Evaluation of the fire potential index utilizing numerical weather prediction fields and remote sensing imagery

Forests are communes of large scale ecological units where many types of vegetation classes, fungi, microorganisms, animals, zooids, and soil live interactively. They cover vast territories and able to modify climatic conditions of their habitat. They serve many products and functions, some of which are utilized without effort. Prominent functions of forest are forestry products, hydrological, climatic, erosion control, medicine, recreational and scientific. Serving such benefits caused them to be an attraction for human population and finally lead to deforestation. It is estimated that national boundaries of Turkey had lost two third of forest assets since the end of ice age. Deforestation has many causes, but one of the most important one is forest fires. Climate change contributes to the increasing pressure over forested areas in many parts of the world and Turkey. Mediterranean countries are specifically more vulnerable to the increased chance of flammability in the forests. Turkey tries to protect its woodlands in those circumstances. Enhanced firefighting methods and strategies caused a decrease in the burnt areas per year despite the fact that annual number of fire occurrences are increasing. Forested fields and vegetation mass has an increasing trend in the last century. This is partially a result of a decrease in the demand for wood as raw material, heat, and energy source. Wildfires are a natural part ecological cycle and have a crucial role in the sustainability of forests. Primary functions of forest fires are dissolving organic biomass, maintain and preserve biodiversity, removing inorganic materials. Besides naturel role of forest fires in ecology, they also have detrimental effects. Those negative consequences might be listed as medical problems, increased carbon dioxide release into atmosphere, loss of habitat, increased vulnerability to erosion, and even fatalities. Due to serious consequences of wildfires they need to be investigated in order to better understand their dynamics, further effects, and avoided. Forest firefighting requires vast resources, and information. Due to severity of forest fires and difficulty of extinguishing efforts, a disaster management approach has to be applied which requires not only response strategies but also mitigation, adaptation, and preparedness. Early warning systems should be established indicating flammability in forested regions. Combustion occurs in the availability heat, oxygen, and fuel. Forest have a prevailing function of photosynthesis activity which turns carbon dioxide (CO2) and water (H2O) into cellulose (C6H10O5). That substance is known as cellulose. In the presence of favorable conditions, stored energy in the bounds of cellulose reveals rapidly in the form of forest fires. Ignition might start due to natural or anthropogenic factors. Most frequent natural cause of ignition is lightning whereas anthropogenic factors vary like intent, accidents, and negligence. Role of accidents is the most dominant factor igniting forests whereas natural causes play a smaller role in fire initiation. Ignition sources might vary regionally. Fuel characteristics are determined by type, age, layering, and season. Atmospheric conditions play a significant role presenting favorable conditions for fire. High temperatures, rainfall deficit, horizontal and upward velocity of air parcels contribute fire risk significantly. As causes and favorable conditions for fire events are well understood, scientists had many attempts to mimic fire potential. Hydrological circumstances, recent condition of atmospheric state, drought status, meteorological gauge readings, remote sensing data are used to estimate fire danger. Parameters chosen to signal fire potential might vary due to availability of data, regions specific dominancy of factors, and computation capability. Whichever method is selected for a fire potential index, it is expected to be easy to compute but still be able to produce reliable results. Fire Potential Index was chosen to estimate fire danger in Turkey in the present study. This method utilizes remote sensing data to determine vegetation status, atmospheric variables of temperature and relative humidity to indicate fuel moisture content, and land use data to distinguish between vegetation classes which have varying fuel characteristics like live/dead fuel loads and more importantly distinctive extinction moisture levels. Parameters of greenness level, water abundance of materials, and vegetation classes were together evaluated to signal likelihood of fire occurrence in a specific location. Daily fire potential mappings had been produced in order to investigate skill of index. General Directorate of Turkey supplied recordings of fire events in 2016. Model used in the current study was able to successfully simulate monthly deviation of fire risk. Correlation between monthly areal average of FPI score and number of forest fire occurrences was calculated. Findings revealed that FPI scores and number of fires were strongly correlated, r = 0.93, p < .01. Rapid changes of FPI scores in consecutive days were in a good agreement with changes in fire occurrences. Fire events were coupled with relevant FPI scores and it is found that FPI was unable to indicate some of the fire events. Specifically, coastal regions experienced lots of fire events when moderate or low fire risk was indicated. Whereas in some cases, inland regions were free of fire events when extreme risk level was indicated. Coastal domains have high population density; therefore more likelihood of ignition is present compared to areas with sparse population. FPI could seriously benefit from consideration of inputs to evaluate ignition chances. Coastal and inland regions are also in different climatic regions which leads modifications in vegetation. Therefore, fire vulnerability of species should be further investigated. Ecoregion specific adaptations to the FPI model would increase the chance of indicating fire risk more accurately.