Conference Agenda

The aim of this project focuses on the development and application of processing methodologies to address the 3D characterization of sub-surface targets using stacks of spaceborne SAR data acquired over natural scenarios. Research activities so far focused on processing data from the BIOMASS Mission. Launched on April 29, 2025, the BIOMASS Mission features the first P-Band Synthetic Aperture Radar (SAR) ever flown in space. Thanks to the penetration capabilities of P-Band waves, BIOMASS has the potential to provide new insights on the response of dense tropical forest to electromagnetic waves, helping scientist understand how to best characterize forest scenario and infer biophysical parameters. In addition to that, BIOMASS data acquired over desert areas have been observed to reveal several features of sub-surface terrain penetration, allowing for the detection of features like drainage channels in polarimetric and interferometric data. So far, a major challenge in processing BIOMASS data appears to be associated with correcting ionospheric propagation. Especially at high latitudes, we observed that it is no possible to restore data quality without assuming multi-layer ionospheric propagation. On this basis, the following results will be shown at the symposium:

• 3D imagery of forested areas in Amazonia, in connection with forest height and forest above ground biomass density (AGBD).
• Evidence of sub-surface features in polarimetric and interferometric data acquired over desert areas.
• Analysis of ionospheric propagation and its effect on BIOMASS data at mid- and high-latitudes.
• Experimental evaluation of P-band SAR data for Digital Terrain Model (DTM) retrieval under dense canopies.

Earth surface deformation and slope instability represent major challenges for hazard mitigation, infrastructure safety, and sustainable land and water management in regions affected by tectonic activity, strong hydrological variability, and increasing human pressure. During the last year, Project 95369 implemented and validated an integrated Earth Observation framework that combines advanced Synthetic Aperture Radar time series analysis, optical remote sensing, geomorphological and hydro‑meteorological information, and Artificial Intelligence methods to support multi‑hazard monitoring and agricultural water management.

The project addressed a wide range of deformation phenomena and spatial scales, including large landslides, unstable slopes, ground subsidence, and localized deformation affecting transport corridors, urban areas, and critical infrastructure. Multi‑temporal InSAR techniques such as Persistent Scatterer InSAR, Small Baseline Subset analysis, enhanced scatterer approaches, and slope‑unit‑based methods were systematically applied using Sentinel‑1 and complementary datasets. These methods enabled the detailed characterization of deformation magnitude, spatial distribution, and temporal evolution in settings that include reservoir banks in the Three Gorges area, tectonically active mountain regions of the Third Pole and Karakoram, major highways, and large airport infrastructures.

Special attention was given to dynamic landslide susceptibility and deformation processes by integrating static terrain factors with time‑dependent variables such as rainfall, vegetation changes, reservoir water level variations, and post‑seismic effects. Machine learning models, including ensemble classifiers and tabular learning approaches, were used to improve prediction capability. Explainable Artificial Intelligence methods were applied to identify the relative contribution of conditioning and triggering factors, showing that seismic disturbances can reduce rainfall thresholds for landslide activation and that observed deformation often results from coupled geological, hydrological, and anthropogenic processes rather than simple fault control.

For infrastructure monitoring, dedicated InSAR strategies were developed to improve deformation detection over low‑coherence surfaces such as runways, engineered slopes, and embankments. Enhanced scatterer techniques and the combination of velocity maps with phase‑gradient information and three‑dimensional visualization allowed the identification and classification of localized deformation associated with soil conditions, structural loading, construction activity, and long‑term subsurface processes. These results provide a sound basis for early detection of instability and for risk‑informed infrastructure management.

In addition to geohazard monitoring, the project expanded SAR‑based approaches to agricultural applications, focusing on soil moisture assessment and water use optimization in water‑limited regions. By combining radar observations with optical data and meteorological information, the project supports the identification of water stress conditions and contributes to more efficient irrigation practices under changing climatic conditions.

The work carried out during the last year demonstrates that the integrated use of multi‑source Earth Observation data, advanced InSAR time series analysis, and Artificial Intelligence enables reliable, transferable, and interpretable solutions for comprehensive monitoring of Earth surface processes. The project provides solid scientific support for operational landslide and infrastructure monitoring, early warning applications, and data‑driven agricultural water management, contributing to risk reduction, infrastructure resilience, and sustainable resource use in line with international Earth Observation strategies and sustainable development goals.