, Volume 27, Issue 8, pp 1091-1108

Mapping large-scale forest dynamics: a geospatial approach

Rent the article at a discount

Rent now

* Final gross prices may vary according to local VAT.

Get Access


Digital map of forest dynamics is emerging as a useful research and management tool. As a key issue to address in developing digital maps of forest dynamics, spatial autocorrelation has been distinguished into “true” and “false” gradients. Previous ecological models are mostly focused on either “true” or “false” gradient, and little has been studied to simultaneously account for both gradients in a single model. The main objective of this study was to incorporate both gradients of spatial autocorrelation in a deterministic geospatial model to provide improved accuracy and reliability in future digital maps of forest dynamics. The mapping was based on two underlying assumptions—unit homogeneity and intrinsic stationarity. This study shows that when the factors causing the spatial non-stationarity have been accounted for, forest states could become a stationary process. A prototype geospatial model was developed for the Alaska boreal forest to study current and future stockings across the region. With areas of the highest basal area increment rate projected to cluster along the major rivers and the lowest near the four major urban developments in Alaska, it was hypothesized that moisture limitation and inappropriate human interference were the main factors affecting the stocking rates. These results could be of unprecedented value, especially for the majority of Alaska boreal region where little information is available.