Using tree clustering method for forestry arrangement planning
No hay miniatura disponible
Fecha
Título de la revista
ISSN de la revista
Título del volumen
Editor
FAO ;
Resumen
Descripción
The difficulties of forestry in Russia are associated with the large territories (894.1 thousand hectares) and various climatic and forest zones. According to this, for efficient forest planning measures, 42 forest areas are defined legislatively. These areas are assumed to have similar conditions for use, conservation, protection, and forest reproduction within their boundaries. Alongside natural and forestry factors, some socio-economic and infrastructural factors need to be considered when planning the optimal scopes of fire-prevention measures in the
forests. In large territories, a significant number of factors and their combinations occur. It does not allow to numerically simulate all the processes connected with the emergence, distribution, and suppression of forest fires accurately. We present a new algorithm for optimizing the wildfire prevention arrangement regulations, which is crucial for increasing forest fire protection's overall efficiency when financial resources are limited.
The new approach discovers and assesses the similar indicators of investigated forest areas based on aggregated factors and their combinations. It enables identifying abnormal deviations, which are considerably below or above the mean values of the group. For the algorithm implementation, key factors are determined, and forest areas are grouped in obedience to the maximum similarity in factors complex. Determining key factors are: the relative number of fires per site, the fire-hazardous season's tension, population density, and transport accessibility of territories. A preprocessed data matrix is using for tree clustering of forest areas.
The constructed dendrogram is applicable for optimizing forestry regulations, a comparative effectiveness assessment of the forest fire protection system in particular areas and many other perspective tasks in decision-making systems in the forestry field.
ID: 3482167