Bandpass Filter Data to Remove Noise.

If the noise and signal are separated in the temporal frequency domain, then bandpass filtering can applied to the traces to remove such noise. Strong noise might be low-frequency surface waves, 60 Hz electrical noise, wind noise, mechanical noise from human environment, cows or sharks munching on cables, etc..

As an example, Figure 1.12 depicts a shot gather before (LHS CSG) and after 50-280 Hz bandpass filtering (RHS image). Much of the low frequency groundroll (i.e., surface waves) has been eliminated to reveal many reflection events with pseudo-hyperbolic trajectories.

  

Figure 1.12: CSG seismograms before (LHS) and after (RHS) 50-280 Hz bandpass filtering. The low frequency surface waves have been suppressed.