Ambient Noise Analyses at Broadband Ocean-Bottom Seismometers: Data Quality and Transfer Function Corrections

Abstract Seismic ambient noise tomography has revolutionized regional crustal and upper-mantle imaging for broadband ocean-bottom seismometer (BBOBS) studies. However, the ocean poses unique challenges including tilt and compliance noise, and interface waves with considerable water-column sensitivity. Transfer function (TF) noise removal techniques enable separation of fundamental Scholte waves from overtone Rayleigh waves, but such noise removal strategies do not always provide the expected signal improvement. Here, we leverage data from recent U.S.-led BBOBS deployments to evaluate variability in ambient-noise surface-wave quality and efficacy of TF corrections across different oceanic environments and instrumentation. We examine the 15–30 and 3–10 s period bands that are typically used in lithospheric imaging. We find that water depth, sediment thickness, instrument design, and ocean basin characteristics are the primary factors influencing ambient-noise data quality and effectiveness of TF corrections. In the 3–10 s period band, enhancement of higher-mode Rayleigh waves is most successful in deep waters with thin sediments. In the 15–30 s period band, signal is improved both in shallow waters likely due to infragravity wave removal, and in instances of high tilt noise. Our results provide new insights into the environmental conditions and instrument designs that shape ambient noise tomography performance, offering practical guidance for optimizing future BBOBS studies as well as novel seafloor-sensing technologies.