Pyroclastic Flow Simulation by a Cellular Automata Model

We developed a minimal three-dimensional Cellular Automata (CA) model for pyroclastic flows, generated by collapsing volcanic columns. Our aim is to simulate numerically simple ideal situations and to test that main phenomenological conditions appear correctly in the simulations. Model applications could be very important in order to assess pyroclastic flow hazard in densely inhabited areas. CA are a paradigm of parallel computing; they are based on a regular division of the space in cells, each one embedding an identical finite automata (fa), whose input is given by the states of neighbouring cells; fa have an identical transition function, which is simultaneously applied to each cell. At time t=0, fa are in arbitrary states and the CA evolves changing the state of all fa simultaneously at discrete times. An empirical method for macroscopic phenomena (applied successfully to lava flows and debris flows) is here considered for modelling pyroclastic flows. Each characteristic, relevant to the evolution of the system and relative to the space portion corresponding to the cell, is individuated as a substate. In our case substates of the cell are: the altitude, the elevation of the pyroclastic column, the energy, the thickness of solid particles deposits, the outflow from the central cell toward a neighbour cell. The values associated to such components can vary depending on interactions among substates inside the cell (internal transformation) and local interactions among neighbouring cells. The model has two internal transformations: degassing and particles deposition, computation of the new values of the substates. The model has been tested on a real case event, namely the 1991 Mt. Pinatubo eruption which occurred in the Philippines. The first simulations are encouraging. They represent a basis to improve this elementary model in order to better describe complex geological characteristics of such flows as degassing, deposition and granulometry.