Fractal and Spectral Analysis of Recent Wildfire Scars in Greece

Fire plays an important role in the complex physical interactions between the processes which define our planet's environments, and wildfires are considered to have a negative and unwanted effect, owing to the destruction of ecosystems and property that they often produce. In this paper, fractal and spectral methods are used to analyze the spatial characteristics of four wildfire scars of recent wildfires in the area of Greece. The fractal dimension, spatial noise and Hurst exponent of each of the chosen wildfire scar boundaries are calculated, and the results are discussed in connection to the real features of the wildfires. The results indicate that the fractal dimension can be used to adjust the fire line length by accounting for inherent roughness in the scar boundary, to define one of two types of propagation patterns that a wildfire exhibited, and to estimate the viability of plant recolonization in the wildfire area-furthermore, wildfire boundary spatial noise has been found to manifest similarities with pink noise, which is supported by the Hurst exponent analysis. The objective is to exemplify the effectiveness of such techniques for analyzing various characteristics of wildfire scars and active wildfires such as propagation patterns, fractality and spatial unpredictability, and for the qualitative determination of wildfire-related environmental factors.