泥石流往往以阵流群的形式出现,不同阵流参数具有统一的统计分布,意味着阵流序列隐含着深层的系统信息。文中根据蒋家沟泥石流观测数据,揭示阵流序列的自相关性和时空变化特征,进而确立其整体性。Hurst指数是刻画序列相关性的指标。阵流的流量、速度和时间间歇序列的Hurst指数值均大于0.5,从而证明序列具有长程相关性;另一方面,将阵流作为其运动参数(如流速、流量、间歇时间)组所确定的点,则一个泥石流事件为参数空间的一个点集,且集中在一个狭小区域,直观表现了阵流在时空的聚集特征,这是序列整体性的又一特征。阵流序列的这些基本特征,意味着不同泥石流事件背后隐藏着同一个动力学机制,即泥石流作为流域行为(而不是流体运动)的系统动力学。

Debris flow , which is usually of high density and viscosity , is conspicuous for it s composition of suc2 cessive surges of various properties. A debris flow is equal to a surge sequence and thus the surge sequence is believed to convey the information about the integrated properties of a system underlying debris flow. Based on observations in J iangjia Gully , in this paper we focus on the integrity of surge sequence by considering the autocorrelation and tempo2spatial interval. The Hurst exponent is employed to express the autocorrelation. Result s show that the Hurst exponent s are exclusively bigger than 015 , the value for random series without longrun af tereffect s , suggesting the long2range correlation of the surge sequences. Furthermore , an event2space is int roduced by taking each surge as a point assigned with a parameter set . Three parameters , the discharge , velocity , and time interval are chosen to define a surge point . Then a debris flow event appears as a group of point s clustered in a small range in the eventspace. This visualizes the concent ration of surges in temporal and spatial dimensions , and further illust rates the tempospatial correlation of the surges. The integrity and autocorrelation of the surge sequence can be taken as the sign of an underlying dynamics of debris flow.