Abstract: The watchtower is an important carrier for forest fire risk monitoring. In order to improve the intensity of forest fire prevention and protect the primary forest, the reasonable site selection of watchtower is crucial. This paper takes the primitive forest area in the northern part of Inner Mongolia Daxinganling as the research object, summarizes the historical fire point data of the forest area for 21 years from 1999 to 2020, uses BWM multi?criteria analysis and ArcGIS raster overlay to classify the fire risk in the forest area; establishes a multi?objective function as the search factor model for watchtower site selection according to the demand situation of forest fire prevention and monitoring in different risk areas; finally selects the optimal watchtower site by Finally, the optimal location of the fire watchtower is selected by the genetic algorithm. The optimized watchtower increased the coverage rate of low, medium and high risk areas of fire risk region by 13.7%, 23% and 41%, respectively. The optimized scheme improved the efficiency of fire risk monitoring and has reference significance to the forest fire prevention work in the northern forest area of China.

Key words: primary forest area, forest fire prevention, watchtower, site selection, ArcGIS