How artificial intelligence can enhance monitoring of volcanoes from space

Cariello, S. and Corradino, C. and Del Negro, C. (2024) How artificial intelligence can enhance monitoring of volcanoes from space. Il nuovo cimento C, 47 (5). pp. 1-4. ISSN 1826-9885

[img]
Preview
Text
ncc13096.pdf - Published Version

Download (143kB) | Preview
Official URL: https://www.sif.it/riviste/sif/ncc/econtents/2024/...

Abstract

The ability to collect and analyze satellite data in near real-time (NRT) is fundamental for monitoring volcanic hazards and identifying the needed mitigation actions. This capability depends mainly on the availability of satellite data, which can range from a few minutes (SEVIRI, MODIS) to several days (MSI, OLI). Volcanology has embraced artificial intelligence (AI), incorporating both Machine Learning (ML) and Deep Learning (DL), allowing computers to learn from historical data and build a knowledge base of volcanic processes. An AI-based platform has been developed for NRT monitoring of volcanic activity, utilizing satellite imagery and offline analysis. ML and DL algorithms automatically evaluate volcano status from extensive multi-spectral satellite data flows. Built-in modules work together towards a common goal and are activated based on AI logic. These modules are adept at diverse tasks, from forecasting eruption initiation to quantifying erupted products using satellite thermal measurements. This platform represents an unprecedented convergence between technological innovation and the scientific community’s experience overcoming limitations of traditional approaches by inferring knowledge from large amounts of heterogeneous satellite data. It enables informed, timely, and sophisticated management of volcanic activity, significantly contributing to enhancing safety measures and minimizing the impact of volcanic events on communities and infrastructure.

Item Type: Article
Subjects: 500 Scienze naturali e Matematica > 530 Fisica
Depositing User: Marina Spanti
Date Deposited: 10 Jan 2025 12:02
Last Modified: 10 Jan 2025 12:02
URI: http://eprints.bice.rm.cnr.it/id/eprint/23324

Actions (login required)

View Item View Item