Search

We report on the extraction of deep ocean travel time variations from time‐lapse cross‐correlations between a hydrophone station and a three‐component broadband seismometer. The signals we cross‐correlate in this study result from repeated activity by the Monowai seamount, one of the most active submarine volcanoes of the Tonga‐Kermadec ridge. In particular, we introduce a specific workflow to exploit repetitive hydroacoustic underwater source activity, which we detail to such an extent that it serves as an example (or "cookbook"). For this reason, we have made the source code publicly available. The workflow proposed in this study (a) overcomes differences in instrument sensitivity and sample rates, (b) involves the selection of eligible cross‐correlations based on a source activity criterium as well as slowness analysis, and (c) extracts the travel time variations in distinct frequency bands. In our case, the two frequency bands are 3–6 and 6–12 Hz. We find that the estimated travel time variations in both frequency bands consist of a complex periodic pattern superimposed on a robust linear trend. This linear trend is decreasing, which we attribute to increasing water temperatures along the propagation path of the hydroacoustic signals. Plain Language Summary: We extract variations in the ocean travel time between submerged microphones and a seismometer on an island. The signals in this study result from repeated activity by the Monowai seamount, one of the most active submarine volcanoes of the Tonga‐Kermadec ridge. The travel time variations result from a time‐lapse similarity analysis. In particular, we introduce a specific workflow to exploit repetitive underwater source activity. We detail it to such an extent that it serves as an example (or "cookbook"). For this reason, we have made the source code publicly available. The workflow proposed in this study (a) overcomes discrepancies in instrument sensitivity and sampling rates, (b) involves the selection of qualified signals based on a source activity criterium as well as directional analysis, and (c) extracts the travel time variations in distinct frequency bands. The estimated travel time variations consist of a complex periodic pattern superimposed on a robust linear trend. Moreover, this linear trend is decreasing, which we attribute to increasing water temperatures along the propagation path of the hydroacoustic signals. Key Points: Travel time variations extracted from time‐lapse cross‐correlations between a hydrophone station and a three‐component broadband seismometerFrequency‐dependent acoustic travel time serves as path‐averaged temperature changes in the Sound Fixing and Ranging channelSelecting eligible cross‐correlations by source activity and slowness estimation is crucial for extracting reliable oceanic travel times [ABSTRACT FROM AUTHOR]