When lightning strikes woody vegetation it can cause a number of phenomena. Single trees can die; single trees can survive but have a characteristic strip of bark torn from top to bottom; single trees can be shattered; and, groups of trees can be killed by a single strike (Taylor 1969). The effects of lightning strikes on trees can be spectacular but description suggests, similarly, that spectacular results can occur when lightning strikes sand. In these cases lightning strikes may produce glassy, dendritic masses called "fulgurites" which are said to be common near Perth (see Kemp 1981).
The consequences of lightning striking trees vary widely. There are references to losses of wood products, creation of outbreak sites for insects (in USA - see Taylor 1969, 1974) and, perhaps of most interest to us, ignition of bushfires.
Cloud-cloud lightning flashes are of no consequence to starting fires, of course; it is the cloud-ground ones that are important. However, not every cloud-ground lightning strike causes a fire. The 'strike' consists of several very short duration return strokes followed by a 'continuing current'. The significant characteristic of the 'strike' for an ignition to result is the duration of the continuing current, not the current flow or diameter of the 'strike' (Latham and Schlieter 1989). Thus it is only those cloud- ground strikes with a sufficiently long continuing current that will cause a fire (given a receptive fuel bed).
Lightning strikes cause many fires in Australia although declaring what proportion of fires have been caused by lightning has been open to question (Gill et al. 1987). In remote hilly areas, lightning-caused fires are more likely to be of concern than are roadside fires caused by people because the former are going to be harder to reach than the latter. In these circumstances, the sizes of fires begun by lightning may be expected to be larger than those caused by people.
In the A.C.T., both McRae (1992) and Cary (1997) have developed models to predict where lightning-caused fires may be expected to start in mountainous terrain. These models may be useful in "fire hazard assessment" (McRae 1992) or for predicting fire regimes and their effects (Cary 1997). The models are only able to predict where lightning fires are likely to occur because lightning can strike twice (or more) in the same place!
Literature cited:
Malcolm Gill
Division of Plant Industry, CSIRO
18 May 2000