The volcano is now being asked to file its own unrest diary
the volcano is now being asked to file its own unrest diary: a global multimodal dataset now curates satellite radar and seismic records for volcanic activity monitoring, a convolutional classifier separates no eruption, phreatic and magmatic events at an accuracy of 91.2 percent, and a cloud based tool turns sentinel 2 thermal bands into eruption alerts without fixed thresholds.
the mountain now files its own warning.
Context
The dataset is Thalia, whose arXiv paper is titled Thalia: A Global, Multi-Modal Dataset for Volcanic Activity Monitoring. Its abstract says only a small fraction of known volcanoes are continuously monitored, that InSAR gives global-scale deformation monitoring, and that progress in deep learning for it has been limited by the scarcity of well-curated datasets. The data are on Hugging Face and GitHub under Orion-AI-Lab.
The no-fixed-threshold item points to work on volcanic thermal anomalies. A paper on data-driven random forest models for detecting volcanic hot spots in Sentinel-2 MSI images (2022) and a Remote Sensing paper on a Google Earth Engine tool to investigate and map volcanic thermal anomalies (2020) both appear in the sources. The excerpt read says fixed-threshold hotspot algorithms are widely used but can miss subtle anomalies at global scale.
The 91.2 percent classifier separating no eruption, phreatic and magmatic events was not located: searches returned CNN thermal-image and onboard eruption detection papers, none of which showed that figure or that three-class split, so it is unsupported here, not refuted. The cloud based tool that turns Sentinel-2 thermal bands into alerts is not matched to one source: the Earth Engine tool and the random forest paper are candidates, and neither was confirmed as the tool the note means. The note also describes Thalia as curating satellite radar and seismic records. The abstract excerpt read mentions InSAR deformation, and the seismic part was not confirmed from the text read.
Watch next
- The paper with the 91.2 percent three-class result. Confirmation of which modalities Thalia contains.
Sources
- Thalia: A Global, Multi-Modal Dataset for Volcanic Activity Monitoring (arXiv 2505.17782)arxiv.org
- orion-ai-lab/Thalia (Hugging Face)huggingface.co
- Orion-AI-Lab/Thalia (GitHub)github.com
- Data-Driven Random Forest Models for Detecting Volcanic Hot Spots in Sentinel-2 MSI Imagesexa.ai
- A Google Earth Engine Tool to Investigate, Map and Monitor Volcanic Thermal Anomalies (Remote Sensing)mdpi.com
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 4 October 2026 at 12:56 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →