The bias involved in a point-process analog (Dijkstra algorithm) to a continuous fire is shown to be related to the number of points involved in the definition of adjacency and the degree of directionality of the rates of spread. If the point-process starts with a fire perimeter rather than a point-origin, the bias is markedly reduced. Forest Sci. 22:61-68.
No takes yet. Share an insight, caveat, or question.
O'Regan et al. (1976) studied this question.