Dr. Claudia Corradino
University of Pittsburgh
Title: Volcanic hazard monitoring using Artificial Intelligence
Volcanic eruptions may pose a great hazard to people living close to active volcanoes. Although the monitoring of volcanic hazards is a vital task, it presents certain challenges, from the detection and quantification of hazardous phenomena during eruptive events, to forecasting their potential impact. The abundance of satellite and ground-based remote sensing image data is an important resource in the attempt to overcome these challenges especially with the ever‐improving temporal, spatial, and spectral resolutions of these instruments. The use of Artificial Intelligence (AI) in volcanology is an emerging field, that is taking advantage of these datasets. AI enables problem-solving by creating systems able to make predictions or classifications based on input data. Different types of AI, such as machine learning, deep learning and computer vision, are gaining importance in volcanology. This talk will highlight the use of remote sensing for volcano monitoring with particular focus on the use of AI algorithms to address volcanic hazard monitoring by using remote sensing. Datasets collected by both ground-based thermal monitoring systems, and passive and active satellite sensors are utilized to characterize volcanic events, with a focus on the application of these methods to volcanic activity from Etna (Sicily, Italy), Stromboli (Sicily, Italy) and Fogo (Cape Verde) volcanoes. Additionally, the capabilities of Google Earth Engine, a cloud-based platform combining a multi-petabyte catalog of satellite imagery and geospatial datasets for planetary-scale analysis, will be shown.