采用IDS方法反演强震数据确定2017年8月8日九寨沟地震的破裂过程

近场强震数据是地震观测的主要数据类型之一，但尚未广泛用于震源参数快速和常规测定.作者利用2017年8月8日九寨沟地震的近场强震数据，分别基于中国地震台网中心和美国地质调查局的定位结果，采用迭代反褶积和叠加（Iterative Deconvolution and Stacking， IDS）方法反演得到了两个破裂模型.反演结果显示，两个破裂模型除在时间上存在大约2.5 s的偏移外，在其他破裂特征上都极其相似，表明强震数据反演并不严重依赖地震定位结果.破裂过程大约持续8~10 s，释放的地震矩在6.9×10¹⁸ Nm左右，对应的矩震级约为M_W6.5.走向方向上，破裂覆盖了震中西北15 km至震中东南10 km的区域，整体表现为不对称的双侧破裂模式，其中西北方向的破裂略占优势.深度方向上，主要破裂集中在0~10 km的较浅区域.这些特征与目前的地震学和大地测量学结果基本一致，表明在当前的数据条件下，即使定位结果存在不确定性，通过反演强震数据也能够独立可靠地确定中强地震的破裂过程.特别地，未来强震数据的数据质量和获取效率得到进一步改进和提升后，这一工作有望成为震后快速确定破裂过程的主要方式之一，在震后应急工作中发挥积极作用.

Estimating the rupture process of the 8 August 2017 Jiuzhaigou earthquake by inverting strong-motion data with IDS method

Near-field strong-motion data are one of the major datasets which well constrain the earthquake source parameters, but have not been widely applied for fast and routine determination purposes. Based on the near-field strong-motion data of the 8 August 2017 Jiuzhaigou earthquake, and the earthquake locations released by China Earthquake Networks Center (CENC) and United States Geological Survey (USGS), we determined two rupture models by using the iterative deconvolution and stacking (IDS) method. Though the two locations are significantly different (ΔH/ΔZ/ΔT=3.8 km/11 km/2.8 s), the obtained two rupture models are quite similar with each other, despite a 2.5 s time shift. This indicates that the rupture model estimated from near-field strong-motion data inversion does not closely depend on the earthquake location.The rupture lasts for about 8~10 s, releasing a scalar moment about 6.9×10¹⁸Nm, equivalent to M_W6.5. Along the strike, the northwestward and southeastward rupture extensions are about 15 km and 10 km, respectively, suggesting a slightly asymmetric bilateral rupture event. Along the down-dip direction, the slip area mainly distributed at a shallow depths ranging from 0 to 10 km. These source characteristics are basically consistent with other existing seismic and geodetic models, implying that the rupture models of moderate-to-strong earthquakes can be well determined by inverting near-field strong-motion seismograms even if uncertainties exist in earthquake locations. Especially, once the acquisition efficiency and quality of the strong-motion data are further improved in the future, the strong-motion data inversion would become another major way to fast estimate the source ruptures, and would play an important role in earthquake emergency responses.