Succession modeling of river floodplain landscapes
The spatial organization of ecological units that constitute river floodplains was modeled using a recent image synthesis method. The model combined a purely spatial process, the stochastic pyramid, with analytical descriptions of ecological successions observed on each landform of the fluvial lands...
Saved in:
Published in | Environmental modelling & software : with environment data news Vol. 13; no. 1; pp. 75 - 85 |
---|---|
Main Authors | , , , |
Format | Journal Article |
Language | English |
Published |
Elsevier Ltd
1998
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The spatial organization of ecological units that constitute river floodplains was modeled using a recent image synthesis method. The model combined a purely spatial process, the stochastic pyramid, with analytical descriptions of ecological successions observed on each landform of the fluvial landscape. Because the stochastic pyramid is a generator of non-homogeneous, anisotropic, multicolored and random patterns, based on concepts of mosaics, it could be articulated with a series of diagrams which related the seral successional states to the spatial configuration of ecological units. Modeling was performed in two steps.
A priori probabilities were determined to transpose theoretical diagrams onto a processing image containing only the channel network of a real map; then after a random draw, a randomly textured image was obtained, this image was then transformed by the controlled stochastic pyramid in order to create a partition of wider regions without texture. As a result of this transformation, each colored pixel tended to expand and to incorporate its neighboring pixels that were randomly deleted. The model simulates various ecological maps of the same floodplain with both deterministic (fluvial network and ecological successions) and stochastic components (decimation and expansion rules). |
---|---|
Bibliography: | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 |
ISSN: | 1364-8152 1873-6726 |
DOI: | 10.1016/S1364-8152(97)00034-0 |