Integrating Remotely Sensed Spatial Heterogeneity with a Three-dimensional Forest Succession Model

Abstract

To address questions concerning the effects of naturally occurring or anthropogenically induced spatial heterogeneity on forested stands and landscapes, it is necessary to couple spatially explicit simulators with georeferenced data. In this study, heterogeneity was quantified within a forested scene from central Maine, USA, by using edge detection algorithms designed for synthetic aperture radar (SAR) imagery. A spatially extended forest gap model that incorporates a three-dimensional solar radiation routine was used to simulate forest stand productivity as a function of edge orientation. The model results were then scaled up to the landscape based on the extent of edge and edge orientations detected in the SAR scene. The estimated increase in productivity, as a function of light-mediated edge effects, averaged approx equal to 26 to approx equal to 4% at the pixel (12.1x12.1-m) and scene (approx equal to 3600-ha) scales, respectively, over a 120-yr period after the edge was created.