Wildfires are a growing threat worldwide, with significant ecological consequences.This study compares the impacts of wildfires on natural ecosystems in the boreal forestsof Sweden and the Mediterranean regions of Portugal using remote sensing techniques.It focuses on two specific wildfire events: the 2018 wildfire in Fågelsjö, Sweden, andthe 2017 wildfire in Orvalho, Portugal, which was later followed by a second fire severalmonths later.

The study assessed burn severity using the differenced Normalised Burn Ratio (dNBR)and Relative differenced Normalised Burn Ratio (RdNBR) indices, and classified theseverity into four classes. Post-fire vegetation recovery was evaluated using a five-yeartime series analysis of the Normalised Difference Vegetation Index (NDVI) for eachseverity class.

The results show that the burn severity patterns differed between the two ecosystems,with the RdNBR index performing better than dNBR in capturing fire impacts acrossheterogeneous landscapes. In the Swedish boreal forests, the recovery rates increasedwith burn severity, with high severity areas estimated to recover within 8-9 years. Incontrast, the occurrence of a second fire in Portugal significantly altered the recoverytrajectories, with the low and moderate severity areas experiencing a more pronouncedimpact compared to the high severity regions. This highlights the complex interplaybetween successive disturbances and vegetation dynamics in Mediterranean ecosys-tems.

The findings provide insights into post-fire recovery patterns between boreal andMediterranean ecosystems and evaluate the effectiveness of Sentinel-2 imagery formonitoring post-fire vegetation dynamics. The study concludes with recommendationsfor future research to enhance understanding and improve remote sensing techniquesin ecological studies.