to 3D measurements. Project allows continuous feedback and data sharing with geological regional services, optimizing system operations and validation of results.

Challenges encountered during the project include ensuring the stability of monitoring equipment in remote locations and addressing environmental factors such as extreme weather conditions. Despite these hurdles, the collaboration with local technicians has facilitated knowledge exchange, fostering the development of photomonitoring techniques and their application in diverse geomorphological contexts.

This research advances monitoring methodologies, improving accuracy in displacement measurements and promoting cost-effective, accessible solutions. By promoting collaboration within the scientific and technical community, it aims to increase the number of monitored landslides and support innovative strategies for landslide risk mitigation.