During recent decades, forests have expanded into new areas throughout the whole of Norway. The processes explained as causing the forest expansion have focused mainly on climate or land use changes. To enable a spatially explicit separation of the effects following these two main drivers behind forest expansion, the authors set out to model the potential for natural forest regeneration following land use abandonment, given the present climatic conditions.

The mean shift (MS) algorithm is based on a statistical approach to the clustering problem. Specifically, the method is a variant of density estimation. We revisit in this article the MS paradigm and its use for clustering of remotely sensed images. Specifically, we investigate further the classification accuracy of remotely sensed images as a function of various MS parameters, such as the variant used, kernel type, dimensionality, kernel bandwidth, etc.

Lidar and orthophotograph-derived land cover are combined with in situ vegetation measurements to assess habitat characteristics typifying four species of butterflies with differing habitat preferences across a large spatial extent (∼30,000 ha) in northern Idaho, USA. Lidar data are employed to characterize both vegetation structure and topography, whereas digital orthophotographs and in situ vegetation measurements are employed to quantify surrounding land use and larval host plant cover, respectively.