Research on the comprehensive control technology system of large-scale debris flows in the area affected by strong earthquake in Longmenshan, Sichuan Province
-
摘要:
强震区泥石流表现出群沟暴发、范围广、持续时间长、规模大、危害重、防治难等特点,文章以强震区数十条典型特大泥石流沟为研究对象,采用野外调查、分析统计、试验验证及示范工程应用等方法,开展了典型泥石流沟特点及其防治工程效果研究,构建了强震区特大泥石流综合防控技术体系,形成了强震区特大泥石流勘查设计技术和防控关键技术体系。勘查设计体系在现有勘查技术基础上提出震裂物源识别新技术、堵塞系数分项取值新思路等;根据泥石流沟谷形态将强震区泥石流分为窄陡型和宽缓型;防控关键技术体系针对不同沟谷形态的泥石流分别进行综合防控关键技术探讨实践,统筹考虑了窄陡型和宽缓型泥石流沟上游、中游、下游沟谷特征。组合多种综合防控措施,建立了“起动控源→过程控量→末端控灾”逐级控制的强震区特大泥石流综合防控体系。结果可为强震区泥石流综合防控提供技术支撑,也可为非震区泥石流综合防控提供参考。
Abstract:In response to the challenges posed by frequent, widespread, long-lasting, large-scale, severe, and difficult-to-control debris flows in strong earthquake areas, this study, based on the results of the National Key R&D Program, investigates dozens of typical mega debris flow gullies in these regions. Through test verification and demonstration engineering application, the characteristics of typical debris flow ditches and their prevention and control engineering effects were studied, leading to the construction of a comprehensive technical system for the prevention and control of mega debris flows in strong earthquake areas. A key technical system for exploration, design, prevention, and control of mega debris flows in strong earthquake areas was formed. The exploration and design system introduced new technologies for identifying earthquake-induced fracturing sources and novel approaches for determining blocking coefficients based on existing exploration technologies. According to the morphology of debris flow valleys in seismic areas, debris flows are classified as narrow and steep or wide and gentle. The key technology system for prevention and control was established to comprehensively discuss and practice key technologies for preventing and controlling debris flows in different valley morphologies. Considering the characteristics of debris flow valleys in the upper, middle, and lower reaches of both narrow and steep or wide and gentle debris flow valleys, a combination of various comprehensive prevention and control measures was proposed. This led to the establishment of a hierarchical control system for major debris flows in seismic areas, from controlling the source to the process and ultimately mitigating disasters. This system not only provides technical support for comprehensive prevention and control of debris flows in seismic areas but also serves as a reference for comprehensive prevention and control of debris flows in non-seismic areas.
-
0. 引言
泥石流是一种自然地质灾害,发生时对当地居民的人身和财产安全造成重大的危害。我国泥石流研究防治工作始于20世纪60年代,早期的泥石流研究依靠实地考察,结合实地数据分析得出防治方法。唐邦兴等[1 − 2]通过对我国山地泥石流的考察研究,绘制了我国早期1∶600万泥石流灾害分布与危险区划图。陈宁生等[3 − 5]、崔鹏等[6 − 8]对四川、云南、新疆、西藏等特定地区泥石流的考察,提出了有效的泥石流判别方法和防治策略。随着遥感技术的发展,从数字高程模型(DEM)和遥感数据中获取与泥石流相关的地形地貌类因子和物源类因子弥补了实地考察的不便,通过提取沟谷的地貌因子、物源和降雨等因子,使用动力学模型FLO-2D、Debris2D、MatDEM等模拟特定沟谷的泥石流动力学过程,能够较准确的研究该沟谷或此类沟谷泥石流形成机理并提出预警和防治方法。
为了能够大面积地研究和评估,将研究区域通过汇水面积或其他阈值划分为小流域单元,统计各单元的泥石流相关因子,通过模型学习泥石流流域因子蕴含的特征来完成其他流域的分类预测。李益敏等[9]提取云南怒江州小流域单元的流域距断裂带距离、岩性、流域水系密度等11个因子,建立模型预测得到泥石流易发性分区,并绘制了怒江小流域易发性分布图。孙滨等[10]选取高程、坡度和坡向等9个影响因子,建立东川地区的小流域评价模型进行泥石流易发性评价。赵岩[11]详细构建了地貌、物源和激发三类泥石流因子的数据库,选取流域10 min降雨量、植被覆盖、6 h均降雨量等8个因子,使用线性回归的方式来对泥石流发生频率进行预测。统计学的方法能够在区域尺度上进行预测,但划分流域的方法不一,并且专家学者收集的影响因子和建模方法不同,导致预警评价结果不一,很难量化其评价预测的准确性。为了规避这些“主观”选择因子的过程,将卷积神经网络自动提取图像特征的方法用于地质灾害影像的研究成为了新热点。刘坤香等[12]提取怒江州泥石流沟谷的DEM图像,使用改进的残差网络学习图像特征,并用其他沟谷的分类预测。杨小兵等[13]基于深度学习方法对泥石流堆积扇的多光谱遥感影像进行训练和识别,以此找寻更多的泥石流隐患点。神经网络学习需要大量的样本数据,而区域尺度上的灾害数据属于小样本问题,小样本学习的提出用于解决样本量少的学习情况。茹颖[14]在遥感影像较少的数据集上,使用小样本学习对遥感图像分类得到了比传统卷积神经网络更好的效果。张萌月[15]在遥感影像识别目标较少的情况下,使用小样本学习方法相对传统的卷积神经网络有绝对优势。
本文以沟谷为评价单元,提取怒江州遥感影像建立泥石流沟谷数据集,使用基于原型网络(Prototypes Net)[16]的小样本学习方法对发生泥石流的沟谷影像特征进行学习,并将模型用于其余大量沟谷的泥石流易发性评价,借助ArcGIS软件将每条沟谷易发性结果和易发性分区进行可视化。
1. 研究区域与实验数据
云南省怒江傈僳族自治州是全国发生泥石流最严重的州(市)之一,怒江州地处云南省西北部横断山脉纵向岭谷区,辖泸水市、福贡县、贡山县和兰坪县,总面积14584.95 km2。州内高黎贡山、怒山山脉、云岭山脉近南北向延伸,澜沧江(湄公河)、怒江(萨尔温江)、独龙江(伊洛瓦底江)等深切峡谷平行南下,河谷谷坡坡度一般35°~45°、部分达60°~70°。按云南省各年减灾年鉴及网络中的不完全统计,截至2022年,有相关文字记载的泥石流灾害中,福贡县至少发生69起泥石流灾害事件、贡山县至少67起、兰坪县至少24起、泸水市至少42起。1950—2021年,福贡县各年平均降水量达到3460 mm、贡山县1632 mm、兰坪县1550 mm、泸水市2596 mm[17]。较高的降水量、高山峡谷的特殊地形及其他因素导致了怒江州泥石流灾害的频发。怒江州的地理位置和地形分布情况如图1所示。
泥石流遥感数据集的建立主要使用谷歌地图(Google Earth)遥感影像,地图的高程数据由SRTM提供,空间分辨率30 m,遥感影像则由Landsat,QuickBird,Spot等遥感卫星提供。Google Earth拥有高精度的时间和空间分辨率,城市区域空间分辨率可达1m,农村地区可达2.5~5 m,边远山区可达15 m[18]。数据的提取和处理如图2所示,利用数字高程模型(DEM)获取每条沟谷的掩膜信息,按照掩膜批量提取Google Earth遥感图像,为了保留沟谷的大小特征,将提取的遥感图像按照DEM比例进行重采样,并按照最大DEM沟谷图像的尺寸进行填充处理。
所使用的Google Earth遥感和DEM数据的信息如表1所示。
表 1 Google Earth遥感和DEM数据信息Table 1. Information on Google Earth remote sensing and DEM data数据信息 Google Earth DEM 数据来源 Google Earth软件 地理空间数据云
http://www.gscloud.cn/home数据类型 遥感图像 数字高程模型 空间分辨率 2~15 m
按照DEM采样为30 mASTER GDEM V2
垂直精度20m,水平精度30m通道数 3 1 时相 泥石流沟谷/与发生时间后最近时间分辨率影像图像
负样本及评测沟谷/最新遥感图像2015年1月6日发布的ASTER GDEM V2版本 泥石流遥感数据集包含正样本、负样本和待评价沟谷,正样本的确定是从《云南减灾年鉴》[19]、相关文献和网络报道中获得发生泥石流山谷和村庄的精确地点,依据坐标在Google Earth中与发生泥石流后最近的时间分辨率进行沟谷图像提取。负样本沟谷的确定依据为:在历史记录中不能检索到发生泥石流信息,影像中没有明显的冲积痕迹和堆积扇的形成且周围村庄密集。最终收集到泥石流沟谷50条作为正样本,未发生泥石流沟谷42条作为负样本,待评价沟谷样本600条。
2. 基于原型网络的沟谷泥石流灾害易发性评价方法
2.1 沟谷泥石流易发性评价流程
基于原型网络的沟谷泥石流易发性评价流程如图3所示,流程包括数据收集、沟谷提取、流域划分、数据集划分、数据集预分类、模型训练和预测、易发性评价和易发性分区图,详细流程如下。
首先收集研究区域的DEM、Google Earth遥感影像和历史泥石流数据,按上一节中的数据提取流程,以沟谷流域为评价单元,提取得到Google Earth沟谷泥石流遥感影像数据集和沟谷流域分区。其次,考虑形态各异的沟谷潜在危险性存在特征差异,直接将沟谷分为正负样本两类训练,在忽略特征差异的同时网络模型也不稳定,按照影响泥石流的重要因子流域面积[20],将正负样本按流域面积大小各预分为小、中、大3类,0、1、2属于正样本,3、4、5属于负样本,如图3中数据集预分类标签所示。以元学习的形式组织正负样本数据并输入到原型网络模型中进行训练,将待评价沟谷样本输入到已训练的模型中,与6个沟谷分类进行相似度量,计算每个评价样本的泥石流易发性指数。最后将易发性指数分区,得到沟谷的泥石流易发性评价等级,在流域单元划分中可视化评价结果,得到易发性评价分区图。
从流程图中可以看出,与基于泥石流因子进行评价的流程相比,基于原型网络的沟谷泥石流易发性评价方法直接使用沟谷影像来学习泥石流沟谷的特征,减少了从小流域计算众多泥石流因子和泥石流因子筛选的繁琐过程,避免了选择不同因子带来的评价差异问题。原型网络在保证性能的前提下,能够快速高效的进行泥石流易发性的评价。
2.2 原型网络模型与泥石流易发性评价算法
原型网络是小样本学习中经典的元学习和度量学习方法[21],网络模型如图4所示。原型网络以元学习的方式组织泥石流数据集进行训练,通过特征提取器(feature extractor)提取支持集(support set)每一类沟谷的特征,计算每类沟谷特征均值作为该类的原型中心(
),将查询集(query set)中待测沟谷 的特征( )与原型中心进行距离度量来完成待测沟谷的分类和预测。同时,按照度量距离计算待测沟谷从属类别的概率,并计算待测沟谷的泥石流易发性指数,从而得到待测样本的泥石流易发性评价等级。2.2.1 基于元学习方法的原型中心计算
元学习分为元训练和元测试。元训练将有限的训练样本分为不同的子任务进行训练,在小样本学习中称为C-way K-shot问题。具体为:在N类训练样本中,每类有少量样本,学习时每个子任务选取C类(
),每类中选取K个样本,共C×K个样本组成支持集(support set),即子任务的训练集。每类选取K个样本后,从每类剩余样本中各选取Q个样本(K+Q小于等于类内样本总数),共C×Q个样本作为查询集(query set),即子任务的测试集。不断从总类N中抽取C-way K-shot组成支持集进行训练,充分地利用了训练样本数据,使得模型在新任务时表现得更好。元测试则使用相同的方法来测试模型性能。在训练和测试泥石流数据时,原型网络将VGG[22]、GoogleNet[23]、ResNet[24]等不带分类器的卷积网络作为特征提取器,计算支持集中每一类沟谷的原型中心作为该类沟谷的代表,原型中心计算如式(1)所示:
(1) 式中:
——每一类沟谷的原型中心; ——训练批次中每一类沟谷的样本数; ——某一样本和对应的标签; ——特征提取器; ——提取器的参数。2.2.2 样本类别概率计算
如果将普通的神经网络分类器训练在小样本学习的任务中,几乎都是过拟合。基于非参数度量的方法(例如:最近邻、k最近邻、K-means、距离度量等)不需要优化参数,可以在元学习框架下构建端到端的小样本分类器。原型网络使用余弦或欧氏距离的距离度量方式作为分类器,通过距离度量沟谷样本
的特征与每类原型中心的距离大小来完成分类,并通过式(2)计算沟谷样本 属于某一类沟谷的概率:(2) 式中:
——某一样本 属于真实类别 的概率; ——距离度量函数; ——样本 通过 提取的特征; ——同一批训练的每一类别。在训练优化阶段,通过将式(3)中的目标函数
反向传播和随机梯度下降来对网络模型参数 进行优化。(3) 2.2.3 沟谷泥石流易发性等级评价
在进行沟谷的泥石流易发性评价时,通过计算沟谷的泥石流易发性指数来判定沟谷易发性等级。易发性指数通过所属类别概率来计算,文中原型网络使用欧氏距离来度量评价样本特征向量与类原型特征向量之间的相似程度,欧式距离越小,相似程度越大,使用式(2)计算所属类别的概率也就越大。
评价沟谷所属正样本中0、1、2类其中一类沟谷的概率越高,说明样本泥石流易发性越高。使用式(4)将所属概率转化为易发性指数。具体方法为:将正样本每类所属概率之和与负样本每类所属概率之和作差,差值大于等于0,表示评价样本与正样本相似程度更大,取正样本中的最大值作为易发性指数,值越大易发性越大。反之,差值小于0,取负样本中最大值的相反数为易发性指数,值越小易发性越小。将沟谷的易发性指数按大小进行分区,即可得到沟谷的易发性等级。
(4) 式中:
——验证样本 的易发性指数; 、 ——第 个样本的正样本所属概率之和、负 样本概率之和;m、k——正样本、负样本的类别数,文中m=k=3。
3. 实验结果与分析
3.1 分类性能
本实验实现的硬件为CPU:Intel Xeon E5-2650 v3、内存:128GB、GPU:NVIDIA GeForce RTX 3090。软件为Ubuntu18.06、Pytorch1.9.0、python3.8、CUDA11.1。实验时设置训练输入为6-way 10-shot,即每轮选6类,每类选10张训练,测试时输入6-way 4shot 1query,即每轮测试选取6类,每类4张计算原型,1张作为测试。选取9个网络模型结构,包括VGG、GoogleNet、ShuffleNetV2[25]、MobileNetV2[26]、Conv4[27]、ResNet12[27]、ResNet18[24]、DenseNet[28]、Rir[29]作为原型网络的特征提取器,分类结果如表2所示。
表 2 多个特征提取器的分类性能Table 2. Classification performance of multiple feature extractors特征提取器 6分类/% 2分类/% VGG 55.43 59.78 GoogleNet 51.09 56.52 ShuffleNetV2 47.82 51.09 MobileNetV2 53.26 55.43 ResNet12 40.22 44.57 ResNet18 44.57 48.91 DenseNet 50.00 64.13 Rir 51.08 57.61 Conv4 58.70 67.39 表2中,2分类为只考虑正负样本的分类正确率,可以看出,提取沟谷特征时,Conv4在原型网络中的6分类和2分类正确率优于其他模型。将正负样本沟谷输入此模型进行测试,得到6分类结果的混淆矩阵和正负2分类的混淆矩阵如表3和表4所示。
表 3 6分类测试混淆矩阵Table 3. Confusion matrix of six-classification6分类混淆矩阵 预测值 0 1 2 3 4 5 真实值 0 9 2 0 2 1 0 1 0 10 1 0 2 6 2 0 2 12 0 0 3 3 2 2 0 9 0 0 4 0 8 0 0 5 2 5 0 0 4 0 1 9 表 4 正负2分类测试混淆矩阵Table 4. Confusion matrix for positive and negative binary-classification test2分类混淆矩阵 预测值 正样本 负样本 真实值 正样本 36 14 负样本 16 26 使用分类指标对分类结果进行衡量,计算6分类的卡帕系数(Kappa)和2分类的准确率(Accuracy)、精确率(Precision)、召回率(Recall)、F1 Score,得到表5指标表。
表 5 2分类和6分类指标表Table 5. Binary-classification and six-classification indicator tableAccuracy/% Precision/% Recall/% F1 Score Kappa 67.39 69.23 72.00 0.71 0.50 表2—4表明,Conv4表现出了良好的分类性能,作为模型的特征提取器能够带来更好的泥石流易发评价结果。
3.2 评价结果及分析
保持同样的模型参数,将600张怒江州待评价沟谷及正负样本输入网络,计算每一个沟谷样本属于6类沟谷的概率,按式(4)计算每个样本的易发性指数
。使用自然断点法,按照易发性指数将易发区分为5个等级:低易发区[−1, −0.7607]、较低易发区(−0.7607, 0]、易发区(0, 0.6036]、较高易发区(0.6036, 0.8288]、高易发区(0.8288, 1]。最后,在ArcGIS中使用DEM进行易发性分区可视化,得到如图5所示分区图。从图5中可以看出,预测为高易发的沟谷主要分布在贡山县、福贡县、泸水市中,兰坪县较少。易发性沟谷分区中,低易发区249条、较低易发区96条、易发区100条、较高易发区124条、高易发区123条。使用ArcGIS计算沟谷的高程差、主沟长度、沟谷投影面积和坡降比。计算时,高程差为ArcGIS中使用沟谷的最高海拔减去沟口海拔得出,主沟长度为ArcGIS中使用流量计算出最长的河流长度,沟谷面积为使用CGCS2000投影坐标得到的栅格面积,坡降比即沟床比降,为高程差与主沟长的比值,反映沟谷的整体陡度。按照易发性分区将高程差、主沟长度、沟谷面积和坡降比分布可视化,如图6所示,作为对比,正样本分布如图6中虚线空心箱线图所示。
从分布情况中可知,在预测的沟谷中,低易发区、较低易发区、易发区、较高易发区在高程差、沟长、面积的分布上呈现下降趋势,坡降比呈上升趋势。易发区和较高易发区分布在高程差小、沟长短、面积小和坡降大的沟谷中。高易发区沟谷分布则较广,面积、高程差、沟长和坡降比从小到大都有高易发沟谷,这和正样本沟谷的分布有着相似的特点。
李益敏[9]使用泥石流因子评价方法,以5 km2的汇水面积为阈值,将怒江州划分为1 414个小流域单元,选取流域单元的距断裂带距离、岩性、Melton指数[30]、流域延伸率、流域高差率、河流弯曲系数、流域水系密度、平均植被覆盖度、年均降水量、距道路距离、距居民点距离共 11 个评价因子,采用确定性系数模型 CF 和多因子叠加权重确定法,对怒江州小流域单元进行泥石流易发性评价分区。与本文易发性分区对比,存在着相同点:预测为高易发及较高易发的沟谷(其文中为小流域)都集中在澜沧江以西,兰坪县较少,这与泥石流历史分布都较为符合。预测为高易发及较高易发沟谷(流域)数分别占全部沟谷(流域)的35.69%和34.72%,非常相近。
3.3 典型沟谷分析
在《云南省减灾年鉴》中检索到4条发生了人员伤亡或多次发生泥石流的沟谷,地理位置如图7中标签A、B、C、D处所示,编号A为贡山县普拉底乡东月谷,于2010年8月18日凌晨1时30分暴发特大山洪泥石流灾害,造成11 212人受灾,死亡96人,失踪53人,重伤9人,轻伤30人,转移安置灾民3 282人,直接经济损失1.4亿元。编号B为贡山县普拉底乡咪谷河,于2010年7月26日凌晨发生泥石流灾害,造成3人死亡、8人失踪、11人受轻伤,直接经济损失200万元。编号C为贡山县独龙江乡巴坡村沟谷,于2010年4月4日12时发生泥石流灾害,造成1人失踪,公路与通讯中断,巴坡村2014年2月20日、2017年8月12日都有发生泥石流的记录。编号D为兰坪县金顶镇七联村练登大沟,于2010年9月19日发生泥石流,对下游基础设施造成破坏,直接经济损失360万元,又于2019年8月7日发生泥石流。4条沟谷的遥感图像和沟口泥石流堆积情况如图7中所示。
4条泥石流沟谷的所属概率和易发性指数如表6所示。
表 6 4条泥石流沟谷原型网络计算的所属概率和易发性指数Table 6. Probability and susceptibility index of four debris flow valleys with prototype networks沟谷流域 编号 所属概率 易发性指数 0 1 2 3 4 5 普拉底乡东月谷 A 0.000 4 0.966 3 0.002 7 0 0.014 5 0.016 2 0.966 3 普拉底乡咪谷河 B 0.000 7 0.999 3 0 0 0 0 0.999 3 独龙江乡巴坡村沟谷 C 0.863 4 0.105 7 0.006 2 0.001 9 0.009 4 0.013 3 0.863 4 金顶镇七联村练登大沟 D 0.315 3 0.504 2 0.002 4 0.060 3 0.113 9 0.003 9 0.504 2 在预测中东月谷、咪谷河和练登大沟与1类泥石流沟谷相似度最高,巴坡村与0类泥石流沟谷相似度最高,东月谷、咪谷河和巴坡村三条沟谷的易发性指数都属于高易发性区间,这说明三条沟谷具有非常高的泥石流易发性。练登大沟易发性指数在易发区间,具有较高的泥石流易发性。
基于因子统计分析是较为成熟的分析方法,文中收集了沟谷相关的高程差、主沟长度、沟谷面积、沟谷周长、沟谷平均坡度、坡降比、Melton指数、切割密度和圆状率共9个地形因子。其中, Melton指数[30]由Melton M A提出,其值为流域高程差与面积平方根的比值,反映了集水区的动力学及其对泥石流的敏感性。切割密度为主沟长度与流域面积的比值,反映集水区内岩石的抗风化能力和地貌发育情况[31]。圆状率由沟谷周长与主沟长度比值所得,影响汇流时间。对9个因子使用随机森林的分类方法进行训练,模型正负样本分类准确率为61.11%,对4条沟谷进行预测,预测结果如表7所示。
表 7 4条泥石流沟谷因子分析方法计算的所属概率和易发性指数Table 7. Probability and susceptibility index of four debris flow valleys with factor analysis methods沟谷流域 编号 所属概率 易发性指数 0 1 2 3 4 5 普拉底乡东月谷 A 0 0.08 0.79 0 0.01 0.12 0.79 普拉底乡咪谷河 B 0.53 0 0.04 0.41 0.02 0 0.53 独龙江乡巴坡村沟谷 C 0.31 0 0 0.69 0 0 −0.69 金顶镇七联村练登大沟 D 0.68 0.02 0.04 0.11 0.13 0.02 0.68 可以看出,虽然使用因子分析和原型网络对4条沟的预测类别不同,但除了巴坡村预测错误,其他3条沟仍被预测为正样本。同样的,与李益敏[9]基于因子分析的易发性分区相比,4条沟谷在其评测中的易发性结果为:东月各为低易发、咪谷河为高易发、巴坡村为低易发、练登大沟为极低易发。对于四条已经发生泥石流的沟谷来说,特别是东月各沟谷发生特大泥石流灾害的特点,4条沟谷在本文原型网络评价中为高易发和较高易发,这与历史情况更为接近。
泥石流的发生是地貌条件、物质条件和激发条件共同作用的结果[11],选取若干小流域易发性评价实验[32 − 34]中认为重要的地貌条件和物质条件因子,对上面4条泥石流沟谷进行分析,分析如表8所示[35 − 36]。
表 8 4条泥石流沟谷的地貌条件和物质条件因子分析Table 8. Factors analysis of geomorphological and material conditions of four debris flow valleys地貌条件和
物源条件泥石流沟谷 东月谷 咪谷河 巴坡村沟谷 练登大沟 主沟长度/km 16.52 14.70 2.45 10.30 面积/km2 45.90 47.22 1.10 16.26 高程差/m 2 854 2 548 1 244 1 386 坡降比 0.17 0.17 0.51 0.13 平均坡度/(°) 15.46 26.49 33.64 16.17 Melton指数 0.42 0.37 1.18 0.34 植被覆盖[35] 有林地41.66%
灌木林10.42%
疏林地31.25%
中覆盖度草地16.67%有林地55.55%
灌木林8.88%
疏林地15.52%
中覆盖度草地5.03%
低覆盖度草地15.02%有林地50.00%
灌木林50.00%疏林地47.06%
高覆盖度草地47.06%
其它建设用地5.88%土壤性质[36] 松软薄层土
简育高活性淋溶土松软薄层土
简育高活性淋溶土简育高活性强酸土
简育高活性淋溶土不饱和雏形土
简育高活性淋溶土地层岩性[36] 花岗岩、板岩、千枚岩、片岩 花岗岩、板岩、千枚岩、片岩 花岗岩、石灰岩和其他碳酸盐岩 板岩、千枚岩、杂砂岩、长石砂岩、砂岩
石灰岩和其他碳酸盐岩在地形地貌和物质条件对泥石流形成的影响中,流域面积大汇水量多,主沟长、坡度大、高程差大、坡降比大、Melton指数大都会为泥石流的冲积提供动能。岩性对泥石流发生的影响中,片岩、混合岩、砂岩都容易风化产生松散物质,花岗岩、板岩、千枚岩、石灰岩、其他碳酸盐岩虽然比片岩、混合岩更硬,但是对泥石流也有较高的敏感性[37 − 40]。灌林和草地的含沙量比林地多,水渗透能力也比林地强,能够提供更多的松散物质和更强的雨水侵蚀能力[41 − 42]。淋溶土和松软薄层土也提供了松散物质[43]。东月谷、咪谷河、练登大沟3条沟谷中,疏林地、草地、松软薄土和淋溶土、砂岩和片岩等为泥石流提供了大量松散物质,练登大沟半个坡面为人工矿区,对植被破坏大,大量裸露的松散泥土,容易被雨水冲刷和搬运。其次,3条沟谷都有较长的主沟长度、较大的汇水面积、高程差、平均坡度和Melton指数,汇集大量雨水且促成强大的冲积能量,从图7中的堆积扇可以看出,冲积泥沙量巨大,阻断河流形成堰塞湖、毁坏公路和农田、摧毁房屋,造成巨大的经济损失。以人员伤亡和经济损失最大的东月谷为例,暴发泥石流时冲出体积约为 60×104 m3[44],破坏力极强。巴坡村虽然沟长和面积较小,但同样有足够的松散物质,并且坡度极大,降雨量达到一定程度,极容易形成小型泥石流或者滑坡,每当松散物质得到补充,只要一定的雨水条件激发,就会周期性的发生泥石流。
4. 讨论
文中考虑到形态各异的沟谷潜在的特征差异及模型的稳定性,按面积大小对泥石流沟谷进行预分类,再进行原型网络的训练,可能存在更为合理的预分类,让神经网络能更好地区分沟谷潜在的图像特征从而使得评价更为准确。其次,能够提取到沟谷图像的还有高分卫星遥感和数字高程模型,这2种图像对泥石流特征是否有着更好的表征能力,或者使用它们的组合能够带来更好的评测效果,是今后研究工作的一个重点。最后,以沟谷为单元进行评价,从历史信息中能够确定发生泥石流的沟谷数量极其有限,特别是极小部分多次发生泥石流的高频沟谷,网络并没有学习到这种高频特征。
李益敏等[9]使用因子评价的方法中,评价为高易发性的流域在贡山县、福贡县和泸水市中,但主要集中在贡山县,而本文中贡山县稍多,同时在贡山县、福贡县和泸水市中的分布比其更加“均匀”,更加符合历史泥石流灾害的分布规律。但是,评价的准确度需要从后续泥石流灾害事件中得到验证。而且,小样本深度学习的方法依赖于数据和模型,在后期的工作中,继续收集更多的泥石流沟谷图像,并不断改进模型,提高模型的精度,以此来减少模型性能带来的误差,提高预测精度。
5. 结论
基于原型网络的小样本学习方法对怒江州所有沟谷的泥石流易发性进行评价,经过实验,得出了以下结论:
(1)基于原型网络的小样本学习模型在泥石流正负样本的分类上精度为67.39%,模型具有良好的评价能力。
(2)通过模型预测,将沟谷泥石流易发性分为低易发区、较低易发区、易发区、较高易发区和高易发区5个等级,高易发和较高易发沟谷主要分布在贡山县、福贡县、泸水市中,兰坪县较少,这与历史泥石流发生点的空间分布较为吻合。
(3)选取4条发生泥石流的沟谷,使用原型网络对其易发性进行评价,同时选取高程差、面积、沟长等9个地形因子使用随机森林的分类方法进行评价,最后再与基于小流域的评价方法进行对比分析,只有原型网络全部评价正确,这说明,原型网络小样本学习的方法有良好的评价性能。
(4)对这4条典型沟谷的地貌条件和物源条件因子进行分析,4条沟谷都有着泥石流发生的地形地貌和物源条件,使用Google Earth遥感图像学习泥石流特征并进行泥石流易发性预测的方法,消除了选取因子不统一的问题,是泥石流易发性评价的一种新的方法,为泥石流易发性研究带来新的思路。
-
表 1 不同堰塞体库容组合下的泥石流堵塞系数取值表[46]
Table 1 Values of debris flow blockage coefficients under different dam reservoir capacity combinations[46]
序号 ⑤堰塞体 ④堰塞体 ③堰塞体 ②堰塞体 ①堰塞体 DC 1 库容组合 × × × × × DC0+DC1 堵塞系数 − − − − DC1 2 库容组合 × × × × √ DC0+DC1+0.8×DC2 堵塞系数 − − − DC2 DC1 3 库容组合 × × × √ × DC0+DC1+0.8×DC2+0.6×DC3 堵塞系数 − − DC3 DC2 DC1 4 库容组合 × √ × × √ DC0+DC1+0.8×DC2+0.6×DC3+0.4×DC4 堵塞系数 DC4 DC3 − DC2 DC1 5 库容组合 √ √ √ √ √ DC0+DC1+0.8×DC2+0.6×DC3+0.4×DC4+0.2×DC5 堵塞系数 DC5 DC4 DC3 DC2 DC1 注:“×”代表该堰塞体为空库,“√”代表该堰塞体为满库,“−”代表该堰塞体为无效堰塞体,不考虑堵塞系数。DC1、DC2、······表示对应堰塞体的堵塞系数。 -
[1] 铁永波. 强震区城镇泥石流灾害风险评价方法与体系研究[D]. 成都:成都理工大学,2009. [TIE Yongbo. The methodology and framework study of urban debris flow risk assessment[D]. Chengdu:Chengdu University of Technology,2009. (in Chinese with English abstract)] TIE Yongbo. The methodology and framework study of urban debris flow risk assessment[D]. Chengdu: Chengdu University of Technology, 2009. (in Chinese with English abstract)
[2] PENG Chao, YIN Zhiqiang, ZHANG Xujiao, et al. A comparative study of the main factors controlling geohazards induced by seismic events in Western China since the wenchuan earthquake[J]. China Geology,2021,5:1 − 15.
[3] 赵佳忆, 田述军, 李凯, 等. 岷江上游汶川地震前后泥石流易发性评价[J]. 中国地质灾害与防治学报,2024,35(1):51 − 59. [ZHAO Jiayi, TIAN Shujun, LI Kai, et al. Susceptibility assessment of debris flow in the upper reaches of the Minjiang River before and after the Wenchuan earthquake[J]. The Chinese Journal of Geological Hazard and Control,2024,35(1):51 − 59. (in Chinese with English abstract)] ZHAO Jiayi, TIAN Shujun, LI Kai, et al. Susceptibility assessment of debris flow in the upper reaches of the Minjiang River before and after the Wenchuan earthquake[J]. The Chinese Journal of Geological Hazard and Control, 2024, 35(1): 51 − 59. (in Chinese with English abstract)
[4] 张宪政, 铁永波, 宁志杰, 等. 四川汶川县板子沟“6•26” 特大型泥石流成因特征与活动性研究[J]. 水文地质工程地质,2023,50(5):134 − 145. [ZHANG Xianzheng, TIE Yongbo, NING Zhijie, et al. Characteristics and activity analysis of the catastrophic “6•26” debris flow in the Banzi catchment, Wenchuan County of Sichuan Province[J]. Hydrogeology & Engineering Geology,2023,50(5):134 − 145. (in Chinese with English abstract)] ZHANG Xianzheng, TIE Yongbo, NING Zhijie, et al. Characteristics and activity analysis of the catastrophic “6•26” debris flow in the Banzi catchment, Wenchuan County of Sichuan Province[J]. Hydrogeology & Engineering Geology, 2023, 50(5): 134 − 145. (in Chinese with English abstract)
[5] 蒋涛,崔圣华,许向宁,等. 基于遥感解译的典型强震区泥石流物源发育及演化——以四川都汶高速沿线为例[J/OL]. 地质通报,(2023-12-13)[2024-02-18]. [JIANG Tao,CUI Shenghua,XU Xiangning,et al. Remote sensing images based observational study on distribution and evolution of debris flow in a typic meizoseismal area-a case study of the Sichuan Duwen expressway[J/OL]. Geological Bulletin of China,(2023-12-13)[2024-02-18]. https://kns.cnki.net/kcms/detail/11.4648.P.20231212.1447.004.html. (in Chinese with English abstract)] JIANG Tao, CUI Shenghua, XU Xiangning, et al. Remote sensing images based observational study on distribution and evolution of debris flow in a typic meizoseismal area-a case study of the Sichuan Duwen expressway[J/OL]. Geological Bulletin of China, (2023-12-13)[2024-02-18]. https://kns.cnki.net/kcms/detail/11.4648.P.20231212.1447.004.html. (in Chinese with English abstract)
[6] 苗晓岐. 多源遥感技术在藏东南艰险复杂山区泥石流物源识别中的应用[J]. 地质通报,2021,40(12):2052 − 2060. [MIAO Xiaoqi. Application of multi-source remote sensing technology in the identification of debris flow source within complex mountainous areas in southeast Tibet[J]. Geological Bulletin of China,2021,40(12):2052 − 2060. (in Chinese with English abstract)] DOI: 10.12097/j.issn.1671-2552.2021.12.008 MIAO Xiaoqi. Application of multi-source remote sensing technology in the identification of debris flow source within complex mountainous areas in southeast Tibet[J]. Geological Bulletin of China, 2021, 40(12): 2052 − 2060. (in Chinese with English abstract) DOI: 10.12097/j.issn.1671-2552.2021.12.008
[7] 胡凯衡,崔鹏,游勇,等. 物源条件对震后泥石流发展影响的初步分析[J]. 中国地质灾害与防治学报,2011,22(1):1 − 6. [HU Kaiheng,CUI Peng,YOU Yong,et al. Influence of debris supply on the activity of post-quake debris flows[J]. The Chinese Journal of Geological Hazard and Control,2011,22(1):1 − 6. (in Chinese with English abstract)] DOI: 10.3969/j.issn.1003-8035.2011.01.001 HU Kaiheng, CUI Peng, YOU Yong, et al. Influence of debris supply on the activity of post-quake debris flows[J]. The Chinese Journal of Geological Hazard and Control, 2011, 22(1): 1 − 6. (in Chinese with English abstract) DOI: 10.3969/j.issn.1003-8035.2011.01.001
[8] BOVIS M J,DAGG B R. Debris flow triggering by impulsive loading:Mechanical modelling and case studies[J]. Canadian Geotechnical Journal,1992,29(3):345 − 352. DOI: 10.1139/t92-040
[9] 马超,胡凯衡,田密. 震后泥石流沟松散物质量与最大冲出总量的关系[J]. 自然灾害学报,2013,22(6):76 − 84. [MA Chao,HU Kaiheng,TIAN Mi. Post-earthquake relationships of maximum runout amount and loose matter mass in debris flow gully[J]. Journal of Natural Disasters,2013,22(6):76 − 84. (in Chinese with English abstract)] MA Chao, HU Kaiheng, TIAN Mi. Post-earthquake relationships of maximum runout amount and loose matter mass in debris flow gully[J]. Journal of Natural Disasters, 2013, 22(6): 76 − 84. (in Chinese with English abstract)
[10] 乔建平,黄栋,杨宗佶,等. 汶川地震极震区泥石流物源动储量统计方法讨论[J]. 中国地质灾害与防治学报,2012,23(2):1 − 6. [QIAO Jianping,HUANG Dong,YANG Zongji,et al. Statistical method on dynamic reserve of debris flow’s source materials in meizoseismal area of Wenchuan earthquake region[J]. The Chinese Journal of Geological Hazard and Control,2012,23(2):1 − 6. (in Chinese with English abstract)] DOI: 10.3969/j.issn.1003-8035.2012.02.001 QIAO Jianping, HUANG Dong, YANG Zongji, et al. Statistical method on dynamic reserve of debris flow’s source materials in meizoseismal area of Wenchuan earthquake region[J]. The Chinese Journal of Geological Hazard and Control, 2012, 23(2): 1 − 6. (in Chinese with English abstract) DOI: 10.3969/j.issn.1003-8035.2012.02.001
[11] 赵松江,赵峥,袁广. 九寨沟震区泥石流物源特征研究[J]. 四川地质学报,2021,41(1):93 − 97. [ZHAO Songjiang,ZHAO Zheng,YUAN Guang. A study on material source characteristics of debris flow in the Jiuzhaigou valley earthquake area[J]. Acta Geologica Sichuan,2021,41(1):93 − 97. (in Chinese with English abstract)] DOI: 10.3969/j.issn.1006-0995.2021.01.019 ZHAO Songjiang, ZHAO Zheng, YUAN Guang. A study on material source characteristics of debris flow in the Jiuzhaigou valley earthquake area[J]. Acta Geologica Sichuan, 2021, 41(1): 93 − 97. (in Chinese with English abstract) DOI: 10.3969/j.issn.1006-0995.2021.01.019
[12] 杨成林,陈宁生,李战鲁. 汶川地震次生泥石流形成模式与机理[J]. 自然灾害学报,2011,20(3):31 − 37. [YANG Chenglin,CHEN Ningsheng,LI Zhanlu. Formation mode and mechanism fordebris flow induced by Wenchuan earthquake[J]. Journal of Natural Disasters,2011,20(3):31 − 37. (in Chinese with English abstract)] YANG Chenglin, CHEN Ningsheng, LI Zhanlu. Formation mode and mechanism fordebris flow induced by Wenchuan earthquake[J]. Journal of Natural Disasters, 2011, 20(3): 31 − 37. (in Chinese with English abstract)
[13] 郝红兵,赵松江,李胜伟,等. 汶川地震区特大泥石流物源集中启动模式和特征[J]. 水文地质工程地质,2015,42(6):159 − 165. [HAO Hongbing,ZHAO Songjiang,LI Shengwei,et al. The star-up mode on large debris flow material source in Wenchuan earthquake region[J]. Hydrogeology & Engineering Geology,2015,42(6):159 − 165. (in Chinese with English abstract)] HAO Hongbing, ZHAO Songjiang, LI Shengwei, et al. The star-up mode on large debris flow material source in Wenchuan earthquake region[J]. Hydrogeology & Engineering Geology, 2015, 42(6): 159 − 165. (in Chinese with English abstract)
[14] 胡涛. 汶川震区震后大型泥石流致灾机理及防治对策研究[D]. 成都:成都理工大学,2017. [HU Tao. Study on disaster mechanism and prevention countermeasures of large-scale debris flow after Wenchuan earthquake[D]. Chengdu:Chengdu University of Technology,2017. (in Chinese with English abstract)] HU Tao. Study on disaster mechanism and prevention countermeasures of large-scale debris flow after Wenchuan earthquake[D]. Chengdu: Chengdu University of Technology, 2017. (in Chinese with English abstract)
[15] 陈多鸿. 基于SPH-DEM-FEM耦合的三维泥石流动力过程模拟研究[D]. 宜昌:三峡大学,2023. [CHEN Duohong. Simulation study of three-dimensional debris flow dynamic process based on SPH-DEM-FEM coupling [D]. Yichang:Three Gorges University,2023. (in Chinese with English abstract)] CHEN Duohong. Simulation study of three-dimensional debris flow dynamic process based on SPH-DEM-FEM coupling [D]. Yichang: Three Gorges University, 2023. (in Chinese with English abstract)
[16] 杨帆. 泥石流冲击工程结构动力过程的三维流固耦合数值分析方法[D]. 长沙:中南大学,2022. [YANG Fan. A three-dimensional fluid structure coupling numerical analysis method for the dynamic process of debris flow impact engineering structures [D]. Changsha:Central South University,2022. (in Chinese with English abstract)] YANG Fan. A three-dimensional fluid structure coupling numerical analysis method for the dynamic process of debris flow impact engineering structures [D]. Changsha: Central South University, 2022. (in Chinese with English abstract)
[17] 张福红. 基于GPU加速的SPH方法溃坝水流数值模拟研究[D]. 郑州:华北水利水电大学,2023. [ZHANG Fuhong. Numerical simulation of dam-break flow by SPH method based on GPU acceleration[D]. Zhengzhou:North China University of Water Resources and Electric Power,2023. (in Chinese with English abstract)] ZHANG Fuhong. Numerical simulation of dam-break flow by SPH method based on GPU acceleration[D]. Zhengzhou: North China University of Water Resources and Electric Power, 2023. (in Chinese with English abstract)
[18] DI GREGORIO S,RONGO R,SPATARO W,et al. High performance scientific computing by a parallel cellular environment[J]. Future Generation Computer Systems,1997,12(5):357 − 369. DOI: 10.1016/S0167-739X(96)00023-4
[19] 刘亚敏. 元胞自动机在自然灾害预警和模拟方面的应用研究[D]. 青岛:中国海洋大学,2005. [LIU Yamin. Researching for predicting and simulating natural disaster method based on cellular automaton[D]. Qingdao:Ocean University of China,2005. (in Chinese with English abstract)] LIU Yamin. Researching for predicting and simulating natural disaster method based on cellular automaton[D]. Qingdao: Ocean University of China, 2005. (in Chinese with English abstract)
[20] 赵莉,杨俊,李闯,等. 地理元胞自动机模型研究进展[J]. 地理科学,2016,36(8):1190 − 1196. [ZHAO Li,YANG Jun,LI Chuang,et al. Progress on geographic cellular automata model[J]. Scientia Geographica Sinica,2016,36(8):1190 − 1196. (in Chinese with English abstract)] ZHAO Li, YANG Jun, LI Chuang, et al. Progress on geographic cellular automata model[J]. Scientia Geographica Sinica, 2016, 36(8): 1190 − 1196. (in Chinese with English abstract)
[21] 殷跃平,王文沛,张楠,等. 强震区高位滑坡远程灾害特征研究——以四川茂县新磨滑坡为例[J]. 中国地质,2017,44(5):827 − 841. [YIN Yueping,WANG Wenpei,ZHANG Nan,et al. Long runout geological disaster initiated by the ridge-top rockslide in a strong earthquake area:A case study of the Xinmo landslide in Maoxian County,Sichuan Province[J]. Geology in China,2017,44(5):827 − 841. (in Chinese with English abstract)] DOI: 10.12029/gc20170501 YIN Yueping, WANG Wenpei, ZHANG Nan, et al. Long runout geological disaster initiated by the ridge-top rockslide in a strong earthquake area: A case study of the Xinmo landslide in Maoxian County, Sichuan Province[J]. Geology in China, 2017, 44(5): 827 − 841. (in Chinese with English abstract) DOI: 10.12029/gc20170501
[22] 张永双,成余粮,姚鑫,等. 四川汶川地震-滑坡-泥石流灾害链形成演化过程[J]. 地质通报,2013,32(12):1900 − 1910. [ZHANG Yongshuang,CHENG Yuliang,YAO Xin,et al. The evolution process of Wenchuan earthquake-landslide-debris flow geohazard chain[J]. Geological Bulletin of China,2013,32(12):1900 − 1910. (in Chinese with English abstract)] ZHANG Yongshuang, CHENG Yuliang, YAO Xin, et al. The evolution process of Wenchuan earthquake-landslide-debris flow geohazard chain[J]. Geological Bulletin of China, 2013, 32(12): 1900 − 1910. (in Chinese with English abstract)
[23] YIN Yueping,CHENG Yuliang,LIANG Jingtao,et al. Heavy-rainfall-induced catastrophic rockslide-debris flow at Sanxicun,Dujiangyan,after the Wenchuan Ms 8.0 earthquake[J]. Landslides,2016,13(1):9 − 23. DOI: 10.1007/s10346-015-0554-9
[24] 冯文凯,胡云鹏,谢吉尊,等. 顺层震裂斜坡降雨触发灾变机制及稳定性分析——以三溪村滑坡为例[J]. 岩石力学与工程学报,2016,35(11):2197 − 2207. [FENG Wenkai,HU Yunpeng,XIE Jizun,et al. Disaster mechanism and stability analysis of shattered bedding slopes triggered by rainfall:A case study of Sanxicun landslide[J]. Chinese Journal of Rock Mechanics and Engineering,2016,35(11):2197 − 2207. (in Chinese with English abstract)] FENG Wenkai, HU Yunpeng, XIE Jizun, et al. Disaster mechanism and stability analysis of shattered bedding slopes triggered by rainfall: A case study of Sanxicun landslide[J]. Chinese Journal of Rock Mechanics and Engineering, 2016, 35(11): 2197 − 2207. (in Chinese with English abstract)
[25] 许强,郑光,李为乐,等. 2018年10月和11月金沙江白格两次滑坡-堰塞堵江事件分析研究[J]. 工程地质学报,2018,26(6):1534 − 1551. [XU Qiang,ZHENG Guang,LI Weile,et al. Study on successive landslide damming events of Jinsha River in Baige Village on Octorber 11 and November 3,2018[J]. Journal of Engineering Geology,2018,26(6):1534 − 1551. (in Chinese with English abstract)] XU Qiang, ZHENG Guang, LI Weile, et al. Study on successive landslide damming events of Jinsha River in Baige Village on Octorber 11 and November 3, 2018[J]. Journal of Engineering Geology, 2018, 26(6): 1534 − 1551. (in Chinese with English abstract)
[26] 邓建辉,高云建,余志球,等. 堰塞金沙江上游的白格滑坡形成机制与过程分析[J]. 工程科学与技术,2019,51(1):9 − 16. [DENG Jianhui,GAO Yunjian,YU Zhiqiu,et al. Analysis on the formation mechanism and process of Baige landslides damming the upper reach of Jinsha River,China[J]. Advanced Engineering Sciences,2019,51(1):9 − 16. (in Chinese with English abstract)] DENG Jianhui, GAO Yunjian, YU Zhiqiu, et al. Analysis on the formation mechanism and process of Baige landslides damming the upper reach of Jinsha River, China[J]. Advanced Engineering Sciences, 2019, 51(1): 9 − 16. (in Chinese with English abstract)
[27] 许强,李为乐,董秀军,等. 四川茂县叠溪镇新磨村滑坡特征与成因机制初步研究[J]. 岩石力学与工程学报,2017,36(11):2612 − 2628. [XU Qiang,LI Weile,DONG Xiujun,et al. The Xinmocun landslide on June 24,2017 in Maoxian,Sichuan:characteristics and failure mechanism[J]. Chinese Journal of Rock Mechanics and Engineering,2017,36(11):2612 − 2628. (in Chinese with English abstract)] XU Qiang, LI Weile, DONG Xiujun, et al. The Xinmocun landslide on June 24, 2017 in Maoxian, Sichuan: characteristics and failure mechanism[J]. Chinese Journal of Rock Mechanics and Engineering, 2017, 36(11): 2612 − 2628. (in Chinese with English abstract)
[28] 殷跃平,高少华. 高位远程地质灾害研究:回顾与展望[J]. 中国地质灾害与防治学报,2024,35(1):1 − 18. [YIN Yueping,GAO Shaohua. Research on high-altitude and long-runout rockslides:Review and prospects[J]. The Chinese Journal of Geological Hazard and Control,2024,35(1):1 − 18. (in Chinese with English abstract)] YIN Yueping, GAO Shaohua. Research on high-altitude and long-runout rockslides: Review and prospects[J]. The Chinese Journal of Geological Hazard and Control, 2024, 35(1): 1 − 18. (in Chinese with English abstract)
[29] 杨东旭,游勇,陈晓清,等. 泥石流排导槽磨蚀行为特征研究[J]. 灾害学,2021,36(1):48 − 53. [YANG Dongxu,YOU Yong,CHEN Xiaoqing,et al. Study on the characteristics of abrasion behavior of debris flow drainage channel[J]. Journal of Catastrophology,2021,36(1):48 − 53. (in Chinese with English abstract)] DOI: 10.3969/j.issn.1000-811X.2021.01.010 YANG Dongxu, YOU Yong, CHEN Xiaoqing, et al. Study on the characteristics of abrasion behavior of debris flow drainage channel[J]. Journal of Catastrophology, 2021, 36(1): 48 − 53. (in Chinese with English abstract) DOI: 10.3969/j.issn.1000-811X.2021.01.010
[30] 陈野鹰,唐红梅,陈洪凯. 公路泥石流防治结构冲磨性能优化研究[J]. 中国地质灾害与防治学报,2007,18(2):37 − 41. [CHEN Yeying,TANG Hongmei,CHEN Hongkai. Reinforcement of ablation resistance for debris flow control structures[J]. The Chinese Journal of Geological Hazard and Control,2007,18(2):37 − 41. (in Chinese with English abstract)] DOI: 10.3969/j.issn.1003-8035.2007.02.008 CHEN Yeying, TANG Hongmei, CHEN Hongkai. Reinforcement of ablation resistance for debris flow control structures[J]. The Chinese Journal of Geological Hazard and Control, 2007, 18(2): 37 − 41. (in Chinese with English abstract) DOI: 10.3969/j.issn.1003-8035.2007.02.008
[31] 秦明,陈洪凯,梁丹,等. 泥石流磨蚀特性试验研究[J]. 人民长江,2019,50(8):184 − 187. [QIN Ming,CHEN Hongkai,LIANG Dan,et al. Experimental study on abrasion characteristics of debris flow[J]. Yangtze River,2019,50(8):184 − 187. (in Chinese with English abstract)] QIN Ming, CHEN Hongkai, LIANG Dan, et al. Experimental study on abrasion characteristics of debris flow[J]. Yangtze River, 2019, 50(8): 184 − 187. (in Chinese with English abstract)
[32] 陈洪凯,唐红梅. 泥石流两相冲击力及冲击时间计算方法[J]. 中国公路学报,2006,19(3):19 − 23. [CHEN Hongkai,TANG Hongmei. Method to calculate impact force and impact time of two-phase debris flow[J]. China Journal of Highway and Transport,2006,19(3):19 − 23. (in Chinese with English abstract)] DOI: 10.3321/j.issn:1001-7372.2006.03.004 CHEN Hongkai, TANG Hongmei. Method to calculate impact force and impact time of two-phase debris flow[J]. China Journal of Highway and Transport, 2006, 19(3): 19 − 23. (in Chinese with English abstract) DOI: 10.3321/j.issn:1001-7372.2006.03.004
[33] BANIHABIB M E,IRANPOOR M. Determination of the abrasion of aprons of dams by debris flow[J]. International Journal of Materials and Structural Integrity,2015,9(4):262. DOI: 10.1504/IJMSI.2015.075832
[34] 何胜庆,郑达,张文. 高海拔地区宽级配泥石流冲击拦砂坝试验研究[J]. 水文地质工程地质,2024,51(1):123 − 134. [HE Shengqing,ZHENG Da,ZHANG Wen. Experimental study on the impact of wide graded debris flow on check dam in high altitude area[J]. Hydrogeology & Engineering Geology,2024,51(1):123 − 134. (in Chinese with English abstract)] HE Shengqing, ZHENG Da, ZHANG Wen. Experimental study on the impact of wide graded debris flow on check dam in high altitude area[J]. Hydrogeology & Engineering Geology, 2024, 51(1): 123 − 134. (in Chinese with English abstract)
[35] 胡卸文,韩征,张友谊,等. 强震区沟道型泥石流不同成因物源起动模式及动储量评价方法结题验收报告[R]. 成都:西南交通大学,2022. [HU Xiewen,HAN Zheng,ZHANG Youyi,et al. Final acceptance report of the initiation pattern and evaluation method for dynamic reserves of differently originated source materials of channelized debris flows in strong earthquake area [R]. Chengdu:Southwest Jiaotong University,2022. (in Chinese)] HU Xiewen, HAN Zheng, ZHANG Youyi, et al. Final acceptance report of the initiation pattern and evaluation method for dynamic reserves of differently originated source materials of channelized debris flows in strong earthquake area [R]. Chengdu: Southwest Jiaotong University, 2022. (in Chinese)
[36] 余斌,李为乐,丁明涛,等. 强震区宽缓与窄陡沟道型泥石流致灾机理及灾害链效应验收报告[R]. 成都:成都理工大学,2022. [YU Bin,LI Weile,DING Mingtao,et al Acceptance report on the disaster mechanism and disaster chain effect of wide and narrow steep gully debris flows in strong earthquake areas [R]. Chengdu:Chengdu University of Technology,2022. (in Chinese)] YU Bin, LI Weile, DING Mingtao, et al Acceptance report on the disaster mechanism and disaster chain effect of wide and narrow steep gully debris flows in strong earthquake areas [R]. Chengdu: Chengdu University of Technology, 2022. (in Chinese)
[37] 周超,常鸣,徐璐,等. 强震区沟道堰塞体失稳模式及其动力学特征[J]. 地球科学,2023,48(8):3115 − 3126. [ZHOU Chao,CHANG Ming,XU Lu,et al. Failure modes and dynamic characteristics of the landslide dams in strong earthquake area[J]. Earth Science,2023,48(8):3115 − 3126. (in Chinese with English abstract)] ZHOU Chao, CHANG Ming, XU Lu, et al. Failure modes and dynamic characteristics of the landslide dams in strong earthquake area[J]. Earth Science, 2023, 48(8): 3115 − 3126. (in Chinese with English abstract)
[38] 徐林荣,刘清华,陈洪凯,等. 强震区宽缓与窄陡沟道型泥石流动力学特征结题验收报告[R]. 长沙:中南大学,2022. [XU Linrong,LIU Qinghua,CHEN Hongkai,et al. Acceptance report on the dynamic characteristics of wide and narrow steep gully debris flows in strong earthquake areas [R]. Changsha:Central South University,2022. (in Chinese)] XU Linrong, LIU Qinghua, CHEN Hongkai, et al. Acceptance report on the dynamic characteristics of wide and narrow steep gully debris flows in strong earthquake areas [R]. Changsha: Central South University, 2022. (in Chinese)
[39] 张亮亮,徐林荣,李永威,等. 强震区沟谷型泥石流峰值流量预测研究及应用[J]. 土木与环境工程学报(中英文),2023,45(2):81 − 88. [ZHANG Liangliang,XU Linrong,LI Yongwei,et al. Prediction of the gully debris flow peak discharge prediction in strong earthquake area[J]. Journal of Civil and Environmental Engineering,2023,45(2):81 − 88. (in Chinese with English abstract)] DOI: 10.11835/j.issn.2096-6717.2022.009 ZHANG Liangliang, XU Linrong, LI Yongwei, et al. Prediction of the gully debris flow peak discharge prediction in strong earthquake area[J]. Journal of Civil and Environmental Engineering, 2023, 45(2): 81 − 88. (in Chinese with English abstract) DOI: 10.11835/j.issn.2096-6717.2022.009
[40] 王文沛,罗刚,蒙明辉,等. 强震区高位滑坡型泥石流运动机理模拟及新型拦挡技术验收报告[R]. 北京:中国地质环境监测院,2022. [WANG Wenpei,LUO Gang,MENG Minghui,et al. Simulation of high-level landslide type debris flow movement mechanism in strong earthquake areas and acceptance report of new interception technology [R]. Beijing:China Institute of Geo-Environment Monitoring,2022. (in Chinese)] WANG Wenpei, LUO Gang, MENG Minghui, et al. Simulation of high-level landslide type debris flow movement mechanism in strong earthquake areas and acceptance report of new interception technology [R]. Beijing: China Institute of Geo-Environment Monitoring, 2022. (in Chinese)
[41] 王文沛,殷跃平,胡卸文,等. 碎屑流冲击下桩梁组合结构拦挡效果及受力特征研究[J]. 地质力学学报,2022,28(6):1081 − 1089. [WANG Wenpei,YIN Yueping,HU Xiewen,et al. Effectiveness and mechanical characteristics of a pile-beam composite structure in blocking debris flows[J]. Journal of Geomechanics,2022,28(6):1081 − 1089. (in Chinese with English abstract)] DOI: 10.12090/j.issn.1006-6616.20222834 WANG Wenpei, YIN Yueping, HU Xiewen, et al. Effectiveness and mechanical characteristics of a pile-beam composite structure in blocking debris flows[J]. Journal of Geomechanics, 2022, 28(6): 1081 − 1089. (in Chinese with English abstract) DOI: 10.12090/j.issn.1006-6616.20222834
[42] 陈洪凯,赵世春. 强震区宽缓与窄陡沟道型泥石流综合防控技术验收报告[R]. 枣庄:枣庄学院,2022. [CHEN Hongkai,ZHAO Shichun. Acceptance report on comprehensive prevention and control technology for wide and narrow steep gully debris flows in strong earthquake areas [R]. Zaozhuang:Zaozhuang College,2022. (in Chinese)] CHEN Hongkai, ZHAO Shichun. Acceptance report on comprehensive prevention and control technology for wide and narrow steep gully debris flows in strong earthquake areas [R]. Zaozhuang: Zaozhuang College, 2022. (in Chinese)
[43] 赵松江,苗晓岐,高延超,等. 强震区特大型泥石流防控标准化体系及示范应用验收报告[R]. 成都:四川省华地建设工程有限责任公司,2022. [ZHAO Songjiang,MIAO Xiaoqi,GAO Yanchao,et al. Acceptance report on the standardization system and demonstration application of large scale debris flow prevention and control in strong earthquake areas [R]. Chengdu:Sichuan Huadi Construction Engineering Co. Ltd. ,2022. (in Chinese)] ZHAO Songjiang, MIAO Xiaoqi, GAO Yanchao, et al. Acceptance report on the standardization system and demonstration application of large scale debris flow prevention and control in strong earthquake areas [R]. Chengdu: Sichuan Huadi Construction Engineering Co. Ltd. , 2022. (in Chinese)
[44] 郭剑,李天涛,孙金坤. 安县高川河流域泥石流物源特征及其活动量预测[J]. 水电能源科学,2015,33(7):151 − 155. [GUO Jian,LI Tiantao,SUN Jinkun. Characteristics of debris flow provenance and activity prediction of Gaochuan River Basin in Anxian County[J]. Water Resources and Power,2015,33(7):151 − 155. (in Chinese with English abstract)] GUO Jian, LI Tiantao, SUN Jinkun. Characteristics of debris flow provenance and activity prediction of Gaochuan River Basin in Anxian County[J]. Water Resources and Power, 2015, 33(7): 151 − 155. (in Chinese with English abstract)
[45] 刘波,胡卸文,何坤. 汶川强震区震裂山体分布特征与成灾模式研究[J]. 灾害学,2024,39(1):177 − 187. [LIU Bo,HU Xiewen,HE Kun. Study on the distribution characteristics and disaster mode of earthquake fracture mountains in Wenchuan strong earthquake area[J]. Catastrophology,2024,39(1):177 − 187. (in Chinese with English abstract)] LIU Bo, HU Xiewen, HE Kun. Study on the distribution characteristics and disaster mode of earthquake fracture mountains in Wenchuan strong earthquake area[J]. Catastrophology, 2024, 39(1): 177 − 187. (in Chinese with English abstract)
[46] 赵松江,胡卸文,余斌,等. 强震区泥石流防治工程 勘查规范:T/CI 055—2023[S]. 北京:中国国际科技促进会,2023. [ZHAO Songjiang, HU Xiewen, YU Bin, et al. Specification of prevention and control engineering for debris flow in meizoseismal area:T/CI 055- 2023[S]. Beijing:China International Association for the Promotion of Science and Technology, 2023. (in Chinese)] ZHAO Songjiang, HU Xiewen, YU Bin, et al. Specification of prevention and control engineering for debris flow in meizoseismal area: T/CI 055- 2023[S]. Beijing: China International Association for the Promotion of Science and Technology, 2023. (in Chinese)
-
期刊类型引用(7)
1. 刘琛尧,晏启祥,孙润方,王绪,邓宝华,陈耀. 基于三维离散-连续耦合的岩溶隧道突水破坏模式研究. 水文地质工程地质. 2024(02): 163-171 . 百度学术
2. 盖孝乾. 隐伏溶洞对隧道围岩应力及地表沉降的影响研究. 工程机械与维修. 2023(01): 67-69 . 百度学术
3. 郭书兰,阎长虹,俞良晨,闫超,李慧,徐源. 无锡浅埋岩溶发育特征及其与隧道安全距离研究. 南京大学学报(自然科学). 2023(05): 890-899 . 百度学术
4. 孙伟. 广东省英德市城南社区岩溶塌陷发育特征及成因分析. 中国地质灾害与防治学报. 2023(05): 74-80 . 本站查看
5. 朱丹晖,聂浩. 大溶腔小跨度公路隧道溶洞处理方案研究. 铁道勘察. 2023(06): 144-150+162 . 百度学术
6. 杨刚. 岩溶地段隧道施工安全管理关键措施研究. 价值工程. 2022(32): 28-30 . 百度学术
7. 王亚强,王先龙,伍尚前,张建博. 隧道岩溶水引排及防治措施研究. 现代隧道技术. 2022(S1): 866-875 . 百度学术
其他类型引用(3)