Using Landsat TM data to estimate carbon release from burned biomass in an Alaskan spruce forest complex

Fire disturbance in boreal forests can release carbon to the atmosphere stored in both the aboveground vegetation and the organic soil layer. Estimating pyrogenic emissions of carbon released during biomass burning in these forests is useful for understanding and estimating global carbon budgets. In this work, we have developed a method to estimate carbon eƒ ux for the burned black spruce in an Alaskan forest by combining information derived from Landsat Thematic Mapper (TM) data and ® eld measurements. We have used the spatial and spectral information of TM data to identify and measure two important factors: pre-burn black spruce density and burn severity. Field measurements provided estimates of aboveground and ground layer carbon per unit area for the pre-burn Landsat spectral classes, and percentage of carbon consumed for the post-burn Landsat spectral classes. Carbon release estimates for the burned black spruce were computed using ® eld data and the co-occurrence of the preburnandpost-burn spectral classes. Theestimated carbon releasedwas39.9 thaÕ 1 , which is 57% greater than an estimate computed using AVHRR data and estimates of pre-burn biomass and carbon fractions consumed that were not site speci® c or spatially varying. We conclude that the spectral bands and spatial resolution of Landsat TM data provide the potential for improved estimates of pyrogenic carbon eƒ ux relative to the coarser spectral and spatial resolution of other multispectral sensors.