滑坡临滑预报的态矢量方法研究

Landslide forecast based on state vector method

态矢量是统计物理学中对连续场进行粗粒化描述的一种方法。将其引入滑坡预报中,通过将滑坡危险区划分为n个子区域,在各个时间窗内对一系列子区域的监测数据进行态矢量分析,将监测数据与降雨量态矢量的模等归一化后叠加,研究一段时间内滑坡位移态矢量的发展变化规律,并提出利用态矢量进行滑坡预报的新方法。以新滩滑坡为例,对累计水平位移进行态矢量分析,发现在滑坡前态矢量的模和态矢量增量的模均发生了明显突变,按态矢量曲线的变化规律,依据滑坡预警分级方法提出3级滑坡预警,最高预警级别预报的滑坡时间比实际滑坡时间提前2天。研究结果表明,态矢量的急剧突变是滑坡发生的前兆,可以用来进行滑坡的短期和临滑预报。

State vector is a method for describing coarse graining of continuous field in statistical physics. Introducing it into landslide forecasting, by dividing landslide hazard into n subregions, the state vector analysis is done to a series of monitoring data of subregion in every time window; through stacking modules of rainfall and monitoring data state vector after normalization, development and variations of state vectors of landslide displacements in a period are studied; and a new method is proposed to forecast landslide by using state vector. Taking Xintan Landslide for example, accumulative horizontal displacement is analyzed by state vector; it is found that the modulus and the incremental modulus of sate vector both make mutation before the landslide obviously, according to the variation of state-vector curves, based on the grading method of landslide early warning, three-grade landslide early warning is put forward; the highest warning level forecasts the landslide time is two days earlier than that of actual landslide. The result shows that the mutation of state vector is the precursory of landslide, which can be used to predict the landslide in short time and imminent sliding.