近30年来湖北洪涝灾害时序特征及预测分析——基于分形与混沌理论的研究

借助分形与混沌理论,对1978-2004年湖北省洪涝灾害成灾面积时间序列进行了分析,重构了其嵌入相空间序列,计算了其关联维数和饱和嵌入维数,确定了模拟相应动力系统所需的基本变量数目;同时,通过计算Kolmogorov熵,得出了对该洪涝灾害时间序列进行预测预报的时间尺度.最后,基于R/S分形分析,对湖北省未来洪涝灾害成灾面积的发展趋势作了预测,对该省洪涝灾害成灾面积可能呈现增大趋势的原因进行了分析,并提出了相应的对策.研究表明,该方法是一种有效的研究洪涝灾害时间序列分形特征的方法,能够客观、合理地反映洪涝灾害系统的分形特征,能为建立洪涝灾害成灾面积时间序列预报模型提供基础信息.

Time series characteristics and forecast analysis of flood disaster in Hubei Province in recent 30 years:fractal and chaos theories-based exploration

Based on fractal and chaos theory, authors analyzed the time-series of inundated area of flood disaster in Hubei Province from 1978 to 2004, rebuilt the time-series imbedding space, calculated the correlative dimension and saturation embedding dimension, determined the basal variable number to simulate correspond dynamics system, calculated the Kolmogorov entropy, and presented the forecast time scale of inundated area of flood disaster. At last, according to R/S analysis,we scientifically forecasted developing trend of inundated area of flood disaster in Hubei Province in the future, and analyzed causes of increased trend of inundated area of flood disaster in the province, as well as put forward the corresponding countermeasures. The results show that the above mentioned method is an effective way of fractal research into flood and waterlogging disaster system, which can reflect fractal characteristics of inundated area of flood disaster objectively and reasonably. The obtained conclusions will provide basic information for establishing the forecast model of inundated area of flood disaster.