Search

Quantification of the interaction of intersecting and non-intersecting fire perimeters over a range of shapes, sizes and orientations is essential to understanding the behaviour of high-intensity wildfires that have become discontiguous as a result of spot fires or effects of broken topography or fuels. One key configuration is that of the V-shaped fire where two individual lines of fire intersect at oblique angles. Previous work under calm conditions in pine needle litter and straw found the speed of propagation of the vertex of the intersection to increase non-linearly as the angle of intersection decreased. The present paper investigates this relation in dry eucalypt forest litter in both the absence and presence of wind (~1.0 m s−1) and found that the increase in vertex speed under calm conditions was no greater than would be expected due to the geometry of the configuration. Conversely, in the presence of wind, the increase in the vertex propagation speed was substantially greater than explained by the geometry alone. Although these results suggest that fire line interactions can influence the behaviour and spread of coalescing fire fronts, further research is required to both identify the precise mechanisms driving this behaviour and quantify the resultant effects. Laboratory experiments were undertaken in a large combustion wind tunnel to investigate the interactions between two obliquely intersecting lines of fire in dry eucalypt forest litter (mean bulk density 66 kg m−3), with and without wind. Strong interactions were observed with wind but little interaction was observed under calm conditions. [ABSTRACT FROM AUTHOR]

The behaviour and rate of spread of a wildfire is strongly affected by local wind conditions depending on topography and surrounding vegetation. The wind speed within dense vegetation can be substantially lower than the open wind speed above the vegetation. This is commonly accounted for by applying a local correction factor to the local wind speed which can be called the wind adjustment factor (WAF). WAFs are often difficult to calculate and are usually based estimates for a particular vegetation type. Variation in the vegetation may result in spatially varying WAFs. We implement a model for spatially-varying WAFs based on leaf area index (LAI) and vegetation height data, which can be derived from remote sensing data sources. The model is implemented within Spark an operational wildfire prediction framework. Simulations of historical fires using the spatially-varying WAF model generally provided improved predictions compared to recorded fire extents than simulations without a spatially-varying WAF. • Wind adjustment factor is estimated using data from real tree canopies. • The model is implemented in the wildfire spread framework, Spark. • Historical fires are used to evaluate predictive improvements of the WAF model. [ABSTRACT FROM AUTHOR]