Key points are not available for this paper at this time.
This study presents an integrated remote sensing and modeling framework to assess the biophysical impacts of a large-scale wildfire that occurred on August 13, 2024, in the Yamanlar region of İzmir, Türkiye. The approach combines GEDI LiDAR-derived aboveground biomass density (AGBD), Sentinel-2 imagery, and the Revised Universal Soil Loss Equation (RUSLE) within the Google Earth Engine (GEE) platform. Pre-fire AGB was estimated using a Random Forest regression model trained on GEDI data and 14 environmental predictors, including vegetation indices, topographic metrics, canopy structure, and soil properties. Fire-induced biomass loss was quantified based on severity classes derived from the differenced Normalized Burn Ratio (dNBR), resulting in estimated biomass losses ranging from 6724 to 7673 Mg depending on the loss rates associated with each severity level. Additionally, pre- and post-fire soil loss was modeled using RUSLE, incorporating CHIRPS rainfall data, SoilGrids soil characteristics, SRTM-based topography, and Sentinel-2-derived NDVI for vegetation cover. The results revealed a clear increase in soil erosion, particularly in moderate-to-high severity zones, where average post-fire loss exceeded 6.7 tons/ha/year. This study demonstrates the utility of cloud-based geospatial analysis in quantifying wildfire impacts on forest ecosystems. The integration of LiDAR, multispectral data, and empirical models in a cloud-based environment enhances the speed and accuracy of post-fire assessments. The results highlight the role of remote sensing in tracking fire-induced forest degradation under changing climate conditions.
Eker et al. (Fri,) studied this question.