高分辨率光学遥感影像地震裂缝特征提取——以昆仑山口西地震为例

Earthquake fissure feature extraction from high resolution optical remote sensing image:taking West Kunlun Mountain earthquake as an example

地震裂缝作为同震地表破裂常见的表现形式之一,会造成房屋等建筑物的开裂倒塌,威胁人民生命财产安全。查明地震裂缝分布规律,有助于深入理解震源断层性质、构造运动状态和地震破裂过程等基础信息。区别于采动地裂缝、滑坡地裂缝,震源断层错动形成的地震裂缝与发震断裂带位置基本重合,具有明显的光谱特征和空间特征。选取昆仑山口西地震同震地表破裂带为研究区,以国产高分二号作为数据源,提出一种基于光谱特征和空间特征的地震裂缝提取方法,并与改进的边缘检测算法、高斯匹配滤波算法进行比较,结果表明该方法结果最优,高斯匹配滤波算法次之,改进的边缘检测算法效果较差。结合里德尔剪切破裂模型分析地震裂缝空间分布及与其他破裂类型的组合关系,以期在震后第一时间掌握地表破裂空间分布规律,确定主破裂带方向和区域应力场,指导地表破裂地面详细调查、灾后应急救援以及地震科学相关研究。

As one of the common forms of coseismic surface rupture,earthquake fissures will cause cracking and collapse of buildings,threatening the safety of people’s lives and property.To find out the distribution law of earthquake fissures is helpful for understanding the basic information such as properties of seismogenic fault,state of tectonic movement,and earthquake rupture process.Different from mining-induced fractures and landslide fissures,earthquake fissures formed by seismogenic fault dislocation basically coincide with the seismogenic fault zone and have obvious spectral and spatial characteristics.Taking the West Kunlun Mountain earthquake surface rupture zone as the research area,and using the GF-2 as the data source,this paper proposes a seismic fracture extraction method based on spectral features and spatial features,and compares it with the noval edge detection algorithm and Gaussian matched filter algorithm.The results show that the proposed method is the best,followed by Gaussian matched filter algorithm,the result of the noval edge detection algorithm is poor.Combined with Riedel shear pattern,the spatial distribution of earthquake fissures and the combination relationship with other rupture types are analyzed in order to grasp the spatial distribution law of surface ruptures,determine the direction of main fracture zone and regional stress field,and guide the detailed investigation of surface ruptures,post disaster emergency rescue and research related to earthquake science.