[S-CG57] Catastrophic eruptions: Mechanism and impact of eruptions

[SCG57-P04] Synergistic approach to robustly reconstruct eruption plume dynamics: application to Campi Flegrei

*Beatriz Martinez Montesinos¹, Yujiro J Suzuki¹, Antonio Costa², Leonardo Mingari³ (1.Earthquake Research Institute, University of Tokyo (ERI), 2.Istituto Nazionale di Geofisica e Vulcanologia (INGV), 3.Centro Superior de Investigaciones Cientificas (CSIC))

Keywords:Campanian Ignimbrite super-eruption, volcanic plumes models, atmospheric dispersal models, SK-3D, FALL3D, volcanic hazard assessment

The ~39 ka Campanian Ignimbrite (CI) super-eruption in Campi Flegrei (CF), southern Italy is the largest explosive eruption to have occurred in Europe in the last 200 ka. Characterized by two distinct plume forming phases, Plinian and co-ignimbrite, the CI eruption is credited with inducing a volcanic winter in Eastern Europe and influencing the Middle to Upper Paleolithic transition. Recent studies indicate a high intensity Plinian explosive phase depositing ~54 km³ of tephra deposits over an area of more than 1 million km², followed by a co-ignimbrite phase of more than 150 km³ of deposits spread over ~3 million km². CF is currently a densely populated area under busy air traffic routes where the monitoring system of the Vesuvius Observatory highlights some variations in the state of the volcanic activity. A similar eruption there would have a major impact on human life and health, air traffic, and infrastructure, at global scale. It is important to forecast what will happen soon after the volcano is erupting or to quantify the potential impacts of a future eruption. Therefore, we aim to reconstruct and understand the dynamics of the CI super-eruption using numerical simulation models. Understanding the CI super-eruption would increase the knowledge of the dynamics of other explosive super-eruptions, such as the Young Toba Tuff, Indonesia or Aso-4, Japan.

Atmospheric dispersal models (ADM) are commonly used to study the impact of tephra fallout needing eruption source parameters (ESPs) and calibration of empirical constants as input for the computation of numerical simulations. ESPs such as plume height and mass eruption rate can be estimated using field data, airborne observations, and analytical models. In addition, 1D plume and gravity current models can be used to calculate the spreading rate of umbrella clouds by adequately choosing the empirical constants. Previous studies used ADM combined with inversion techniques to reconstruct from geological deposits the different phases associated with CI super-eruption. However, ESPs and empirical constants are still variable among models and poorly constrained. The ash dispersion model FALL3D allows us to determine source conditions by inverting field data within an ensemble-based method. The eruption cloud dynamics model SK-3D accurately resolves the relevant turbulent scales of the volcanic plumes and the transport of volcanic ashes. Both models feedback each other using the estimation of the ESPs obtained from one approach to be used in the other to better reconstruct the eruption plume dynamics of the Plinian and co-ignimbrite phases of CI eruption and their impacts. In a first step, we use FALL3D to create an ensemble of forward simulations of ash dispersion by perturbing the input ESPs and meteorological conditions and, applying inversion techniques to the deposits thickness and grain-size data, obtain the best fit for the ESPs and the wind fields. Then, we use these constrained key eruption parameters as input for SK-3D to understand the eruption dynamics and calculate 1D plume profiles and, also, to investigate limits where gravitational spreading of the umbrella cloud dominates tephra transport due to wind winds.