Abstract

Natural hazards are significant problem that every year cause important loses. We report on both theoretical models and simulations aimed at better understanding of disaster spreading in various networks. The structure of the networks in this work is obtained either through neighbor analysis in real space or using models which reproduce generic features of real networks (i.e., power, telecommunication, or road networks). Our investigations are focused on the understanding of interaction between network structure and disaster spreading mechanism. The probability that fire will propagate through fire protection strip is investigated and a model is introduced based on finite-size considerations in percolation theory. Also, the uncertainty in prediction of fire propagation rate due to the local inhomogeneities of the vegetation cover is investigated. Finally, uncertainty in cascade failures of network infrastructure is analyzed for a model where edges have limited capacity. The importance of the results for disaster prevention and control is discussed as well.