Summary The ocean is the primary source of seismic ambient noise. Therefore, seismic recordings at seafloor stations should reveal noise characteristics more directly than land stations. However, due to a lack of broadband seismic instrumentation, seafloor noise studies using seafloor stations have been inadequate compared to land-based instrumentation. In this study, we use seismic data collected at the South China Sea (SCS) seafloor by newly developed Ocean Bottom Seismographs (OBSs) to analyze the ambient noise features in this marginal sea. The broadband OBS, dubbed ‘Pankun’, has unique shielding to isolate its sensor from the influences of bottom currents. A side-by-side land test between the OBS sensor unit and a standalone seismometer showed that the self-noise caused by the gimbal and the pressure case is insignificant. The recordings on the SCS seafloor have distinct noise spectra. The Double Frequency Microseisms (DFMs) have a single instead of double peak like that seen for Pacific stations. The peak appears in a lower period range (1–5 s) than in the global noise model, indicating that the primary source region for the DFM is the SCS itself. The high-frequency content of the DFM is attenuated more as it propagates from its source region (seafloor) to land stations. The Single Frequency Microseism (SFM) peak on the spectrum is weak, reflecting that SFMs, generated in shallow water along the coast, have difficulties propagating back into the deep ocean due to the substantial increase in seafloor depth. A long-period Earth's hum signal is also identifiable on the vertical component at periods greater than 50 s, probably due to the anti-current design of the OBS. Although the seasonal sea state mainly affects the noise level, extreme events such as typhoons can produce short-term abnormally high DFMs in the basin. However, the DFM highs caused by such events exhibit complex patterns, depending on the wind speed, duration, and area covered by the events.

The ocean‐bottom currents are believed to be largely responsible for the high noise level in ocean bottom seismograph (OBS) data, in particular on the horizontal components. Due to the lack of in‐situ experiments and measurements, the generation mechanism and characteristics of the current‐induced noise are still poorly understood. In this paper, we designed an experiment to explore the features of current‐induced noise. A sensor module from a typical passive‐source OBS was installed in a water flume that can produce controllable steady water flows. We measured and analyzed the recorded noise with changing current velocities, with and without shielding. With other noise sources, such as infragravity waves precluded, this experiment demonstrates that the currents can generate low‐frequency noise, particularly <0.1 Hz. The noise level depends on the current velocity as well as the frequency. While the current‐induced noise affects the horizontal components significantly, its impacts on the vertical component appear negligible. The experiment shows that current‐induced noise has a dominant direction approximately perpendicular to the currents, a pattern consistent with an actual OBS on the flank of Mariana Trench, where the current direction is supposed to be along the trench. Shielding the sensor module with a plastic casing can substantially suppress the noise, indicating that shielding is a practical, low‐cost scheme to reduce the current‐induced noise.

No abstract

No abstract

No abstract