Environmental Management https://doi.org/10.1007/s00267-020-01389-z Factors Affecting the Behavior of Large Forest Fires in Turkey İsmet Daşdemir 1 Fuat Aydın 2 Mertol Ertuğrul 1 Received: 25 April 2020 / Accepted: 28 October 2020 © Springer Science+Business Media, LLC, part of Springer Nature 2020 Abstract Fire behavior should be estimated accurately and multidimensionally to reduce the dangers and harmful effects of forest res and to develop effective re management strategies. Therefore, in this study, the most important factors affecting re behavior were determined by statistical methods using data of 59 large forest res that occurred between 1977 and 2017 in Turkey. Some data were obtained from the General Directorate of Forestry, and the remaining were obtained from meteorological stations and Google Maps. The large forest res were investigated with descriptive statistics. Correlation, regression with generalized models, and factor analyses were performed. It was discovered that the rate of spread is particularly related to the wind direction, wind speed, and stand canopy. The most important factors affecting the behavior of large forest res were determined as (1) stand age, (2) stand canopy, (3) wind direction, (4) wind speed, (5) altitude, and (6) aspect, and 76% of re behavior is linked to these factors. Moreover, a forest re behavior index (FFBI) was developed for large forest res. They were categorized into three classes using the FFBI, and the characteristics of each class were identied. Finally, the ndings were discussed, and suggestions for re management were presented. Keywords Fire behavior Fire management Large forest re Multidimensional analysis Turkey Introduction Forest res have ecological, biological, technical, eco- nomic, administrative, and social dimensions. Forest res are natural events of a forest ecosystem and they may have both positive and negative effects on it. They can renew vegetation, regenerate forests, eliminate diseases and harmful threats, and reduce the amount of ammable materials. On the other hand, they may have destructive effects, and cause economic, social, and environmental losses. Therefore, forest res, which are a subcomponent of forest resource management, should be considered together with their ecological, economic, environmental, social, and other dimensions when preparing re management strate- gies and plans. The characteristics of re caused by environmental fac- tors are called re behavior (Bilgili et al. 2002). This con- cept is used to describe the path and development of res that follow. Indeed, re behavior can be used to estimate the characteristics and danger potential of an existing re in a given time and space (Hawkes et al. 1997). Fire behavior describes the movement (rate of area increase), intensity (ame length), and indicators of rapid combustion (spotting, crowning, and re whirls) of a re. It is expressed as a real time or predicted condition for ongoing res (NFDRS National Fire Danger Rating System 2002). The rate of spread (ROS) or ame length is commonly used as an indicator of re behavior. ROS denotes the difculty in controlling the re and the danger level (Albini 1976), and the ame length denotes the re intensity (Radtke 1983). The most important factors affecting the behavior of forest res are ammable material (fuel), weather (climate), and topography (Eron 1988; GDFGeneral Directorate of Forestry 1995; Bilgili et al. 2002; Güngöroğlu 2018). Fuel includes dead cover, debris, cutting wastes, grasses and grassy plants in the lower layer, various bushes, and trees. The nature, quantity, size, humidity, density, continuity, and quality of the fuel and the stand characteristics affect the re behavior. The weather consists of variables, such as pre- cipitation, relative humidity, temperature, wind speed, and * İsmet Daşdemir isdasdemir@hotmail.com 1 Faculty of Forestry, Department of Forest Engineering, Bartın University, Bartın, Turkey 2 Department of Fighting Forest Fires (Retired), General Directorate of Forestry, Ankara, Turkey 1234567890();,: 1234567890();,: