A GIS-remote sensing approach for forest fire risk assessment: case of Bizerte region, Tunisia

In this era of climate change and global warming, forest fires are increasing around the world and especially in areas with arid and semi-arid climate. Hence, prevention is vital and it is considered as the best solution to protect forest areas. This paper presents a multi-criteria approach for the assessment and mapping of fire risk using three indicators: topomorphology index, climatic index, and human one. For each indicator, sub-indicators such as slope, morphology, exposure, number of fires, groundwater reserve, and evapotranspiration are chosen to generate a forest fire risk index in Bizerte region. Spatial data on all these indicators have been aggregated and organized in a geographic information system (GIS) framework. Results show that 33% of the total area of Bizerte forest is highly vulnerable to fire risk and an increasing of risk from 2013 to 2016. Sensitivity analyses indicated that the removal of the climatic (ICL) and the human indexes (HI) from the forest fire risk index causes large variation in the risk assessment. As a consequence, it should have higher weights than other indicators, which proves that triggering of wildfires is in the whole part caused by human activities and accelerated by climatic conditions. The remote sensing approach using NBR index confirms that severity of burned area increases throughout the time and the most changes are observed in the Northeast of Bizerte forest. These results can serve as a planning tool for decision makers to save the lives of residents and forest resources.