A comprehensive approach to soil burn severity mapping for erosion susceptibility assessment

Burn severity maps are typically generated using spectral indices and used in classifying the spatial distribution of damage caused by fires. In densely vegetated forests, even when overstory crowns are severely affected by the high-intensity fire, the topsoil may not experience high temperatures which makes spectral indices inadequate for assessing soil burn severity. On the other hand, field observations of soil burn severity can be subjective. For this reason, horizon-based soil sampling and extensive soil testing (physical, hydrological, chemical, mineralogical, and mechanical properties) were conducted in this study. Statistical tests have been employed to identify the most representative soil parameters of soil burn severity in the area. The remote sensing data (differential spectral indices and land surface temperature), field observations, and site-specific burned soil data were combined through weighted overlay analysis in the Geographical Information System (GIS). Accordingly, an improved soil burn severity map for the area affected by a forest fire in Kavaklıdere, Muğla, Türkiye was produced to show the post-fire soil erodibility potential. The findings of this study indicated that the effect of fire on soil properties was limited to the upper 0–4 cm of the soil profile with surface temperatures reaching a maximum of 300 °C for the high burn severity. The liquid limit, shear strength, organic matter, water repellency, and mean grain size were determined to be promising parameters to represent the soil burn severity. The map produced using the novel approach outperformed conventional burn severity maps. In addition, the high soil burn severity class can serve as a parameter to indicate erosion susceptibility after a wildfire.