Comparisons of Source Parameter Estimates Based on Moment Tensor and Coda Analysis Methods in the Sea of Marmara, NW Türkiye

Reliable, stable, and internally consistent moment magnitude (M_w) estimates are essential for seismic hazard assessment and earthquake source characterization. Full waveform inversion provides physically robust seismic moment estimates; however, its applicability is often constrained by data quality, network geometry, and modeling assumptions. Conversely, coda-based approaches are a more efficient alternative that are less sensitive to radiation pattern and path effects. This study presents a systematic comparison between coda-derived moment magnitudes (M_w-coda) estimated using the Qopen framework and those (M_w-ISOLA) obtained from moment tensor inversion conducted through the ISOLA software.

Analysis is based on the same dataset examined by Özkan et al. (2024), comprising 303 local earthquakes (2.5 ≤ M_L ≤ 5.7) recorded by 49 broadband stations in the Sea of Marmara region, NW Türkiye. Qopen is utilized to jointly invert S-wave and coda-wave envelopes by employing Radiative Transfer Theory. This process yields frequency-dependent attenuation parameters and coda-derived source spectra, from which the M_w-coda is estimated. For a subset of these events with sufficient waveform quality and azimuthal coverage, independent seismic moment estimates are obtained through time-domain waveform inversion using ISOLA.

Overall, we aim to evaluate the agreement and consistency between M_w-coda and M_w-ISOLA in terms of magnitude scaling, scatter, and potential systematic bias, with particular emphasis on magnitude range, source depth, and signal-to-noise conditions in a tectonically complex region characterized by strong lateral heterogeneity and frequency-dependent attenuation.