流域降雨径流时间序列的混沌识别及其预测研究进展

混沌和随机在本质上是两种不同的特征,对这两种特征的描述方法各不相同,确定流域降雨径流时间序列的混沌性和随机性是对其进行模拟和预测的重要基础。近10多年来,许多学者相继开展了流域降雨径流时间序列的混沌识别及其预测研究。着重回顾其中最为重要的相空间重构、混沌识别和混沌预测方法,对将混沌理论应用于降雨径流时间序列的限制条件(序列的数据量大小和数据噪声)也进行了探讨。

Study advances in diagnosis of chaotic behaviour and its prediction for rainfall and streamflow time series in watershed

Chaotic and stochastic behavior have different character in essentiality, the describing methods for the two characters are not same each other, and identification of chaotic and stochastic behavior of rainfall and streamflow time series in the watershed is the most important basement for modeling and forecasting it. In the last decade, many hydrologists have engaged in the study of the diagnosis of chaotic behavior and the chaotic prediction methods of rainfall and streamflow time series. The reconstruction of the phase-space and the diagnosis of chaotic behavior as well as the chaotic prediction methods of rainfall and streamflow time series are reviewed. The two limiting conditions, the problem of data size and data noise are also discussed in chaos theory application to rainfall and streamflow time series.