A linkage between fire temperature and satellite-derived fire severity index in a north-central appalachia mixed-oak forest

The Differenced Normalized Burn Ratio (dNBR) is a widely used satellite-based index in fire severity assessment. However, gaps remain in the linkage between dNBR and fire temperature. This study aims to investigate the relationship between dNBR and fire temperature, demonstrating the potential of dNBR to retrospectively infer fire temperatures and generate spatially continuous fire temperature maps, which are challenging to achieve through fieldwork. A total of 209 plots were established in Zaleski State Forest, southeast Ohio. Three prescribed fires were conducted during spring and fall. The recorded fire temperatures were analyzed with both Landsat 8–9 and Sentinel 2-derived dNBR. Seasonal and spatial patterns of fire severity were also examined. The results indicate that both Landsat 8–9 and Sentinel-2 derived dNBR displayed significant positive relationships with fire temperature (R 2 = 0.378–0.509) for the spring burns, with Sentinel 2 fitting better than Landsat 8–9. No significant relationships were found in the fall burn between fire temperatures and dNBR for both sensors. The comparison between spring and fall burns revealed seasonal variation in satellite performance, with Sentinel 2 capturing higher dNBR values compared to Landsat 8–9 in the spring but Landsat 8–9 capturing higher dNBR values in the fall. These findings suggest a potential method to create spatial continuous fire temperature maps by retrospectively inferring fire temperature using satellite-based dNBR, filling the gaps that fieldwork can only measure discrete fire temperature. • Linking satellite-derived fire severity to field-measured fire temperature. • Landsat 8–9 & Sentinel-2 dNBR correlate with fire temperature in spring burns. • Sentinel-2 and Landsat 8–9 performed differently over spring and fall burns. • Seasonal and spatial patterns of fire severity were detected in the study area.