On December 18, 2023, an Ms6.2 earthquake struck Jishishan County, Gansu Province, in western China. The China Earthquake Early Warning Network (CEEWN) captured extensive near-field ground motion data using high-density microelectromechanical system (MEMS) sensors and force-balanced accelerographs (FBAs). Through noise level and usable frequency range assessments of MEMS/FBA recordings, we compiled a strong- motion dataset encompassing the Ms6.2 mainshock and 13 aftershocks (Ms >= 3.0). Analysis of this dataset revealed distinct source characteristics and site effects through spatial distributions and attenuation patterns of peak ground acceleration (PGA, up to 1.1 g at station N002B), peak ground velocity (PGV), and spectral accelerations (SAs) across various periods. The mainshock's near-fault motions exhibited pronounced short-period energy, with 0.2 s SAs exceeding 1.0 gin intensity zones VII-VIII due to hanging wall effects, soil amplification, and topographic influences. Site-to-reference ratio (SSR) analysis identified site nonlinearity above 1 Hz and amplification between 1 and 10 Hz. Observed PGAs and short-period SAs surpassed ground motion model (GMM) predictions with faster attenuation rates, while long-period SAs (>1.0 s) remained below predictions. Residual analysis of intensity measures (IMs) and horizontal-to-vertical spectral ratios (HVSRs) demonstrated progressive site nonlinearity, showing HVSR frequency reductions and amplitude declines at PGAs >500 cm/s(2). This dataset advances regional ground motion model (GMM) development, while our findings on strong ground motion characteristics offer critical insights for earthquake damage assessment and post-disaster reconstruction.

来源平台：SOIL DYNAMICS AND EARTHQUAKE ENGINEERING