Computational comparison and pattern visualization of forest fires

This paper analyses forest fire (FF) patterns based on information collected by the Canadian National Fire Database. The space-time dynamics of FF reveals characteristics that require advanced mathematical and computational methods. The study adopts the perspective of dynamical analysis to process the FF data considering amplitude, space and time. Several distinct techniques, such as dimensionality reduction, clustering and computer visualization unveil the relationships embedded in the data. The results reveal long term memory effects that demonstrate the usefulness of the proposed tools and motivate collecting further data related with this complex phenomenon.