Jump to Main Content
Coupling fire behaviour modelling and stand characteristics to assess and mitigate fire hazard in a maritime pine landscape in Portugal
- Botequim, B., Fernandes, P. M., Garcia-Gonzalo, J., Silva, A., Borges, J. G.
- European journal of forest research 2017 v.136 no.3 pp. 527-542
- Pinus pinaster, computer software, coniferous forests, databases, fire behavior, fire damage, fire hazard, fire suppression, fire weather, forest stands, fuels (fire ecology), landscapes, managers, national forests, regression analysis, simulation models, stand characteristics, tree mortality, Portugal
- Silvicultural models are often developed and applied without due consideration of fire modelling. Yet, this information is important for designing treatment options to lower fire hazard. We used the FlamMap software to assess potential fire behaviour under extreme fire weather conditions within a 10,881-ha maritime pine landscape in central Portugal, the Leiria National Forest. Models describing fire hazard and providing information to assess potential benefits of stand-level fuel treatments were developed based on fire behaviour simulation. These models use as predictors stand variables and may assist forest managers in identifying hazardous areas in pine forests. Models were built from a database comprising 94,207 unique combinations of variables to detect significant fire-landscape interactions between stand-level features and fire behaviour. A set of compatible models that express crown fire likelihood and tree mortality were fitted using logistic regression. Additionally, classification tree analysis was used to model the type of fire, fire suppression difficulty, and tree mortality. The results highlight the potential of this methodology to explain the influences of fuel- and stand-related variables on fire hazard. This approach allowed the identification of straightforward discrimination rules to implement fuel treatments that prevent crown fires, enhancing the effectiveness of fire suppression and thereby reducing fire damage in fire-prone forest stands. Results further allow developing specific hazard-reduction prescriptions based on common forest metrics without resorting to advanced simulation modelling.