Abstract
A Digital Twin (DT) is an accurate virtual representation of a physical object, phenomenon, or process. DTs are utilized in geosciences to simulate and analyze complex terrains. Developing DTs of rock glaciers is essential due to their significant role in addressing challenges related to climate change, monitoring permafrost, water resource management, and the dynamics of mountain ecosystems. Creating a DT of a rock glacier is very challenging because of the monitoring strategy (regarding spatial, temporal, and spectral resolution), the selected tools for processing, analyzing, and simulating data collected, and the computational infrastructure.
In this work, we study the rock glacier of Lazaun, South Tyrol, Italy (southern Ötztal Alps). The actively moving rock glacier (elevation range 2,480 to 2,700 m a.s.l.; 0.12 km2) is a prominent site for long-term monitoring and research, being one of the most investigated rock glaciers worldwide in terms of internal structure, motion, and hydrological behaviour.
Here, we present a scale-down strategy starting from remote sensing products using differential synthetic aperture radar interferometry (DInSAR) to analyze extensive areas and identify active zones. Over these identified zones, we introduced a proximal sensing approach using drones equipped with specialized sensors.
Using high-resolution cameras, we captured and combined overlapping images through photogrammetry techniques to generate detailed orthomosaics, 3D models, and Digital Surface Models. Additionally, we incorporated thermal imaging from UAV sensors to detect land surface temperature variations, inspect the presence of subsurface ice, and identify areas of activity.
These data sources offer unparalleled spatial resolution and detail, which is crucial for building an accurate DT. Using GNSS to determine displacement and velocity, we continued a long-lasting in-situ method to measure the coordinates of specific features (boulders). We integrate ground photography to identify their shapes in drone products for further automatic shape identification.
Finally, we introduced the use of FLAC3D (Fast Lagrangian Analysis of Continua for 3D modeling) to understand the propagation and evolution of the rock glacier movement by using a viscous constitutive model whose parameters have been calibrated by matching the velocity field of the central part of the glacier. We propose the use of Azure Digital Twins tool to visualize the possible combinations of data and scenarios.