Advancements in landslide monitoring by leveraging satellites observations using Offset Tracking algorithms

Satellite remote sensing techniques have emerged as crucial tools in monitoring and analysing Earth's surface, enabling insights into high-risk natural phenomena and enhancing decision-making processes. The PARACELSO project (funded by ASI – Italian Space Agency) aims to leverage satellites observations and innovative data analysis approaches to improve the mapping and surveillance practices of the Po River Basin Authority with respect to the dynamics of rivers, landslides and rock glaciers. As of landslides, the project aims to introduce the usage of techniques such as interferogram stacking and offset tracking (OT) for detecting and monitoring large-scale slope movements characterized by displacement rates higher than these allowing the application of multi-interferometric techniques. In such framework, and with reference to moderate velocity active earthslides and earthflows in the northern Apennines of Italy, this presentation deals with the application of OT algorithms implemented in Python (such as Normalized Cross-Correlation, Phase Correlation and Optical Flow), to imagery from the Sentinel 2 (multispectral), Prisma (hyperspectral) and Cosmo-SkyMed (X-band SAR) missions. Results obtained so far, validated by ground-based evidences and monitoring, confirm that offset tracking can become a powerful tool for leveraging satellite data for characterizing landslide dynamics over both short and extended periods of time. Furthermore, they evidence some limitations and the need for an optimization of data pre-processing routines (e.g., co-registration and terrain correction) and of the OT algorithms (so to reduce computing times). On such basis, it is concluded that using OT algorithms with satellite imagery can effectively allow the extraction of relevant motion distribution at the slope scale for specific landslides and, possibly, allow the identification of unrecognized active phenomena over quite large areas, so to advance the possibility of slope movements detection for hazard and risk management both for researchers and decision-makers.