Unstructured Cellular Automata and the Application to Model River Riparian Vegetation Dynamics
Cellular automata (CA) have proved to be a robust approach to spatially-explicit modeling of ecosystems. Conventionally the CA models applied a structured square grid. However, due to the anisotropic properties of environmental conditions, the capability
- PDF / 1,012,767 Bytes
- 8 Pages / 430 x 660 pts Page_size
- 57 Downloads / 165 Views
Abstract. Cellular automata (CA) have proved to be a robust approach to spatially-explicit modeling of ecosystems. Conventionally the CA models applied a structured square grid. However, due to the anisotropic properties of environmental conditions, the capability of CA method was not fully explored when using the regular squared lattice. This research investigated the unstructured cellular automata (UCA) by implementing a irregular triangular grid and used it to develop a vegetation dynamics model. The model was then coupled with a two-dimensional hydrodynamic model to simulate the riparian vegetation dynamics due to flow modifications by the reservoirs operations. The integrated model was applied to a compound channel of the Lijiang River in the Southwest China, which has been affected by the flow regulations for navigation purpose. Through the simulations, the previous evolutions of the riparian vegetations were recaptured and their future developments under the new flow regulation scheme were predicted. In particular, the potentials of UCA in ecosystem modeling were illustrated in the research.
1 Introduction Cellular automata constitute a mathematical system, in which the simple local components interact together to produce global complicated dynamics. Cellular automata have become a viable alternative approach to ecological modelling [1, 2], in particular after the emergence and application of object-oriented programming language. Conventionally cellular automata models apply a structured grid, for example the regular squared lattice. However, ecosystems are mostly characterised by the spatial anisotropy [3, 4], which indicates that some important local variations might be lost in these models. Although, the hexagon grid had been investigated [5, 6], they are still structured. Besides, they have not been widely used in the ecological models. The riparian zones are highly dynamic systems governed by interrelating physical and biological processes. The physical template of riparian zone is characterized by several typical geomorphic features, mainly including channel bed, channel shelf, floodplain and terrace [7]. Plant communities of different characteristics are distributed along the river side according to the ecological gradients. However, it is challenging to model the riparian successions, especially when flow patterns are largely modified by river regulations, for instance, the reservoir operations. H. Umeo et al. (Eds): ACRI 2008, LNCS 5191, pp. 337–344, 2008. © Springer-Verlag Berlin Heidelberg 2008
338
Q. Chen and F. Ye
Some previous attempts to this subject include the use of cellular automata approach [1, 8]. However, these studies all applied a squared grid, thus some key local features such as turbulence intensity and microscopic topography of riverbed were not well presented. The simulated spatial vegetation patterns on the shoals and floodplains are different from the surveys. This research developed an UCA model for vegetation dynamics by implementing the irregular triangular grid and inte
Data Loading...
