Preprints
https://doi.org/10.5194/egusphere-2025-6097
https://doi.org/10.5194/egusphere-2025-6097
30 Dec 2025
 | 30 Dec 2025

Technical note: Expanding ensemble forecasts with generative AI: application to volcanic clouds

Leonardo Mingari, Arnau Folch, Heribert Pascual, and Manuel Titos

Abstract. Ensemble-based modelling of the atmospheric dispersal of volcanic clouds enables more realistic forecasts by explicitly accounting for uncertainties in eruption source parameters, meteorological data, and systematic errors in transport models. Many ensemble applications, including quantification of forecast uncertainties, data assimilation, or probabilistic hazard assessments, require a large number of members to mitigate sampling errors and to properly capture probability distributions. However, running large ensembles with Volcanic Ash Transport and Dispersal (VATD) models can be computationally demanding, even for high-performance computing clusters. As a result, operational forecasting is typically restricted to smaller ensembles in order to fit time-to-solution requirements. In contrast, generative AI models can produce large volumes of physically-consistent data samples with minimal computational cost. In this work, a convolutional Variational AutoEncoder (VAE) is trained on an ensemble of 256 forecasts simulated with the FALL3D model and subsequently used to generate larger ensembles, effectively augmenting physics-based ensemble modelling capacity. Ensembles with up to 8192 members were generated nearly instantaneously using the trained neural network, with no reliance on HPC resources. The statistical properties of the expanded ensembles are characterised in detail, and the VAE performance is evaluated against a test dataset composed of 2048 numerical simulations. The VAE-generated ensembles closely approximate the actual (target) probability distribution as well as key sample statistics, such as ensemble mean and spread, with minimal degradation in the evaluation metrics. Finally, we discuss possible future applications of this work, including latent space data assimilation via deep learning.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.
Share

Journal article(s) based on this preprint

11 Nov 2026
Technical note: Integrating generative AI with physics-based simulations for probabilistic modelling of volcanic clouds
Leonardo Mingari, Arnau Folch, Heribert Pascual, and Manuel Titos
Atmos. Chem. Phys., 26, 15953–15973, https://doi.org/10.5194/acp-26-15953-2026,https://doi.org/10.5194/acp-26-15953-2026, 2026
Short summary
Leonardo Mingari, Arnau Folch, Heribert Pascual, and Manuel Titos

Interactive discussion

Status: closed

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on egusphere-2025-6097', Anonymous Referee #1, 25 Feb 2026
  • RC2: 'Comment on egusphere-2025-6097', Anonymous Referee #2, 01 May 2026
  • AC1: 'Comment on egusphere-2025-6097', Leonardo Mingari, 27 Aug 2026
  • AC2: 'Comment on egusphere-2025-6097', Leonardo Mingari, 15 Sep 2026

Peer review completion

AR – Author's response | RR – Referee report | ED – Editor decision | EF – Editorial file upload
AR by Leonardo Mingari on behalf of the Authors (15 Sep 2026)  Author's response   Author's tracked changes   Manuscript 
EF by Mario Ebel (16 Sep 2026)  Manuscript   Author's tracked changes   Supplement 
ED: Referee Nomination & Report Request started (16 Sep 2026) by Peer Nowack
RR by Anonymous Referee #1 (29 Sep 2026)
ED: Publish as is (01 Nov 2026) by Peer Nowack
AR by Leonardo Mingari on behalf of the Authors (02 Nov 2026)  Manuscript 

Interactive discussion

Status: closed

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on egusphere-2025-6097', Anonymous Referee #1, 25 Feb 2026
  • RC2: 'Comment on egusphere-2025-6097', Anonymous Referee #2, 01 May 2026
  • AC1: 'Comment on egusphere-2025-6097', Leonardo Mingari, 27 Aug 2026
  • AC2: 'Comment on egusphere-2025-6097', Leonardo Mingari, 15 Sep 2026

Peer review completion

AR – Author's response | RR – Referee report | ED – Editor decision | EF – Editorial file upload
AR by Leonardo Mingari on behalf of the Authors (15 Sep 2026)  Author's response   Author's tracked changes   Manuscript 
EF by Mario Ebel (16 Sep 2026)  Manuscript   Author's tracked changes   Supplement 
ED: Referee Nomination & Report Request started (16 Sep 2026) by Peer Nowack
RR by Anonymous Referee #1 (29 Sep 2026)
ED: Publish as is (01 Nov 2026) by Peer Nowack
AR by Leonardo Mingari on behalf of the Authors (02 Nov 2026)  Manuscript 

Journal article(s) based on this preprint

11 Nov 2026
Technical note: Integrating generative AI with physics-based simulations for probabilistic modelling of volcanic clouds
Leonardo Mingari, Arnau Folch, Heribert Pascual, and Manuel Titos
Atmos. Chem. Phys., 26, 15953–15973, https://doi.org/10.5194/acp-26-15953-2026,https://doi.org/10.5194/acp-26-15953-2026, 2026
Short summary
Leonardo Mingari, Arnau Folch, Heribert Pascual, and Manuel Titos
Leonardo Mingari, Arnau Folch, Heribert Pascual, and Manuel Titos

Viewed

Total article views: 3,291 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
2,278 819 194 3,291 527 470
  • HTML: 2,278
  • PDF: 819
  • XML: 194
  • Total: 3,291
  • BibTeX: 527
  • EndNote: 470
Views and downloads (calculated since 30 Dec 2025)
Cumulative views and downloads (calculated since 30 Dec 2025)

Viewed (geographical distribution)

Total article views: 3,279 (including HTML, PDF, and XML) Thereof 3,279 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 18 Nov 2026
Download

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Short summary
Forecasting volcanic clouds is challenging because accurate modelling can require many numerical simulations which are computationally demanding. We use a generative AI model to create large ensembles of realistic forecasts quickly and inexpensively. This approach expands traditional modelling capabilities, helping scientists better understand uncertainty and improve volcanic cloud predictions, unleashing generative AI modelling for atmospheric transport of volcanic clouds.
Share