A study on the effects of effusion rate on 2D numerical simulations of lava flow evolution

The effusion rate plays a crucial role in controlling the evolution of volcanic eruptions and it is thus critical to quantify the associated volcanic hazards. It represents the volume of lava emitted over time and can be estimated using satellite remote sensing data, such as Volcanic Radiative Power (VRP) measurements, deriving the Time Averaged Discharge Rate (TADR). Variations in the effusion rate can influence the resulting eruption style, ranging from effusive or Strombolian to lava fountaining activity. Thanks to the abundance of free data collected through fieldwork and remote sensing, it is now possible to numerically simulate eruptive behaviors with Computational Fluid Dynamics (CFD) models, allowing for a detailed study of lava flow evolution in time and space, without the dangers of collecting field data. We have already largely simulated lava flows with a Lagrangian mesh‑free particle CFD method, known as Smoothed Particle Hydrodynamics (SPH). This model takes in input the physical parameters of the fluid and returns in output a numerical simulation of its spatio‑temporal evolution. Here, we propose a framework to simulate lava flows with a range of values for some physical parameter, to better understand their results on numerical simulations, bringing to a deeper knowledge of the different eruptive behaviors in function of these parameters. In particular, we conduct a study on the effusion rate effects over 2D numerical simulations of lava using numerical physical‑mathematical models that return as output not only the evolution of the flow front but also the behavior of the vertical flow section. Using the TADR derived from VRP data and the SPH method, we simulate the fluid flow under different effusion rate conditions. This sensitivity analysis demonstrates how varying effusion rates affect the evolution of the flow front and vertical cross‑section. In detail, we show how time‑independent and time‑dependent effusion rates, obtained by satellite data, can make possible to qualitatively analyze and reproduce different eruptive styles, as effusive, Strombolian, or explosive activity and variable flows. The results highlight the potential for integrating real observations with numerical models and lay the groundwork for future applications that combine these approaches with artificial intelligence to enhance the model performance.