Spectral burn severity does not fully capture post-fire forest structural loss: a case study of the 2020 Holiday Farm Fire

Background Wildfire can substantially alter forest vertical structure and aboveground biomass, yet commonly used spectral burn-severity indices may not directly represent three-dimensional structural change. Using the 2020 Holiday Farm Fire in Oregon, USA, as a case study, we integrated GEDI observations with Sentinel-1, Sentinel-2, and topographic data to characterize pre- and post-fire canopy height and aboveground biomass density (AGBD). We quantified structural losses, evaluated their relationships with the differenced normalized burn ratio (dNBR), and examined whether these relationships varied with pre-fire forest structure. Results Mean canopy height and AGBD decreased by 27.0% and 44.8%, respectively, following the fire. Structural losses generally increased with dNBR-based burn severity, but their distributions overlapped substantially among severity classes. Across spatially matched GEDI footprints, dNBR was positively correlated with canopy height loss ( r = 0.30) and AGBD loss ( r = 0.36), whereas relationships within individual severity classes were generally weak. Pre-fire canopy height and AGBD showed stronger associations with their respective absolute losses ( r = 0.57 and 0.71). Significant dNBR × pre-fire structure interactions further indicated that the relationship between spectral burn severity and structural loss varied with initial forest structure. Conclusions dNBR captured broad gradients of fire effects but did not fully characterize within-class variability in canopy height and biomass loss. Spectral burn severity and three-dimensional structural change therefore represent complementary dimensions of fire effects. Incorporating pre-fire structural information with dNBR can provide a more ecologically informative assessment of post-fire forest condition and support field assessment and recovery monitoring.