每年出版 6 期
ISSN 打印: 1543-1649
ISSN 在线: 1940-4352
Indexed in
MULTISCALE MODELING FOR STOCHASTIC FOREST DYNAMICS
摘要
This paper investigates a multiscale characterization of individual-based dynamics. At the fine scale, an individual-based and spatially explicit point process model is used to describe ecological processes relevant to the evolution of individuals and their mutual interactions. At the coarse scale, a stochastic model is synthesized to describe the dynamics of the system on larger spatial scales. The upscaling technique is designed to efficiently coarse-grain very fine-scale variables in order to express new state dynamics that no longer require access to fine details. The coarse dynamics are carefully constructed to preserve fluctuations from localized competition by retaining subgrid information obtained from precomputed Monte Carlo simulations of the fine-scale dynamics. Implications to computational efficiency and cross-scale coupling are significant. The present technique is then described in detail as it is applied to an individual-based and spatially explicit forest dynamics model.