Volcanic Deformation

Volcanic deformation studies aim to understand and predict volcanic eruptions by analyzing ground movements preceding them, primarily using satellite imagery like InSAR data. Current research heavily utilizes deep learning, employing various architectures such as autoencoders, Patch Distribution Modeling (PaDiM), vision transformers, and contrastive learning methods to detect subtle deformation anomalies automatically and efficiently, often incorporating synthetic data to address data scarcity issues. These advancements improve the accuracy and speed of volcanic unrest detection, leading to more effective hazard mitigation and improved early warning systems for at-risk populations.