Imaging the subsurface location of a hidden fault using autocorrelations of ambient seismic noise recorded by a dense linear seismic array

Ambient noise autocorrelations functions (ACFs) can elucidate the body-wave reflectivity of the local structure beneath single seismic stations without the need for active sources or earthquakes. However, the large-amplitude signal of the near-zero-lag time in the ACF can interfere with the target reflection(s), thereby hindering the ability to determine the precise location of hidden faults in shallow sediments. In this study, a ∼ 1.2-km-long dense linear seismic array consisting of 30 three-component geophones was deployed orthogonal to the Puqian–Qinglan Fault in Hainan province, China, to image the shallow structure around the fault. Using three components of the noise records, we calculate the ACFs beneath single stations and then subtract the average source time function from the individual ACFs to obtain the body-wave reflectivity section. We then convert the zero-offset reflectivity from the time domain to the depth domain using a constant S-wave velocity of 200 m/s. We also analyse the zero-offset reflectivity by stacking the cross-correlation functions of nearby traces originating from virtual source gathers. The results reveal that the depth to basement increases suddenly from ∼60 m to ∼110 m at ∼360 m profile distance, which we ascribe to offset on the Puqian–Qinglan Fault. This conclusion is verified by an engineering geological study, thereby demonstrating that the ambient noise technique used here may be particularly useful for imaging the near-surface basement structure.