Improve Signal-to-Noise Ratio by Varying Daily Truncation Segmentation of Cross-Correlation Data for Ambient Noise Tomography Study

Abstract Ambient seismic noise tomography (ANT) has been the most discussed topic in the last 20 years. The Idea of ANT comes from interferometry cross correlation (CC) between pairs of seismographs. Biases are mitigated from the CC result by determine the signal to noise ratio (SNR). The process is necessary for cutoff of useable CC. Each CC that doesn’t meet the SNR cutoff are removed, as they typically introduce artifacts into the group velocity dispersion curve, which later used to determine the group velocity maps. SNR commonly measure by comparing the amplitude of signal relative to rms noise from stack result of CC of several day, month or even years. Recent process including segmentation in a day of record for CC. the segmentation of CC varies from 30 minute up to 24 hours. This study tries to compare the SNR from 5 length of segment using two permanent stations in Java Island as part of the tomography crustal study in the Java Island. The Pairs station refences are located at West Java. Segmentation length is 30-minute, 1 hour, 2-hour, 3 hour, and 4 hours. Our result, 30 minutes of CC showing increment of SNR 4.5 times higher compared to 4-hour segmentation. By this result we could reduce the amount of removed data by improve the SNR with smaller segmentation of CC.