Assessment of long-term post-fire vegetation recovery in Türkiye using multi-temporal NDVI and EVI

Wildfires are major disturbance factors in Mediterranean ecosystems, and monitoring post-fire vegetation condition is essential for ecosystem management, restoration planning, and wildfire mitigation. This study assessed long-term post-fire changes in spectral vegetation conditions in Türkiye using multi-temporal Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI) data derived from Landsat imagery processed on the Google Earth Engine platform. A 10-year post-fire assessment framework was adopted for 40 wildfire areas larger than 50 ha. Burned area boundaries were obtained from the European Forest Fire Information System, and analyses were conducted for the pre-fire, immediate post-fire, 10-year post-fire, and 2025 periods. The differenced Normalized Burn Ratio (dNBR) was also calculated to assess burn severity and its relationship to loss and recovery of vegetation spectral condition. The 40 wildfires covered 44,333 ha. Mean NDVI-based vegetation condition scores were 0.517 before the fire, 0.265 immediately after the fire, 0.498 ten years after the fire, and 0.492 in 2025. Mean EVI-based vegetation condition scores were 0.289, 0.137, 0.295, and 0.277 for the respective periods. Both indices showed a sharp decline immediately after the fire, followed by a substantial increase in later periods. NDVI- and EVI-based scores showed strong, statistically significant positive correlations, indicating consistent temporal patterns in vegetation spectral conditions. Friedman tests confirmed significant temporal differences for both indices, while Wilcoxon signed-rank post-hoc tests showed that immediate post-fire scores differed significantly from those in the other periods. The dNBR-based analysis indicated that higher burn severity was strongly associated with greater immediate loss of spectral vegetation condition. In contrast, its relationship with long-term recovery by 2025 was weaker and not statistically significant. Burned area size alone was not significantly related to NDVI- or EVI-based recovery metrics. Overall, the study highlights the value of integrating NDVI, EVI, and dNBR for satellite-based post-fire ecosystem monitoring in Türkiye.