The amount and spatial distribution of aboveground forest biomass (AGB) are required inputs to forest carbon budgets and ecosystem productivity models. While emerging satellite remote sensing data and empirical methods can estimate forest canopy structure and AGB, they do so to varying degrees of success, and few of these methods can account for the effects of terrain on observed reflectance. The presence of terrain can affect the accuracy of estimates in sub-alpine and montane environments. This paper introduces a new method for obtaining AGB from forest structure estimates using a multiple-forward-mode (MFM) canopy reflectance model inversion that include constraints that account for the effects of terrain. This approach first estimates average tree crown dimensions and stem density for satellite image pixels through indirect inversion of a geometric-optical canopy reflectance model. These crown dimension and stem density estimates are then related to average tree biomass and AGB for the pixel. The estimates of AGB from the MFM approach were evaluated for 40 field validation sites at Kananaskis, Alberta, along the eastern slopes of the Canadian Rocky Mountains. On average, AGB estimates were within 50 tonnes/ha (RMSE) of the field plot values where biomass ranged from 70 to 250 tonnes/ha. This result was similar, or an improvement over other NDVI and Spectral Mixture Analysis empirical methods tested.
展开▼