Forest Fire Probability Prediction Taking Into Account Different Reasons of Anthropogenic Load

Introduction

Both the consequences and the causes of forest fires are diverse. For practical purposes, among natural causes, it is sufficient to single out the occurrence of forest fires as a result of thunderstorm activity. Reasons such as volcanic activity and self-ignition of forest fuel can be neglected. The first reason in mind is the extremely small share in the total number of forest fires. The second one is for physical reasons, since self-ignition of the ground cover and further formation of the center of the surface forest fire is impossible (Kuznetsov and Baranovskiy, 2006). Anthropogenic causes of forest fires are quite diverse. A model should be developed for a differentiated assessment of forest fire danger for anthropogenic reasons, taking into account both the specifics of our state and attracting information from foreign sources (Gorev, 2004; Pew and Larsen, 2001).

A comparative analysis of Russian and foreign data shows that, at least in the Tomsk Region, the vast majority of man-made fires are caused by the fault of the local population and other reasons can be neglected. However, on the territory of the Canadian island of Vancouver, the proportion of forest fires from other anthropogenic sources is high. Moreover, most fires have a mixed cause. Recent events in Spain allow us to talk about what needs to be included in the model and such a reason as arson. Therefore, in order to develop a universal geographically independent prediction model for forest fire danger as a result of anthropogenic load, all causes of fire outbreaks should be taken into account.

An analysis of the forest fire danger predicting methods existing in Russia and abroad shows that almost all methods have a weak physical basis (Baranovskiy and Kuznetsov, 2017). Thus, there is no doubt the need to develop a new methodology for determining the probability of forest fires. For example, in the nuclear industry, a probabilistic safety criterion is used (Grishin, 2002). Currently, the need to develop such a criterion and the corresponding methodology has arisen in forestry. As shown, the analysis of real forest taxation descriptions of specific forest stands should be detailed at the level of the minimum forest taxation unit (site). The forest fire danger scales currently used are usually tied to a specific region (Schetinskiy, 2002).

Thus, it will be relevant to develop such a system of scales that would unambiguously and uniformly interpret the level of forest fire danger in any region in the form of scales of the forest fires probability (Kuznetsov and Baranovskiy, 2009). A study of forest fires resulting from human activities on the territory of Vancouver Island (Canada) is significant (Pew and Larsen, 2001). The causes of forest fires from anthropogenic load on Vancouver Island are presented in Table 1. The causes of forest fires in the Tomsk Region are also quite diverse. Table 2 summarizes the causes of foci of forest fires in the region in 1993 – 2002 (Gorev, 2004).