Fire disturbance data improves the accuracy of remotely sensed estimates of aboveground biomass for boreal forests in eastern Canada

Accurate estimation of aboveground biomass (AGB) using remote sensing data is still challenging and an approach based on an understanding of forest disturbance and succession could help improve AGB estimation. In the boreal forest of North America, time since last fire (TSLF) is seen as a useful variable to explain post-fire successional change and aboveground biomass (AGB). Within a large study area (>200 000km2) located in the northeastern American boreal forest, we compared remotely sensed biomass estimates of MODIS (Moderate Resolution Imaging Spectroradiometer), GLAS (Geoscience Laser Altimeter System) and ASAR (Advanced Synthetic Aperture Radar) with inventory-based estimates derived from ground plots, and forest maps at a spatial resolution of 2-km2. We identified that TSLF could explain the error observed in remotely sensed AGB estimates (MODIS (45%), GLAS (47%) or ASAR (23%)) when associated with surficial geological substrate information at that scale. Our results therefore show the importance of TSLF as a potential ancillary variable for improving the accuracy of remotely sensed AGB estimates in North American boreal forests.