HIERARCHICAL DECOMPOSITION OF VARIANCE WITH APPLICATIONS IN ENVIRONMENTAL MAPPING BASED ON SATELLITE IMAGES

A quadtree-based image segmentation procedure (HQ) is presented to map complex environmental conditions. It applies a hierarchical nested an alysis of variance within the framework of multi-resolution wavelet ap proximation. The procedure leads to an optimal solution for determinin g mapping units based on spatial variability with constraints on the a rrangement and shape of the units. Linkages to geostatistics are point ed out, but the HQ decomposition algorithm does not require any homoge neity criteria. The computer implementation can be parameterized by ei ther the number of required mapping units or the maximum within-unit v ariance, or it can provide a ''spectrum'' of significances of nested A NOVA. The detailed mathematical background and methodology is illustra ted by a salt-affected grassland mapping study (Hortobagy, Hungary), w here heterogenous environmental characteristics have been sampled and predicted based on remotely sensed images using these principles.