山岳地区强烈地震诱发的滑坡、崩塌等地震次生灾害在造成严重人员伤亡的同时, 对社会经济的发展也构成了严重威胁.而震后滑坡分布的快速评估, 尤其是滑坡重灾区的确定, 则可为救援工作的科学部署和有效开展提供决策支持并减轻地震灾害的损失.2017年8月8日发生于四川省阿坝州九寨沟风景区的MS7.0级地震诱发了大量的滑坡、崩塌灾害, 造成了一定的人员伤亡和财产损失, 凸显出进行震后滑坡快速评估的重要性.本文通过对九寨沟地震灾区震前、震后"北京二号"遥感影像的对比分析, 解译此次地震诱发的滑坡分布概况.尽管震后影像在局部区域有云团覆盖, 影响了震区滑坡解译的详尽程度, 但是对于主要发生的滑坡地震高烈度地区 (≥Ⅻ度) , 本文所用影像基本满足解译要求.本文共解译出194个面积大于700m2的滑坡, 这些滑坡主要沿地震烈度长轴方向分布, 灾害体平面面积达5.6km2, 影响范围超过600km2.通过对九寨沟地震灾区的地形、岩性及地震动加速度进行分析, 本文采用Newmark刚体滑块模型对该区震后滑坡危险区域进行了预测.预测结果按照危险程度不同划分为5个级别, 即高度危险、较高危险、中度危险、较低危险和低度危险.滑坡分布与评估结果呈现出较好的一致性:解译的滑坡主要分布在评估为滑坡高度危险的区域, 表明本文所采用方法的有效性.本文对该方法的局限性也进行了讨论, 并提出改进建议.%Landslides triggered by strong earthquakes in mountainous areas can cause severe casualties and also pose threats to socio-economic development in the affected region.Therefore, an effective evaluation model, not only for a quick assessment of landslide distribution, but also for identification of most hazardous areas, is needed to provide critical assistance in the deployment and execution of rescue efforts.On August8, 2017, aMS7.0 earthquake struck Jiuzhaigou County, Aba Prefecture, Sichuan Province.The earthquake triggered massive landslides causing casualties and property loss.After this event, we quickly analyzed the pre-and post-earthquake remote sensing images (produced by the Beijing No.2 satellite) in the affected region and interpreted the landslide distribution.Although, in some areas, the post-seismic images were unable to provide detailed interpretation of landslide due to cloud coverage, they met the primary requirements for landslide interpretation at locations of high seismic intensity where landslides mainly occurred.In total, we interpreted 194 landslides with individual area>700 m2.These landslides were mainly distributed along the long axis of seismic intensity, with an accumulative disaster area of 5.6 km2 and impact range exceeding 600 km2.Based on topography and lithology analyses of affected region, as well as peak ground acceleration data, we used the Newmark rigid block model to predict the landslide hazardous area after the earthquake.The prediction results were divided into five levels according to degree of potential danger, i.e.high, moderately high, moderate, low and very low.Comparison study shows that the predicted landslide hazard zones agreed with the actual distribution of earthquake-triggered landslides that interpreted landslides mainly distributed in areas evaluated as high danger zone, validating the effectiveness of using the Newmark rigid block model to quickly evaluate earthquake-triggered landslides.In addition, we also discussed limitations and made suggestions for future improvement in the model.
展开▼
