【作者】 曾裕平；

【导师】 许强；

【作者基本信息】 成都理工大学 ， 地质工程， 2009， 博士

【摘要】 我国是滑坡灾害频繁发生的国家,尤其是近年来随着人类工程活动的日益加剧,再加上全球气候变暖、变乱,重大突发性滑坡灾害不断发生。如仅四川省境内,2004年发生了宣汉天台乡特大型滑坡;2005年丹巴县城后山出现明显的滑坡迹象;2006年绵阳北川县白什乡出现重大滑坡隐患;2007年发生了达州青宁乡岩门村滑坡、达县杨柳垭变电站滑坡;2008年汶川地震更是诱发了数以万计的滑坡灾害。重大突发性滑坡灾害已对人民生命财产构成严重威胁,急需有针对性地建立一套重大突发性滑坡灾害的预报预警理论和技术方法体系。由于影响滑坡发展演化的因素众多,导致滑坡运动具有复杂性和不确定性,滑坡的预测预报仍是世界性难题。目前对滑坡时间预报的主要作法是根据滑坡位移-时间曲线三阶段演化规律,当变形接近或进入临滑阶段时,结合宏观变形迹象进行滑坡时间预报。但目前在作时间预报时并没有定量标准,不易把握;而滑坡的运动范围一直是人们所关注的问题,因为这直接关系到滑坡对灾区人民的危害程度。众多的学者进行过滑坡滑速与滑距的计算研究,但提出的各种假说以及计算方法不具有普遍性,没有形成较为系统的预测方法。针对目前重大突发性滑坡预测预报所存在的问题,本文在搜集多个突发性滑坡完整监测资料的基础上,对滑坡各变形演化阶段的规律和特征以及滑坡下滑后可能的运动范围进行较深入系统地研究,获得如下创新性研究成果:(1)结合大量的滑坡监测实例,对滑坡位移-时间曲线特征进行了仔细地研究,并按形态将其分为六类:平稳型,直线型,曲线型,阶跃型,回落型以及收敛型。不同成因类型和不同演化阶段的滑坡,其位移-时间曲线是不同的。掌握滑坡位移-时间曲线特征,对定性判断滑坡的成因模式、变形发展阶段、外界因素影响程度,具有重要的指导意义。(2)为了克服滑坡位移-时间曲线切线角会随着纵横坐标尺度变化而变化的问题,本文提出通过坐标变化使位移-时间曲线各处切线角成唯一值的办法。在对改进切线角特征进行深入研究的基础上,提出了利用改进切线角判别滑坡的初始变形、等速变形以及加速变形阶段的定量划分标准。为实现对滑坡进行准确地临滑预报,进一步将滑坡的加速变形阶段划分为初加速、中加速和临滑阶段三个亚阶段,给出了三个亚阶段的切线角划分标准。同时发现,滑坡下滑前的最大切线角约等于89o。(3)对滑坡位移-时间曲线的加速度、加加速度特征和变化规律进行了深入地研究,提出了利用累积加速度和累积加加速度定量判别滑坡各变形阶段的方法,在此基础上,提出了滑坡临滑预报判据:累积加速度∑a >0且持续增长;累积加加速度∑? >0,且持续增长或保持高值。(4)提出了根据滑坡所处的不同变形阶段,采用不同预报模型的滑坡定量预报方法,即在滑坡的中长期变形阶段可采取神经网络法进行变形的趋势性预测;滑坡进入临滑阶段可采用Verhulst模型定量预报。本文同时提出了滑坡短期预报的抛物线法和临滑预报的加速度法。(5)对滑坡体滑面形态和裂缝类型及其各阶段的特点进行了较为系统地研究,总结出滑坡裂缝的分期配套规律和特征。根据裂缝的分期配套特征,可通过对滑坡宏观变形破坏迹象的研究,较为准确地判断和把握滑坡的所处演化阶段和稳定性状况,由此进行滑坡的定性预测预报。(6)为了分析评价滑坡下滑后的空间运动范围,将滑坡下滑分成5类模式:整体无碰撞滑行、整体碰撞滑行、滑落堆积、凌空飞跃和坡脚弧形滑动,并采用能量守恒定理和动量守恒定律,推导各类滑坡运动模式的运动距离估算公式。(7)在综合本文对滑坡位移-时间曲线特征、滑坡各阶段适宜的定量预报模型、滑坡裂缝分期配套特征等各方面研究成果的基础上,将滑坡预警级别分为蓝色-注意级、黄色-警示级、橙色-警戒级、红色-警报级共四级,提出了滑坡分级综合预警预报方法,给出了滑坡灾害分级综合预警表。在表中给出了各级的变形曲线特征、宏观变形破坏迹象、预报判据、滑坡对外界因素的响应特征、适宜的预报模型和方法以及灾害的应对措施。最后,选取了一些典型重大突发性滑坡实例,通过滑坡预测预报实例研究,对本文提出的理论和方法进行了实际的应用和检验验证。更多还原

【Abstract】 China is a country where landslide hazards are very active and bring a huge amount of damage to poverty and great threat to people’s lives. Especially affected by global warming, lots of landslides occur when the climate change abnormally in recent years. For example, in Sichuan province, the Tiantai village landslide(large scale) occurred in 2004; a large crack had been discovered on the mountain behind Danba town in 2005; the Baishi village landslide(large scale) happened in 2006, Mianyang city, Beichuan county; Landslides take placed at Yanmen (in Qingning village) and Yangliuya substation, Dazhou county in 2007, and more, lots of landslides broke out after Wenchuan 8.0 earthquake in 2008. It is urgent that a theory and technique system of forecast and prediction of large scale and sudden landslides should be established during the emergency situation at large disasters which damage to both people’s lives and property.Due to the complex triggering factors, landslide movement has lots of uncertainty and comprehensive influence elements, it is too difficult to predict landslide in a precise way which is still a global problem. Currently, the main method to predict landslide is according to the three developing stages of landslide in the displacement-time curve. When the deformation becomes close to or enters into the sliding stage, the prediction should be given combining the macro-deformation signs. However, the recent method only stays on the empirical aspect, and it is hard to be accurate to predict for no quantitative standard. The calculation of run out distance of landslide is also a big issue which has been concerned for a long time, because it is directly related with the disaster degree to peoples. Lots of scholars have provided several equations to calculate the speed and run out distance of landslide, but they are not available for other cases and still can not build a completed system to forecast it.Aiming to solve the problems in forecast and prediction of large scale and sudden landslides by gathering investigation and monitoring data from several cases, the paper studies detailly and deeply the evolution laws and characteristics in every deformation phases as well as possible run out distance after the fall of landslide, and provides some new results of research:(1) Based on lots of monitoring data, it is categorized the displacement-time curves into six types after thinking over carefully: stable, straight line, curve, stairs, up and down and Convergence. It is different among displacement-time curves according to causes and evolution stages of different landslides. To analyze landslides according to different types of curves is very important in describing the mechanisms of landslides, deformation stages and degree of disturbed by outer factors.(2) In order to solve the difficulties of change of displacement-time curves with the various coordinates, the paper shows the solution of finding out the specific tan angle of certain point of the curves through changing coordinates. Based on the adjusted tan angle characteristics, it is provided the standard criterions for distinguishing the different stages----the initialization deformation; equal speed deformation and accelerated deformation, of the deformation process after in-depth study. In order to realize the accurate prediction, it is given the criterions of recognizing the three sub stages----initial acceleration; middle acceleration and temporary slide, of the acceleration deformation. In the meanwhile, the max sliding angle is reaching to 89o before landslides fall。(3) For further research, the paper studies the characteristic and regulation of the acceleration and accelerated acceleration of displacement-time curves and provides the quantitative methods of how to recognize every deformation stage of slope and give the criterion of prediction of the sliding of landslide: Accumulate acceleration∑a >0 and keep on increasing; Accumulate accelerated acceleration∑? >0 and keep on increasing or stay at a high value.(4) The paper provides the different quantitative methods of prediction of the landslide with models according to the different deformation stages. For instance, it should be applied the artificial neuron network for the long term deformation to predict the deformation tendency; and applied the Verhulst methods to predict when the landslide enters into the sliding stage. The methods for predict the landslide in short period is also mentioned in this paper which is the parabolic curves modeling method as well as the acceleration method in sliding stage.(5) Based on the systematic researches on situations of different sliding planes of landslides and characteristics of cracks in various stages, it is summarized the laws and characteristics of cracks matching with various stage. According to the characteristics, it is judged and understood the deformation stages and the stability of landslide and applied to the forecast and prediction in empirical methods, based on the research of macro-deformation signs.(6) In order to analyze and assess the run out distance of landslide after it slides down, it is divided the types of sliding into 5 groups: whole sliding without colliding; whole sliding with colliding; sliding and depositing; flying, sliding at a arc plane of slope toe. It is also deduced the calculation equations of the landslide’s run out distance with the theories of the law of conservation of energy and the law of conservation of momentum.(7) Based on the synthesis of characteristics of displacement-time curves, appropriate landslide prediction models of stages and characteristics of cracks matching with various stage, It is mentioned in the paper to divide the forecast and prediction stages into four levels depending on the degree of dangerous: blue-notice level; yellow-warning level; orange–dangerous level and red-alarming level. The synthetic methods of dealing with each level are provided in the table with lots of information including characteristics of deformation curves, macro-deformation signs, prediction criteria, response characteristics of landslide to the outer factors, appropriate prediction models and methods and disaster response measures.In the end, based on the researches of cases of landslide forecast and prediction, the paper shows some large scale and sudden landslide cases with the application of those methods and theories which are mentioned in the former of the paper.更多还原