基于B-G分割算法的河川年输沙量突变分析

Analysis on annual sediment transport abruption of river basin based on B-G segmentation algorithm

引入一种适合非线性和非平稳时间序列突变检测的新方法——Bernaola-Galavan(B-G)分割算法,结合二项系数滤波,对嘉陵江流域年输沙量变化进行分析。在B-G算法的基础上,对嘉陵江北碚站近50年的实测年输沙量资料进行突变年检测,并从流域的泥沙来源、年内分配和年际变化、以及气候变化、水利水保工程对产输沙的影响等方面来探讨输沙量发生突变的物理背景。研究结果表明:1984年和1993年为嘉陵江北碚站年输沙量时间序列的两个突变点,北碚站的年输沙量明显减少,气候变化对流域产输沙的影响是主要的、直接的;20世纪90年代以来修建的一批大中型水库的减沙作用也十分显著;北碚站年输沙量发生突变是气候变化和人类活动共同作用的结果。

This article introduces a new detecting method for abrupt change, the Bernaola-Galavan（B-G）segmentation algorithm, which is adaptive to analyze the nonlinear and non-stationary time series. The abrupt changes of average annual sediment discharge in the Jialing River Basin from 1954 to 2005 were detetced, based on the introduced algorithm and binomial coefficient filtering. The sediment source composition, distribution characteristics in a year and annual variation, sediment yield and transport influenced by climatic change and water conservancy and soil conservation projects, were studied to find out the physical mechanism of the abrupt changes. The results show that abrupt points occurred at 1984 and 1993, and the sediment discharge has a significant change after 1993. The abruption of average annual sediment discharge in the Jialing River Basin is resulted from both climate change and human activity. Sediment reduction effect of climate change is direct, and the hydraulic engineerings built since 1990s have remarkable effect on sediment discharge change.