Physics-based broadband ground-motion simulations are generated for the strong mainshocks that occurred in the region of the Central Ionian Islands, on 26th January 2014 in Kefalonia (M_w6.1) and 17th November 2015 in Lefkas (M_w6.5). The study area is associated with frequent strong earthquakes both in the historical and instrumental era. During the last decades the network of strong-motion accelerographs in the area has been densified, and thus provided an adequate number of strong ground motion records as a means to better examine the related ground motion characteristics. In the present study, broadband ground motions for the two case studies are simulated both at selected sites and at a dense grid of points covering the affected areas. The low-frequency part of the synthetics is computed using a discrete wavenumber finite element method by convolving Green’s functions with a kinematic slip model in the frequency domain. A stochastic finite fault model approach based on a dynamic corner frequency is considered to calculate the ground motions for the higher frequencies. The broadband synthetic time series are generated after merging the results obtained from the two separate techniques, by performing a weighted summation at intermediate frequencies. The simulated values are validated by comparison with both recorded Peak Ground Acceleration (PGA) and Peak Ground Velocity (PGV) values and the estimated ones by using widely accepted Ground Motion Prediction Equations (GMPEs). Our results indicate that both the spatial distribution and the amplification pattern of the simulated ground motions, in the near-field, in terms of PGA and PGV are highly influenced by the slip heterogeneity and the maximum slip patches within the seismic source.

2014 年 1 月 26 日在凯法利尼亚 (M _w 6.1) 和 2015 年 11 月 17 日在莱夫卡斯 (M _w6.5)。研究区与历史和仪器时代的频繁强震有关。在过去的几十年里，该地区的强震加速度计网络已经变得更加密集，因此提供了足够数量的强地震动记录，作为更好地检查相关地震动特征的手段。在本研究中，两个案例研究的宽带地面运动在选定地点和覆盖受影响区域的密集点网格中进行了模拟。合成的低频部分是使用离散波数有限元方法通过将格林函数与频域中的运动滑移模型进行卷积来计算的。考虑基于动态角频率的随机有限故障模型方法来计算较高频率的地面运动。通过在中频执行加权求和，合并从两种单独技术获得的结果后生成宽带合成时间序列。模拟值通过与记录的峰值地面加速度 (PGA) 和峰值地面速度 (PGV) 值以及使用广泛接受的地面运动预测方程 (GMPE) 的估计值进行比较来验证。我们的结果表明，在近场中，就 PGA 和 PGV 而言，模拟地震动的空间分布和放大模式都受到震源内滑动非均质性和最大滑动斑块的高度影响。通过在中频执行加权求和。模拟值通过与记录的峰值地面加速度 (PGA) 和峰值地面速度 (PGV) 值以及使用广泛接受的地面运动预测方程 (GMPE) 的估计值进行比较来验证。我们的结果表明，在近场中，就 PGA 和 PGV 而言，模拟地震动的空间分布和放大模式都受到震源内滑动非均质性和最大滑动斑块的高度影响。通过在中频执行加权求和。模拟值通过与记录的峰值地面加速度 (PGA) 和峰值地面速度 (PGV) 值以及使用广泛接受的地面运动预测方程 (GMPE) 的估计值进行比较来验证。我们的结果表明，在近场中，就 PGA 和 PGV 而言，模拟地震动的空间分布和放大模式都受到震源内滑动非均质性和最大滑动斑块的高度影响。