The degree of accuracy in model predictions of wildland fire behaviour characteristics are dependent on the model's applicability to a given situation, the validity of the model's relationships, and the reliability of the model input data. While much progress has been made by fire behaviour research in the past 35 years or so in addressing these three sources of model error, the accuracy in model predictions are still very much at the mercy of our present understanding of the natural phenomena exhibited by free-burning wildland fires and the inherent temporal and spatial variability in the fire environment. This paper will serve as a state-of-the-art primer on the subject of error sources in model predictions of wildland fire behaviour and includes a short historical overview of wildland fire behaviour research as it relates to model development.

A 3-m between crown spacing is a commonly cited criterion found in the wildland-urban interface fire literature for minimizing the likelihood of a fully-developed crown fire from occurring in a conifer forest on level terrain. The validity of this general recommendation is examined here in light of our current state-of-knowledge regarding crown fire propagation in relation to canopy bulk density. Given the characteristics of the overstory structure for 20 lodgepole pine (Pinus contorta Dougl. ex Loud. var. latifolia Engelm.) stands located in Alberta, as sourced from the literature, the canopy fuel properties following a virtual thinning to a 3-m crown spacing and then to a targeted canopy bulk density of 0.05 kg/m3 were computed. On the basis of these computations, crown fire potential was then analyzed and interpreted. The conclusion reached is that, in the majority of cases, a less widely spaced stand would be adequate for preventing crown fire development in lodgepole pine forests.

No abstract

No abstract

No abstract