Gino Bella and Salvatore Filippone
Universita di Roma ``Tor Vergata''
Roma, Italy
and
Alessandro De Maio
Numidia s.r.l. Roma, Italy
email: salvatore.filippone@uniroma2.it
The control of forest fires is a very important problem for many countries around the world. Proper containment and risk management depend on the availability of reliable forecasts of the flame front propagation under the prevailing wind conditions, keeping into account the terrain features and other environmental variables.
Ideally one would like to have a single integrated system that can proceed from available observation sources, such as satellite images, from geographical map data, physical characteristics of the woods, providing in near real time an accurate forecast; such a system poses tremendous requirements in terms of coordination complexity and computing power.
In this paper we discuss our development of the central part of an integrated system, a tool to simulate the advancement of the flame front; in particular we discuss the pyrolysis model, the wood characteristics taken into account, and the modeling of heat exchange phenomena. Our modeling tool is derived from the well known fluid dynamics code Kiva, originally developed for engine design applications.
We highlight the computing requirements that a minimal simulation poses, the computational and parallelization techniques employed to reach our results, and the limitations of the current formulation; moreover we present some test cases and their significance with respect to realistic situations.
Finally, we give an overview of the future directions of development for this activity, with special attention to the models of combustion employed.
References:
1. R.H. White and M.A. Dietenber,
Wood product: thermal degradation and fire,
Elsevier Science Ltd., 2001
2. S. Filippone, G. Bella, P. D'Ambra, Parallelizing KIVA-3 to get the right simulation time in engine design, Proc. of Applied Simulation and Modeling, pp. 571-575, Crete, June 2002.
3. S. Filippone, M. Colajanni: PSBLAS: A library for parallel linear algebra computation on sparse matrices, ACM Trans. Math. Softw., 26(4):527-550, December 2000.