Topographic Amplification at the Eastern Anatolia Observatory Site: A Comparative Analysis

This study investigates topographic site amplification at the Eastern Anatolia Observatory (DAG) site. It uses data from the 16 September 2022 Çat earthquake of magnitude Mw 4.9. The event was recorded by internationally registered seismic stations, ETEKE and EKONA. ETEKE is at the mountain base, while EKONA is approximately 55 m below the summit. We apply two methods to estimate frequency‑dependent amplification. These are the Standard Spectral Ratio (SSR) and the Horizontal‑to‑Vertical Spectral Ratio (HVSR). The SSR compares spectra at the target site to a reference rock site and provides absolute amplification values. The HVSR method provides a single-station estimate of the fundamental resonant frequency. We also compare the results with two different seismic codes, which provide provisions for topographic amplification. Our results reveal a clear topographic resonance at 0.85 Hz. The SSR amplification factors at this frequency are 3.35 (E), 4.52 (N), with a geometric mean (GM) of 3.89. The HVSR analysis also identifies the same 0.85 Hz peak. While the two methods are not independent, the absence of the 0.85 Hz peak at ETEKE and consistency with response spectra support the topographic amplification. The pseudo-acceleration response spectra (PSA) ratio at T = 1.18 s is 1.80. This value exceeds the maximum topographic amplification factors in the seismic codes by 28%. While these results are derived from a single earthquake, the absence of the peak at the base station, the response-spectra consistency, and an independent ambient-noise H/V measurement supports the findings. However, the findings should be considered preliminary and require validation with future earthquake recordings. A moving-block bootstrap analysis yields uncertainty estimates that support the statistical significance of the results. The findings highlight the potential importance of site-specific, frequency-dependent topographic amplification for infrastructure at steep terrain. However, because the analysis is based on a single event at one azimuth, the comparison with the codes should be treated as preliminary, requiring validation with additional earthquake recordings.