2017年8月8日四川九寨沟地震诱发地质灾害空间分布规律及易发性评价初步研究

A PRELIMINARY STUDY ON SPATIAL DISTRIBUTION PATTERNS OF LANDSLIDES TRIGGERED BY JIUZHAIGOU EARTHQUAKE IN SICHUAN ON AUGUST 8TH,2017 AND THEIR SUSCEPTIBILITY ASSESSMENT

2017年8月8日四川省九寨沟发生7.0级地震,是继2008年汶川大地震和2013年芦山地震之后,四川省发生的又一次7.0级以上的强震。为了给现场救援和震后地质灾害防治提供科学依据,本文作者第一时间赶赴现场,并基于震前和震后高精度遥感影像,完成了震区地质灾害解译和复核工作,共解译地质灾害1883处,主要以中小型浅层滑坡和崩塌为主。基于解译结果,对同震灾害的空间分布规律和控制因素(距断层距离、地面峰值加速度PGA、高程、坡度和坡向等)进行了分析,研究表明地质灾害主要沿北西—南东向呈带状分布,且沿公路、沟谷较为发育,在野外推测发震断层2km范围内高度集中,呈现明显的断层效应,但与塔藏断层之间的空间相关性相对较弱。在上述分析的基础上,采用逻辑回归模型,利用地震、地形和地质3大因素(8个因子)对地质灾害易发性进行了快速评价和分区,经统计校验证明该模型的准确率达0.851,模型精度较高。

A magnitude 7.0 earthquake struck Jiuzhaigou on August 8th, 2017 in Sichuan province, which is another strong earthquake happened in Sichuan province after the 2008 great Wenchuan earthquake and the 2013 Lushan earthquake. In order to provide scientific evidence for on-the-spot rescue work and geologic hazard prevention after the earthquake, the author arrived quickly on the field and finished geologic hazard interpretation and review work in the earth-stricken areas based on high resolution remote sensing images before and after the earthquake. As a result, 1833 earthquake-induced geohazards have been obtained in the study area of 840km2, which are mainly medium-scale and small-scale shallow landslides and collapses. In terms of the results of the interpretation, the spatial distribution patterns and control factors in the cosesmic geohazards （ distance from the fault, PGA, elevation, slope, aspect etc.） are analyzed, the research shows that the geohazards triggered by earthquake have the features of zonal distribution mainly distributed along north-west and east-south, especially along the roads and valleys. The width of strong development zone along the field inferred seismogenic fault is about 2km, showing a distinct faulting effect, but the spatial correlation betweem geohazards and Tazang fault is weak. Based on the above-described analysis, logistic regression model is taken quick comments and divisions of geologic hazard susceptibility are made. After statists and testing, the accuracy of this model has reached 0. 851, a comparatively high precise data.