Extended Data Fig. 4: The effect of fire size on different postfire biogeophysical and surface energy flux changes for summer (red, June–August), winter (blue, December–February) and the annual time scale (black) for up to 14 years in northern forests (40°N–70°N).
From: Forest fire size amplifies postfire land surface warming
Linear regression models (y = α + β × log10(fire size)) were fitted across the whole study domain, where y stands for postfire changes in land surface temperature (βΔT, a), outgoing longwave radiation (βΔLWout, b), surface albedo (βΔα, c), reflected shortwave radiation (βΔSWout, d), ecosystem evapotranspiration (βΔET, e), latent heat flux (βΔLE, f), the sum of sensible and ground heat fluxes (βΔ(H+G), g), and net radiation (βΔRn, h) for different years after fire. Solid dots represent the β values from significant regressions (p < 0.05, the student’s t-test) while the empty dots represent insignificant regressions (p ≥ 0.05). Shading indicates the 95% confidence intervals, which are shown for all variables, although in some cases may be difficult to discern due to the narrow width of the confidence interval. Figure developed using the Python open-source tools.
