Real-time 3D visualization of forest fire spread based on tree morphology and finite state machine

Abstract

State-of-the-art approaches to forest fire spread are based on either 2D numerical simulations of trees on GIS or rough 3D visualization. In this paper, we approximate the tree form by dynamically changing sets of tree-shape modules according to the morphological structure and wind fields. Guided by finite state machine, we define the states of equilibrium, heating, pyrolysis, cooling and destruction of tree-shape modules. Interactions between tree-shape modules drive the state transfer to achieve forest fire spread. Additionally, Loose Quadtrees are adopted to the spatial distribution of trees, which allows us to maintain the visual fidelity of the representation while rendering the forest fire spreads in real-time. Our method allows us to construct the Jiufeng forest example about 10km x 10km extent at interactive rates. The capabilities of tree-shape modules and forest fire spread visualization are demonstrated by numerous examples.