北京平原区夏垫断裂的多源遥感影像特征

Characteristics of Multiple Remote Sensing Images of Xiadian Fault,Beijing Plain

1679年9月2日三河—平谷8.0级地震在地表形成总体走向N40°～45°;E,倾角60°～70°;,长达10 km的破裂带。该文采用Landsat－8、SPOT－5、SAR、GDEM、Google Earth多源遥感影像研究夏垫断裂的遥感影像特征。利用ERDAS软件,对Landsat－8影像按7∶4∶2依次为R、G、B分量作彩色合成增强处理,突出影像的线性构造特征；对Landsat－8和SPOT－5影像、Landsat－8和SAR影像分别进行融合处理,丰富影像的光谱信息；对GDEM与SPOT影像进行三维叠加,突出影像的三维特征；对SAR影像进行斑点压缩、方向滤波处理,突出影像平行断裂方向的线性特征。文中分析、对比了 Google Earth 和 SPOT－5影像上地震陡坎的特征,陡坎在SPOT－5影像上表现较明显,但在影像上表现不出较低陡坎,同时,在两种影像上容易将绿色植被、狭窄道路或其它地物误认为陡坎；SAR滤波后的影像表现出暗色的线性条带；Landsat－8和SAR融合后的影像表现出较宽的异常条带,该条带指示断裂带所在位置。

Sanhe-Pinggu Ms8. 0 earthquake on September 2nd, 1679 forms a whole trend N40°～ 45°;E, dip angle 60°～ 70°;, 10 km-long rupture zone at the surface. This paper analyzes image characteristics of Xiadian Fault through multiple remote sensing images, such as Landsat-8, SPOT-5, SAR, GDEM and Google Earth. By ERDAS, we use color enhanced processing of Landsat-8 image, R：G：B following 7：4：2 as color component,to stress linear structural features;we make a fusion of Landsat-8 and SPOT-5 images, Landsat-8 and SAR images, respectively, to enrich spectral information. For GDEM and SPOT images, we use three-dimensional stack to stress three-dimensional features. For the SAR image, we use speckle compression and direction filtering enhancement to stress linear features of paralleling fracture direction. After processing, we analyze and compare the earthquake scarp features on Google Earth and SPOT-5 images. Scarps appear on SPOT-5 image significantly, but lower scarps have no obvious features. Meanwhile, it is easy to consider green vegetation, narrow road or other surface features as scarps on images. After filtered, SAR image shows a dark linear strip. Fusion image of SAR and Landsat-8 images show a wide abnormal band, indicating the strip fault location.