集成灰色分析和元胞自动机用于景观动态模拟

New approach to dynamic modeling to landscape eco-pattern changes using grey analysis and cellular automata

针对景观生态系统"复杂性"的特点,采用灰色局势确定邻域转换规则,构建元胞自动机,用于模拟和分析景观生态动态变化。在信息不完全的情况下,提高了元胞自动机景观生态动态模型的可靠性和可行性。以中国云南省怒江流域中段作为实例,确定影响元胞转换规则的三因素主要有元胞邻域的聚集程度,土地适宜程度和人类活动影响程度,根据景观这一类复杂系统的动态变化特征其权重随地理空间位置和时间而变化,同时用Monte Carlo法考虑了模拟时转换的随机性。通过计算显示模拟未来发展的情景总的趋势是符合实际的,不但模拟了微观的景观单元的自组织机制,而且在一定程度上反映了宏观的社会经济因素影响,因而更具有针对性、典型性及准确性。

In this article, a new technique of simulating and analyzing dynamic changes in landscape eco-patterns is proposed based on grey-situation decision-based cellular automata. When only incomplete information is available, such a method can add credibility and feasibility to the dynamic modeling. Based on a case study on the middle reach of the Nujiang River in Yunnan province, three factors, including concentration of cellular neighbors, land suitability, and human disturbance, are found to be the major drivers of cellular transition, whose weight changes along with spatial and temporal alterations in accordance with the dynamic properties of such a complex system as a landscape. Randomness of transition in simulating is also taken into consideration by employment of the Monte Carlo method. Calculation results show that general tendencies seen in a simulated future scenario conform to the actual processes, not only simulating the micro self-organizing mechanisms of landscape units, but also reflecting to some extent the macro social and economic impacts. Hence, it can be concluded that such an approach is more problem-targeted, representative, and accurate than methods currently in use.