Identification method of forest fire risk factors and their coupling relationship driven by attribute dependence

Identifying forest fire risk factors and their coupling relationships is crucial for revealing the mechanism of fire occurrence and assessing risks. However, previous studies on forest fire risk factors often ignore the differences and coupling effects between risk factors, resulting in inaccurate risk prediction. Given this, this study introduces the attribute dependence of fuzzy rough set into forest fires and constructs a new method to identify key risk factors and measure their coupling effects. Taking the forest fire risk in the Thompson area of Canada from 1984 to 2023 as the research object, the influence of each risk factor on forest fire is quantified by the attribute dependence function. Then, according to the influence degree of each risk factor, the weighted processing is carried out, and the coupling effect of risk factors is measured by combining the overlapping idea of fuzzy mutual information. The results show that climate factors have a significant impact on forest fires, while topographic factors, vegetation factors, and human factors are relatively weak. The coupling effect of topographic factors, vegetation factors, and human factors is also weak, but the combination of climate factors and topographic factors, vegetation factors, and human factors significantly enhances its influence.