ESTIMATION OF RELATIVE AND ABSOLUTE TIME DELAY USING A MODIFIED ADAPTIVE STACKING METHOD

Abstract

Accurate estimation of seismic wave arrival times plays a fundamental role in earthquake analysis, seismic tomography, and lithospheric studies. However, waveform misalignment caused by station variability, subsurface heterogeneity, and ambient noise remains a challenge, especially in tectonically complex regions such as eastern Indonesia. To overcome this limitation, a modified adaptive stacking method was applied for the estimation of both relative and absolute P-wave delay times. The approach utilizes semi-automatic P-wave picking, cross-correlation with a high signal-to-noise reference trace, linear time shifting, and iterative stacking refinement. The reference trace is updated at each iteration using the previous stack to progressively enhance alignment. The method was applied to four earthquake events two teleseismic (Mw 7.9) and two regional (Mw 4.9) recorded by 21 broadband stations in the Flores region. Results indicate improved waveform coherence and higher cross-correlation coefficients across iterations, particularly in teleseismic data. Travel-time residuals, derived by comparing observed arrival times with IASP91 theoretical predictions, reveal spatial variations in travel-time residuals that may reflect underlying structural complexity, potentially influenced by variations in lithospheric properties, such as changes in composition, temperature, or the presence of subducted material. The findings demonstrate that the modified adaptive stacking method enhances phase alignment and is suitable for use in structurally complex seismic environments.