Self-similar spatial clustering of wildland fires: the example of a large wildfire in Spain

Euclidean analytical tools for quantifying the patchiness of complex natural patterns such as wildfires in general do not recognize that measures of object densities over the landscape vary as a function of the size of the spatial elements. Therefore, improved fractal statistical methods are needed to quantify patch structures in ways that capture the spatial variation of the analysed pattern in relation to scale. The aims of this research are to show first, that remotely sensed wildfire scars are fractal objects, and second, how the resulting fractal dimension can be interpreted to give insight into the dynamics of fire spread over the landscape.