EAGLE
Enhanced understanding of Alpine mass movements Gathered through machine LEarning
Abstract
Mass movements cause significant surface changes to the landscape and impact on the entire ecosystem including the anthroposphere. These processes are potentially associated to the current phase of global warming; thus, their investigation is of major importance to understand the implications of climate change. The project EAGLE (Enhanced understanding of Alpine mass movements Gathered through machine LEarning) is aimed at efficiently exploiting the increasing amount of information provided by Earth Observation (EO) satellites to study the status and the future evolution of mass movements in the Swiss Alps. The goal of EAGLE is to develop, validate, and distribute machine learning algorithms to be applied on EO datasets. The main data sources considered by EAGLE will be Digital Elevation Models (DEMs), optical/multispectral imagery, and radar imagery acquired from satellite sensors. EAGLE will deliver homogeneous inventories of slope movements, including also information on their status of activity. As additional outcome, EAGLE will deliver enhanced information on relevant infrastructures (e.g., urban areas, transportation corridors, hydropower dams, etc.) which are in areas potentially affected by current or foreseen changes due to glacier shrinkage, permafrost degradation, and/or landslide hazard. EAGLE will rely on data sources available worldwide, and thus the implementations resulting of this project will be exportable also to other environments and applicable to respond to different scientific goals.
People
Collaborators
PI | Partners:
description
Motivation
Slow-moving mass movements pose a major concern in the Swiss Alps, since they can result in significant hazards such as landslides, rock avalanches, and debris flows, endangering both human lives and infrastructure. Currently, in Switzerland, the detection and monitoring of these processes heavily rely on manual methods, where experts analyse Earth Observation (EO) data or conduct field measurements. Large constellations of satellites are constantly circling the Earth, capturing and storing high-resolution EO data, leading to an exponentially growing archive of data. Thus, the core objective of our project is to harness the power of machine learning to unlock the potential of this rich repository of EO data. By developing innovative algorithms and techniques, we aim to automate the detection and monitoring processes, providing more accurate countrywide mass movement catalogues for the Swiss Alps.
Proposed Approach / Solution
The SDSC takes the lead role in developing machine learning approaches for mapping alpine landforms of interest using various EO data, focusing specifically on Differential Interferometric Synthetic Aperture Radars (D-InSAR) data. While some models rely on optical images (Prakash et al., 2021; Prakash et al., 2020) or stacked D-InSAR data (Liang et al., 2023) to demonstrate promising mapping capabilities (refer to Figure 1), only a single approach has attempted to map mass movements directly from raw D-InSAR data (Bralet et al., 2024). Our efforts involve the utilisation of cutting-edge segmentation models from Computer Vision, such as U-Net, SegFormer, among others. In addition to segmentation, we are researching novel techniques for classifying the dynamics and activities associated with these processes.
Impact
Advancements in automating the mapping of alpine landforms promise to revolutionise the compilation of mass moments catalogues, and enhance Switzerland's ability to assess hazards, manage risks, and respond to disasters effectively.
Presentation
Gallery
Annexe
Additional resources
Bibliography
- Prakash, N., Manconi, A., & Loew, S. (2021). A new strategy to map landslides with a generalized convolutional neural network. Scientific Reports, 11, 9722. https://doi.org/10.1038/s41598-021-89015-8
- Prakash, N., Manconi, A., & Loew, S. (2020). Mapping Landslides on EO Data: Performance of Deep Learning Models vs. Traditional Machine Learning Models. Remote Sensing, 12(3), 346. https://doi.org/10.3390/rs12030346
- Liang, Y., Zhang, Y., Li, Y., & Xiong, J. (2023). Automatic Identification for the Boundaries of InSAR Anomalous Deformation Areas Based on Semantic Segmentation Model. Remote Sensing, 15(21), 5262. https://doi.org/10.3390/rs15215262
- Bralet, A., Trouvé, E., Chanussot, J., & Atto, A. M. (2024). ISSLIDE: A New InSAR Dataset for Slow SLIding Area DEtection With Machine Learning. IEEE Geoscience and Remote Sensing Letters, 21, 1-5. https://doi.org/10.1109/LGRS.2024.3365299
News
Latest news
Smartair | An active learning algorithm for real-time acquisition and regression of flow field data
Smartair | An active learning algorithm for real-time acquisition and regression of flow field data
The Promise of AI in Pharmaceutical Manufacturing
The Promise of AI in Pharmaceutical Manufacturing
Efficient and scalable graph generation through iterative local expansion
Efficient and scalable graph generation through iterative local expansion
Contact us
Let’s talk Data Science
Do you need our services or expertise?
Contact us for your next Data Science project!