Radar modelling of forest spatial patterns.

The understanding of ecosystem processes and the impacts of natural and anthropogenic induced change upon these processes requires techniques to monitor structure, composition, and function of ecosystems. Radar backscattering is a useful monitoring technique since it is sensitive to vegetation water content, its arrangement within the 3-D canopy structure as leaves, branches, trunks, and the underlying surface characteristics. In this research we used a 3-D radar backscatter model and lacunarity spatial data analysis to study the effects of forest patterns caused by natural processes or forest management practices on spatial properties of radar images. The simulated results were compared with AIRSAR data from an area of managed forests. The findings from this theoretical study will direct our further research on spatial characterization of forests using SAR imagery. [ABSTRACT FROM AUTHOR]